From acesouth at frii.com Mon Oct 25 21:49:55 1999 From: acesouth at frii.com (David Willcox) Date: Mon, 25 Oct 1999 21:49:55 -0600 Subject: [lug] Chronology settings Message-ID: <005801bf1f65$2dfd5bc0$219e11d8@davewill> Does anyone happen to know and I know it's probably quite simple. I am trying to set up (my already tested 'tar backup script' from Unix OS) to be set to use every day at 6:00 AM/. Where and how do I install the date;time;etc. for this script to run. In Xenix/Unix it was a crontab file. Thanks ahead for any help on this, Dave Willcox :-) -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Daivd Willcox.vcf Type: text/x-vcard Size: 562 bytes Desc: not available URL: From fschmid at archenergy.com Mon Oct 25 22:00:54 1999 From: fschmid at archenergy.com (Ferdinand P. Schmid) Date: Mon, 25 Oct 1999 22:00:54 -0600 Subject: [lug] Chronology settings References: <005801bf1f65$2dfd5bc0$219e11d8@davewill> Message-ID: <38152776.19A7653B@archenergy.com> An HTML attachment was scrubbed... URL: From phil at sysmgrs.com Mon Oct 25 22:27:39 1999 From: phil at sysmgrs.com (Phil Hildebrand) Date: Mon, 25 Oct 1999 22:27:39 -0600 Subject: [lug] Chronology settings References: <005801bf1f65$2dfd5bc0$219e11d8@davewill> <38152776.19A7653B@archenergy.com> Message-ID: <38152DBB.D3B07125@sysmgrs.com> You can also use crontab -e from the command line as root. (will throw you into the default editor). crontab -l to get current listings. "Ferdinand P. Schmid" wrote: > That is correct. In redhat Linux this file is located at: > /etc/crontab > > The format for entries into that file is: > # "minute" "hour" "dayofmonth" "month" "day of week" "script and shell > commands go here", > > with commands looking like this: > root /etc/scripts/updateMcAfee > (that is root is the userID that this job runs as and then your script > file). > > Ferdinand > > David Willcox wrote: > >> Does anyone happen to know and I know it's probably quite simple. I >> am tryingto set up (my already tested 'tar backup script' from Unix >> OS) to be set to use every day at 6:00 AM/. Where and how do I >> install the date;time;etc. for thisscript to run. In Xenix/Unix it >> was a crontab file. Thanks ahead for any help on this, Dave Willcox >> :-) > > -- > Ferdinand Schmid > (Staff Engineer) > > Architectural Energy Corporation > http://www.archenergy.com > 2540 Frontier Avenue, Suite 201 > Boulder, CO 80301 > Phone: (303) 444-4149 > Fax: (303) 444-4304 > e-mail: mailto:fschmid at archenergy.com > _______________________________________________ Web Page: > http://lug.boulder.co.us Mailing List: > http://lists.lug.boulder.co.us/mailman/listinfo/lug -------------- next part -------------- An HTML attachment was scrubbed... URL: From mjhammel at graphics-muse.org Mon Oct 25 22:40:48 1999 From: mjhammel at graphics-muse.org (Michael J. Hammel) Date: Mon, 25 Oct 1999 22:40:48 -0600 (MDT) Subject: [lug] Chronology settings In-Reply-To: <38152776.19A7653B@archenergy.com> from "Ferdinand P. Schmid" at Oct 25, 99 10:00:54 pm Message-ID: <199910260440.WAA31741@feynman.graphics-muse.org> Thus spoke Ferdinand P. Schmid > > Ick. Turn off the HTML or turn on the ASCII text please. I still use elm (by choice). -- Michael J. Hammel | Women serve but to keep a man from better company. The Graphics Muse | William Wycherley, English dramatist. mjhammel at graphics-muse.org | Horner, in The Country Wife, act 1. http://www.graphics-muse.com From behapy at shinbiro.com Tue Oct 26 06:05:49 1999 From: behapy at shinbiro.com (behapy) Date: Tue, 26 Oct 1999 21:05:49 +0900 Subject: [lug] I'm looking for a kind person Message-ID: <3815991D.C1C45BFA@shinbiro.com> Hi, all There is strange network construction in my work. This works fine, but I can not understand. router+---21.101.2.1----------------21.101.2.122(IP), 255.255.255.0(SUBNETMASK), 21.101.2.1(GW) | 21.101.4.234 , 255,255,255,0 21.101.4.1 | 21.101.3.221 , 255,255,255,0 21.101.3.1 | 21.101.2.229 , 255,255,255,0 21,101.4.1(?) | . | . | . | . | +---21.101.3.1----------------21.101.3.112 , 255,255,255,0 21,101.3.1 | 21.101.3.224 , 255,255,255,0 21.101.3.1 | 21.101.3.124 , 255,255,255,0 21.101.3.1 | 21.101.4.24 , 255,255,255,0 21.101.4.1(?) | . | . | . | . | +---21.101.4.1----------------21.101.4.24 , 255,255,255,0 21.101.4.1 21.101.4.24 , 255,255,255,0 21.101.4.1 21.101.4.24 , 255,255,255,0 21.101.4.1 21.101.4.24 , 255,255,255,0 21.101.4.1 . . . . As you see, I do not think the network construction is correct. I didn't confirm that the (?) works fine. It might be not perfect but still works. The all of above works fine in my work. I think 21.101.2.1 has to be connected to 21.101.2.* and 21.101.3.1 has to be connected to 21.101.3.*. I think this is right. The above is wrong. Sometimes I see uncommon subnetmask like 255.255.255.128 at a pc game room. I want to know what subnetmask is for. So, I'm looking for a person who'll send me a book. Of course I'll pay the all of cost. The book may concern tcp/ip, subnetmask, router, hub etc.. And it must be easy to understand in general, because I'm not a native speaker. I don't think the book is very easy to understand, because it is a technology book. I don't care but I wish the book is easy. Now, I'm waiting for your help. Have a nice day. Best Regards, Jingun From hbrown at byu.edu Tue Oct 26 07:41:18 1999 From: hbrown at byu.edu (Hugh Brown) Date: Tue, 26 Oct 1999 07:41:18 -0600 Subject: [lug] I'm looking for a kind person References: <3815991D.C1C45BFA@shinbiro.com> Message-ID: <3815AF7E.86409ECB@byu.edu> The subnet mask allows the computer to know what traffic to consider local. It is a logical AND (I believe). So you would convert the IP address and the mask to binary AND them and that tells you what ranges are acceptable. Go to http://www.oreilly.com/catalog/tcp2/noframes.html for a book on TCP/IP networking. Hugh > Sometimes I see uncommon subnetmask like 255.255.255.128 at a pc game > room. I want to know what subnetmask is for. So, > > I'm looking for a person who'll send me a book. Of course I'll pay the > all of cost. The book may concern tcp/ip, subnetmask, router, hub > etc.. And it must be easy to understand in general, because I'm not a > native speaker. I don't think the book is very easy to understand, > because it is a technology book. I don't care but I wish the book is > easy. From rosterburg at uswest.net Tue Oct 26 07:57:33 1999 From: rosterburg at uswest.net (Rob) Date: Tue, 26 Oct 1999 07:57:33 -0600 Subject: [lug] Font installation question Message-ID: A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1847 bytes Desc: not available URL: From zeb at utalk.org Tue Oct 26 08:57:01 1999 From: zeb at utalk.org (Lamar Zabielski) Date: Tue, 26 Oct 1999 08:57:01 -0600 Subject: [lug] I'm looking for a kind person In-Reply-To: <3815991D.C1C45BFA@shinbiro.com> Message-ID: <4.1.19991026084811.04029eb0@pop.dimensional.com> Take a look at http://www.3com.com/nsc/501302.html and you might save yourself a buck or two. If you want to print this out in 3 PDF parts goto http://www.3com.com/nsc/501302s.html zya, Lamar At 09:05 PM 10/26/99 +0900, behapy wrote: >Hi, all > >There is strange network construction in my work. This works fine, but >I can not understand. > >router+---21.101.2.1----------------21.101.2.122(IP), >255.255.255.0(SUBNETMASK), 21.101.2.1(GW) > | 21.101.4.234 , >255,255,255,0 21.101.4.1 > | 21.101.3.221 , >255,255,255,0 21.101.3.1 > | 21.101.2.229 , >255,255,255,0 21,101.4.1(?) > | . > | . > | . > | . > | > +---21.101.3.1----------------21.101.3.112 , >255,255,255,0 21,101.3.1 > | 21.101.3.224 , >255,255,255,0 21.101.3.1 > | 21.101.3.124 , >255,255,255,0 21.101.3.1 > | 21.101.4.24 , >255,255,255,0 21.101.4.1(?) > | . > | . > | . > | . > | > +---21.101.4.1----------------21.101.4.24 , >255,255,255,0 21.101.4.1 > 21.101.4.24 , >255,255,255,0 21.101.4.1 > 21.101.4.24 , >255,255,255,0 21.101.4.1 > 21.101.4.24 , >255,255,255,0 21.101.4.1 > . > . > . > . > >As you see, I do not think the network construction is correct. I >didn't confirm that the (?) works fine. It might be not perfect but >still works. The all of above works fine in my work. > >I think 21.101.2.1 has to be connected to 21.101.2.* and 21.101.3.1 has >to be connected to 21.101.3.*. I think this is right. The above is >wrong. > >Sometimes I see uncommon subnetmask like 255.255.255.128 at a pc game >room. I want to know what subnetmask is for. So, > >I'm looking for a person who'll send me a book. Of course I'll pay the >all of cost. The book may concern tcp/ip, subnetmask, router, hub >etc.. And it must be easy to understand in general, because I'm not a >native speaker. I don't think the book is very easy to understand, >because it is a technology book. I don't care but I wish the book is >easy. > >Now, I'm waiting for your help. Have a nice day. > >Best Regards, >Jingun > >_______________________________________________ >Web Page: http://lug.boulder.co.us >Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug From fschmid at archenergy.com Tue Oct 26 09:22:33 1999 From: fschmid at archenergy.com (Ferdinand P. Schmid) Date: Tue, 26 Oct 1999 09:22:33 -0600 Subject: [lug] Chronology settings References: <199910260440.WAA31741@feynman.graphics-muse.org> Message-ID: <3815C739.FA36C879@archenergy.com> Michael: I only reply using HTML to incoming messages of that type. I guess those defaults don't work for lists - sorry about that. Ferdinand "Michael J. Hammel" wrote: > Thus spoke Ferdinand P. Schmid > > > > > > Ick. Turn off the HTML or turn on the ASCII text please. I still use elm > (by choice). > -- > Michael J. Hammel | Women serve but to keep a man from better company. > The Graphics Muse | William Wycherley, English dramatist. > mjhammel at graphics-muse.org | Horner, in The Country Wife, act 1. > http://www.graphics-muse.com > > _______________________________________________ > Web Page: http://lug.boulder.co.us > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug -- Ferdinand Schmid (Staff Engineer) Architectural Energy Corporation http://www.archenergy.com 2540 Frontier Avenue, Suite 201 Boulder, CO 80301 Phone: (303) 444-4149 Fax: (303) 444-4304 e-mail: mailto:fschmid at archenergy.com From lynnd at techangle.com Tue Oct 26 10:37:11 1999 From: lynnd at techangle.com (Lynn Danielson) Date: Tue, 26 Oct 1999 10:37:11 -0600 Subject: [lug] Font installation question References: <199910261400.IAA00644@ta5.techangle.com> Message-ID: <3815D8B7.C2AC6428@techangle.com> Rob wrote: > > Greetings, > Rob Osterburg wrote: > After following the brief instructions on the gimp site, I did NOT succeed installing the fonts. Please understand that when it comes to Linux system administration I'm total novice. So here are the steps I took, please let me if you can see where I went astray. > > 1) Working as root untarred the archive which created the /usr/X11R6/lib/X11/fonts/URW on my RedHat 6.0 system. The directory is full off files and tar reported no errors. > > The URW line follows the FontPath lines for "misc: unscaled", "75dpi: unscaled", and "100dpi: unscaled". In what appeared to me to be a distribution > dependent distinction, the Section "Files" was contained in XF86Config.eg, not XF86Config. What are the permissions on these fonts? They should be mode 666. Did you create a fonts.dir file using mkfontdir? If these are scaleable fonts, you will also need a fonts.scale file in your URW directory. You should have a Files section in XF86Config. My FontPath is a comma delimited list, but individual FontPath lines should work. I don't know what XF86Config.eg is, but unless you're running an unusual configuration of XFree86, the only config file that counts is /etc/XF86Config. Running xset as the user who started X should work fine. Lynn Danielson From cmorrison at info2000.net Tue Oct 26 10:40:52 1999 From: cmorrison at info2000.net (Charles Morrison) Date: Tue, 26 Oct 1999 10:40:52 -0600 Subject: [lug] devpts Message-ID: <99102610472004.00590@morrison.home> Greetings, I've had some problems that stem from compiling a kernel (2.2.10) and not getting devpts included. I've searched throughout the kernel and haven't found this as an option, but when I compile and then boot from the new kernel, I get error messages on bootup that it has no support for filesystem devpts when mounting my disk's partitions. While it appears to boot normally other than those messages, when I shutdown the box, it is unable to unmount / and upon rebooting (with the original mandrake kernel) it requires that I repair the partition. I have lost data in the process. could anyone enlighten me on how to get this compiled in? I usually use "make menuconfig'" when I set up the compile options. -- Chuck Morrison Sr. Systems Engineer VA Linux Inc. From mjhammel at graphics-muse.org Tue Oct 26 11:44:49 1999 From: mjhammel at graphics-muse.org (Michael J. Hammel) Date: Tue, 26 Oct 1999 11:44:49 -0600 (MDT) Subject: [lug] Chronology settings In-Reply-To: <3815C739.FA36C879@archenergy.com> from "Ferdinand P. Schmid" at Oct 26, 99 09:22:33 am Message-ID: <199910261744.LAA00663@feynman.graphics-muse.org> Thus spoke Ferdinand P. Schmid > Michael: > I only reply using HTML to incoming messages of that type. I guess those defaults > don't work for lists - sorry about that. Not a big deal. Its just a little hard on my eyes since I use a text based reader. As to whether the HTML works for lists, its a matter of who is reading them. For many Linux user group lists, you'll find a lot of old Unix hands on board who prefer ordinary text files. Other lists may be more modern. :-) -- Michael J. Hammel | "I don't like people. They can't be easily The Graphics Muse | categorized, referenced or explained." mjhammel at graphics-muse.org | Michael "Spooky" Hammel http://www.graphics-muse.com From mjhammel at graphics-muse.org Tue Oct 26 11:55:49 1999 From: mjhammel at graphics-muse.org (Michael J. Hammel) Date: Tue, 26 Oct 1999 11:55:49 -0600 (MDT) Subject: [lug] Font installation question In-Reply-To: <199910261400.HAA15013@graphics-muse.org> from "Rob" at Oct 26, 99 07:57:33 am Message-ID: <199910261757.LAA00701@feynman.graphics-muse.org> Thus spoke Rob > 1) Working as root untarred the archive which created the > /usr/X11R6/lib/X11/fonts/URW on my RedHat 6.0 system. The > directory is full off files and tar reported no errors. Thats good. > 2) Again as root, I added added the FontPath line to the Section > "Files" as specified in the gimp directions: > > FontPath "/usr/X11R6/lib/X11/fonts/URW" > > > The URW line follows the FontPath lines for "misc: unscaled", > "75dpi: unscaled", and "100dpi: unscaled". In what appeared to me > to be a distribution dependent distinction, the Section "Files" was > contained in XF86Config.eg, not XF86Config. Overkill. You don't necessarily need to update the XF86Config file, but I suppose that works. I just use xset fp+ /usr/X11R6/lib/X11/fonts/URW from the command line. That should make the fonts available immediately, even if you're already running under X (actually, I install new fonts under /usr/local/fonts, but thats really a user preference thing). > 3) Logged out and restarted the machine. After logging in as a Restarting the machine is not necessary - you only need to restart the X session, which means logging out of X and then logging back in. > regular user, I used xfontsel to see the new set of fonts were being > available. Unfortunately, I didn't find any URW foundry entries, nor > a dingbats family. Oh, the java problem remained as well. xfontsel isn't the easiest tool to use. Its often easier to start up the Gimp and look through its font list since the font names are more human readable. There are some other tools for previewing fonts out there (I think KDE has one thats supposed to be pretty good) but I just use the Gimp. > Do I need to run xset as root? Is there a better way to tell if the > fonts were successfully installed than running xfontsel? Your > suggestions are appreciated. No, xset works for the current user and the current X session. Adding the FontPath entry to XF86Config is supposed to make sure the font path is added each time you restart X. When you log back into your X session, in an xterm run "xset q". At the bottom of the output you'll see the current paths that the server is using for finding fonts. If your URW path is not listed, then that may be the problem. If it is listed, then you may need to run mkfontdir in the URW directory and restart the X session. At least I think its mkfontdir - I use a script called Type1Inst, but thats not available on most distributions (you have to grab it off the net). mkfontdir is delivered with all installations of X. -- Michael J. Hammel | "I don't like people. They can't be easily The Graphics Muse | categorized, referenced or explained." mjhammel at graphics-muse.org | Michael "Spooky" Hammel http://www.graphics-muse.com From shawd at hex.net Tue Oct 26 12:55:07 1999 From: shawd at hex.net (Doug Shaw) Date: Tue, 26 Oct 1999 12:55:07 -0600 Subject: [lug] devpts In-Reply-To: <99102610472004.00590@morrison.home> Message-ID: <3.0.6.32.19991026125507.00855aa0@pop3.hex.net> At 10:40 AM 10.26.1999 -0600, you wrote: >I've had some problems that stem from compiling a kernel (2.2.10) and not >getting devpts included. I've searched throughout the kernel and haven't found >this as an option, but when I compile and then boot from the new kernel, I get >error messages on bootup that it has no support for filesystem devpts when >mounting my disk's partitions. Pts is a method of defining and accessing terminals. It has been included in newer version of Linux to work toward Unix98 branding. Pts inclusion in the kernel is under 'Character Devices.' The help for that option recommends also adding the '/dev/pts filesystem for Unix98 PTYs' under 'Filesystems.' It sounds as if you've not done this latter step. Caveat: this is from a 2.2.8 kernel. Doug From acesouth at frii.com Tue Oct 26 12:55:10 1999 From: acesouth at frii.com (David Willcox) Date: Tue, 26 Oct 1999 12:55:10 -0600 Subject: [lug] Ca;n't access CDROM Message-ID: <001501bf1fe3$f2d2c600$049e11d8@davewill> Does anyone know what I'm doing wrong. I can;t seem to mount my cdrom after a new Linux 6.0 install. It shows I have it in 'etc'fstab as shown below. /dev/cdrom /mnt/cdrom iso9660 noauto,ro 0 0 And I have it as a directory under '/mnt' And I don't know the syntax to create a symbloic link or what to link it to if linking was my problem. I try 'mount /mnt/cdrom' I try 'mount /mnt/cdrom /dev/cdrom' Everything I try I get 'not a valid block device' Thanks ahead for any help, Dave Willcox :-) -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Daivd Willcox.vcf Type: text/x-vcard Size: 562 bytes Desc: not available URL: From Mike.Vigorita at commerce.com Tue Oct 26 13:15:06 1999 From: Mike.Vigorita at commerce.com (Mike Vigorita) Date: Tue, 26 Oct 1999 13:15:06 -0600 Subject: [lug] Ca;n't access CDROM In-Reply-To: <001501bf1fe3$f2d2c600$049e11d8@davewill> Message-ID: <000301bf1fe6$697c5bc0$5430460a@commerce.com> I realize that this may be an idiotic suggestion, but do you have a CD in the drive before executing the "mount" command? And, is the CD read-able by the drive? Viggy -----Original Message----- From: lug-admin at lug.boulder.co.us [mailto:lug-admin at lug.boulder.co.us]On Behalf Of David Willcox Sent: Tuesday, October 26, 1999 12:55 PM To: LINUX User group Subject: [lug] Ca;n't access CDROM Does anyone know what I'm doing wrong. I can;t seem to mount my cdrom after a new Linux 6.0 install. It shows I have it in 'etc'fstab as shown below. /dev/cdrom /mnt/cdrom iso9660 noauto,ro 0 0 And I have it as a directory under '/mnt' And I don't know the syntax to create a symbloic link or what to link it to if linking was my problem. I try 'mount /mnt/cdrom' I try 'mount /mnt/cdrom /dev/cdrom' Everything I try I get 'not a valid block device' Thanks ahead for any help, Dave Willcox :-) -------------- next part -------------- An HTML attachment was scrubbed... URL: From HowelJS at LOUISVILLE.STORTEK.COM Tue Oct 26 13:22:07 1999 From: HowelJS at LOUISVILLE.STORTEK.COM (Howell, Jeff S) Date: Tue, 26 Oct 1999 13:22:07 -0600 Subject: [lug] Can't access CDROM Message-ID: <5C97A373607ED31188CD002048406DEE12772F@lsv-msg05.stortek.com> A few other questions... is /dev/cdrom linked to the right /dev/hdx or /dev/srx device for your cd? if you're mounting it as a user, have you added the 'user' (Next to noauto,ro in the example you gave) option to mount in fstab? -----Original Message----- From: Mike Vigorita [mailto:Mike.Vigorita at commerce.com] Sent: Tuesday, October 26, 1999 1:15 PM To: lug at lug.boulder.co.us Subject: RE: [lug] Ca;n't access CDROM I realize that this may be an idiotic suggestion, but do you have a CD in the drive before executing the "mount" command? And, is the CD read-able by the drive? Viggy -----Original Message----- From: lug-admin at lug.boulder.co.us [mailto:lug-admin at lug.boulder.co.us]On Behalf Of David Willcox Sent: Tuesday, October 26, 1999 12:55 PM To: LINUX User group Subject: [lug] Ca;n't access CDROM Does anyone know what I'm doing wrong. I can;t seem to mount my cdrom after a new Linux 6.0 install. It shows I have it in 'etc'fstab as shown below. /dev/cdrom /mnt/cdrom iso9660 noauto,ro 0 0 And I have it as a directory under '/mnt' And I don't know the syntax to create a symbloic link or what to link it to if linking was my problem. I try 'mount /mnt/cdrom' I try 'mount /mnt/cdrom /dev/cdrom' Everything I try I get 'not a valid block device' Thanks ahead for any help, Dave Willcox :-) From bcollins at csd.net Tue Oct 26 14:40:42 1999 From: bcollins at csd.net (Bob Collins) Date: Tue, 26 Oct 1999 14:40:42 -0600 Subject: [lug] devpts References: <99102610472004.00590@morrison.home> Message-ID: <381611CA.EAA62AB3@csd.net> Hi, I had the same thing happen. I had put devpts in the kernel and later removed it. It is under "character devices". Delete the entry for it from /etc/fstab and you will no longer get the error message. Charles Morrison wrote: > > Greetings, > I've had some problems that stem from compiling a kernel (2.2.10) and not > getting devpts included. I've searched throughout the kernel and haven't found > this as an option, but when I compile and then boot from the new kernel, I get > error messages on bootup that it has no support for filesystem devpts when > mounting my disk's partitions. > > While it appears to boot normally other than those messages, when I shutdown > the box, it is unable to unmount / and upon rebooting (with the original > mandrake kernel) it requires that I repair the partition. I have lost data in > the process. > > could anyone enlighten me on how to get this compiled in? I usually use "make > menuconfig'" when I set up the compile options. > > -- > Chuck Morrison > Sr. Systems Engineer > VA Linux Inc. > > _______________________________________________ > Web Page: http://lug.boulder.co.us > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug -- Regards, Bob Collins bcollins at csd.net From harkerii at sprintmail.com Tue Oct 26 14:41:13 1999 From: harkerii at sprintmail.com (John Edwards) Date: Tue, 26 Oct 1999 14:41:13 -0600 Subject: [lug] Chronology settings In-Reply-To: <199910261744.LAA00663@feynman.graphics-muse.org> References: <3815C739.FA36C879@archenergy.com> Message-ID: <3.0.5.32.19991026144113.007c5b70@pop.a001.sprintmail.com> At 11:44 AM 10/26/1999 -0600, you wrote: > >Not a big deal. Its just a little hard on my eyes since I use a text based >reader. As to whether the HTML works for lists, its a matter of who is >reading them. For many Linux user group lists, you'll find a lot of old >Unix hands on board who prefer ordinary text files. > >Other lists may be more modern. :-) I don't think it's the lists, but rather the email program one is using. Myself, since I still read mail on my {Gasp!} windows boot, I didn't even notice it. Other email programs don't translate the html code. Just my .02 John If Life is just a highway, then the soul is just a car and objects in the rear-view window may appear closer han they are. Meatloaf From shawd at hex.net Tue Oct 26 15:23:40 1999 From: shawd at hex.net (Doug Shaw) Date: Tue, 26 Oct 1999 15:23:40 -0600 Subject: [lug] Ca;n't access CDROM In-Reply-To: <001501bf1fe3$f2d2c600$049e11d8@davewill> Message-ID: <3.0.6.32.19991026152340.0085b680@pop3.hex.net> '/sbin/lsmod' to make sure that your CDROM-related modules are loading, if you're using modules. If you're running a stock redhat-6 install, you probably are. '/sbin/lsmod' should show iso9660, and perhaps others (although I cannot recall them from memory and don't have a module-using system handy). Have you checked to make sure there IS a /dev/cdrom? It should be a symbolic link to a hd?? device, and this latter piece of information is dependent on where your CDROM drive is on your IDE controller. If you know this latter piece of information and there is no /dev/cdrom, you can set it up with: ln -s /dev/hd?? /dev/cdrom (Of course you'll fill in hd?? with the proper name). Try a specific mount command: mount -t iso9660 /dev/cdrom /mnt/cdrom or mount -t iso9660 /dev/hd?? /mnt/cdrom Rule number one in troubleshooting: trust nothing to be working as it should. To wit: in the above examples, mount should pull out the '-t iso9660' information, and more, from /etc/fstab. If, for some reason, it is failing to do so (malformed syntax in /etc/fstab?) this will provide a workaround as well as a clue. You said you had been using things like: mount /mnt/cdrom mount /mnt/cdrom /dev/cdrom mount's parameters should be , so in this latter situation you'd be trying to mount the directory /mnt/cdrom on the hardware device /dev/cdrom. That won't work. From fschmid at archenergy.com Tue Oct 26 15:23:11 1999 From: fschmid at archenergy.com (Ferdinand P. Schmid) Date: Tue, 26 Oct 1999 15:23:11 -0600 Subject: [lug] Chronology settings References: <3815C739.FA36C879@archenergy.com> <3.0.5.32.19991026144113.007c5b70@pop.a001.sprintmail.com> Message-ID: <38161BBF.44590B8E@archenergy.com> To clarify my statement: Lists referred to an unknown number of users receiving my message versus a direct response to an individual who sent text in HTML. Ferdinand John Edwards wrote: > At 11:44 AM 10/26/1999 -0600, you wrote: > > > >Not a big deal. Its just a little hard on my eyes since I use a text based > >reader. As to whether the HTML works for lists, its a matter of who is > >reading them. For many Linux user group lists, you'll find a lot of old > >Unix hands on board who prefer ordinary text files. > > > >Other lists may be more modern. :-) > > I don't think it's the lists, but rather the email program one is using. > Myself, since I still read mail on my {Gasp!} windows boot, I didn't even > notice it. Other email programs don't translate the html code. > > Just my .02 > > John > If Life is just a highway, then the soul is just a car > and objects in the rear-view window may appear closer > han they are. Meatloaf > > _______________________________________________ > Web Page: http://lug.boulder.co.us > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug -- Ferdinand Schmid (Staff Engineer) Architectural Energy Corporation http://www.archenergy.com 2540 Frontier Avenue, Suite 201 Boulder, CO 80301 Phone: (303) 444-4149 Fax: (303) 444-4304 e-mail: mailto:fschmid at archenergy.com From cmorrison at info2000.net Tue Oct 26 17:03:22 1999 From: cmorrison at info2000.net (Charles Morrison) Date: Tue, 26 Oct 1999 17:03:22 -0600 Subject: [lug] devpts References: <381611CA.EAA62AB3@csd.net> Message-ID: <99102617062003.00783@morrison.home> On Tue, 26 Oct 1999, you wrote: > Hi, > > I had the same thing happen. I had put devpts in the kernel and later > removed it. It is under "character devices". Delete the entry for it > from /etc/fstab and you will no longer get the error message. > Yes, that would get rid of the messages, but would that clean up the umount problem too? Could it be refusing to umount / because it can't umount a nonexistant devpts filesystem? I really hate trashing my / partition :-( > Charles Morrison wrote: > > > > Greetings, > > I've had some problems that stem from compiling a kernel (2.2.10) and not > > getting devpts included. I've searched throughout the kernel and haven't found > > this as an option, but when I compile and then boot from the new kernel, I get > > error messages on bootup that it has no support for filesystem devpts when > > mounting my disk's partitions. > > > > While it appears to boot normally other than those messages, when I shutdown > > the box, it is unable to unmount / and upon rebooting (with the original > > mandrake kernel) it requires that I repair the partition. I have lost data in > > the process. > > > > could anyone enlighten me on how to get this compiled in? I usually use "make > > menuconfig'" when I set up the compile options. -- Chuck Morrison Sr. Systems Engineer VA Linux Inc. From alanr at bell-labs.com Tue Oct 26 21:24:23 1999 From: alanr at bell-labs.com (Alan Robertson) Date: Tue, 26 Oct 1999 21:24:23 -0600 Subject: [lug] devpts References: <381611CA.EAA62AB3@csd.net> <99102617062003.00783@morrison.home> Message-ID: <38167067.F4B943FB@bell-labs.com> Charles Morrison wrote: > > On Tue, 26 Oct 1999, you wrote: > > Hi, > > > > I had the same thing happen. I had put devpts in the kernel and later > > removed it. It is under "character devices". Delete the entry for it > > from /etc/fstab and you will no longer get the error message. > > > > Yes, that would get rid of the messages, but would that clean up the umount > problem too? Could it be refusing to umount / because it can't umount a > nonexistant devpts filesystem? I really hate trashing my / partition :-( It will only complain about being unable to unmount things that are actually mounted. Things that aren't actually mounted aren't a problem. -- Alan Robertson alanr at bell-labs.com From wmorriso at csd.net Tue Oct 26 22:16:41 1999 From: wmorriso at csd.net (Whit Morrison) Date: Tue, 26 Oct 1999 22:16:41 -0600 Subject: [lug] Can't access CDROM References: <5C97A373607ED31188CD002048406DEE12772F@lsv-msg05.stortek.com> Message-ID: <38167CA9.A6DE0266@csd.net> Howell, Jeff S wrote: > A few other questions... is /dev/cdrom linked to the right /dev/hdx or > /dev/srx device for your cd? if you're mounting it as a user, have you added > the 'user' (Next to noauto,ro in the example you gave) option to mount in > fstab? > > -----Original Message----- > From: Mike Vigorita [mailto:Mike.Vigorita at commerce.com] > Sent: Tuesday, October 26, 1999 1:15 PM > To: lug at lug.boulder.co.us > Subject: RE: [lug] Ca;n't access CDROM > > I realize that this may be an idiotic suggestion, but do you have a CD in > the drive before executing the "mount" command? And, is the CD read-able by > the drive? > > Viggy > > -----Original Message----- > From: lug-admin at lug.boulder.co.us [mailto:lug-admin at lug.boulder.co.us]On > Behalf Of David Willcox > Sent: Tuesday, October 26, 1999 12:55 PM > To: LINUX User group > Subject: [lug] Ca;n't access CDROM > > Does anyone know what I'm doing wrong. I can;t seem to mount my cdrom > after a new Linux 6.0 install. It shows I have it in 'etc'fstab as shown > below. > > /dev/cdrom /mnt/cdrom iso9660 noauto,ro 0 0 > > And I have it as a directory under '/mnt' > And I don't know the syntax to create a symbloic link or what to link it to > if linking > was my problem. > > I try 'mount /mnt/cdrom' > I try 'mount /mnt/cdrom /dev/cdrom' > > Everything I try I get 'not a valid block device' > > Thanks ahead for any help, > > Dave Willcox :-) > > > _______________________________________________ > Web Page: http://lug.boulder.co.us > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug Have you tried "mount /dev/cdrom /mnt/cdrom I think the device is identified before the location. From gmasters at devcg.denver.co.us Wed Oct 27 07:19:16 1999 From: gmasters at devcg.denver.co.us (Gary Masters) Date: Wed, 27 Oct 1999 07:19:16 -0600 (MDT) Subject: [lug] Font installation question In-Reply-To: <199910261757.LAA00701@feynman.graphics-muse.org> Message-ID: On Tue, 26 Oct 1999, Michael J. Hammel wrote: > > 2) Again as root, I added added the FontPath line to the Section > > "Files" as specified in the gimp directions: > > > > FontPath "/usr/X11R6/lib/X11/fonts/URW" > > > > > > The URW line follows the FontPath lines for "misc: unscaled", > > "75dpi: unscaled", and "100dpi: unscaled". In what appeared to me > > to be a distribution dependent distinction, the Section "Files" was > > contained in XF86Config.eg, not XF86Config. > > Overkill. You don't necessarily need to update the XF86Config file, but I > suppose that works. I just use > Doesn't RedHat 6.0 use the font server (xfs)? Will it work correctly to have part of the fonts accessed via xfs and others via a more typical FontPath directive? If the answer to the second question is no, perhaps it would work out better to add that font to the xfs configuration, and restart xfs. -- Gary Masters gmasters at devcg.denver.co.us From khchan at cyberdude.com Tue Oct 26 13:21:30 1999 From: khchan at cyberdude.com (Chan Kar Heng) Date: Wed, 27 Oct 1999 03:21:30 +0800 Subject: [lug] ppp on gnome dsktop In-Reply-To: <3812F29B.2CF77C8@bell-labs.com> References: <99102322451100.00462@spiritship.localdomain> <3.0.6.32.19991024162654.007ebdf0@pop.SoftHome.net> Message-ID: <3.0.6.32.19991027032130.007fa5d0@pop.SoftHome.net> hi thanks a million for the tip! i figured half how to configure the thing, but not the other half.. it doesn't seem to work yet. and i can't seem to access the help files too.. sigh. the lights on the modem lights applet don't work.. i can up a connection using modem lights but not bring it down. don't know how to configure ppp dialer applet as well. and the way i make modem lights applet up/down an interface is by executing /etc/sysconfig/network-scripts/ifup ppp0 pointers pls? thanks in advance. khchan >> how do i open a dialup (ppp) connection to my isp from within gnome? >> anything like usernet or kppp? perhaps something called gppp? >> (i tried searching for something called gppp, non-existent though) >> >> tried using usernet.. doesn't work. noticed that there's an applet >> in gnome that has a play/pause and stop button, but can't seem to >> find where to configure it. >> >> dug around some HOWTOs and FAQs too.. no mention. then browsed thru >> the whole gnome online help. nothing on it... >> >> configured ppp0 in linuxconf though. >> >> i'm sure i can set up the connection manually (scripts, ifconfig, pppd, >> etc), but thought i should just try out stuffs on gnome. > >I use the cool applet that comes with Gnome for this purpose. It's called the >RH PPP dialer. You can add it by going to the gnome foot, and going to >Panel/Add Applet/Network/RH PPP Dialer. It will then add an amazing little >applet to the panel which has a button for connect/disconnect, modem lights and >a PPP throughput bar graph. > -------------------------------------------- i don't suffer from stress... i'm a carrier. http://home.backroom.net/~bozo From khchan at cyberdude.com Wed Oct 27 08:06:26 1999 From: khchan at cyberdude.com (Chan Kar Heng) Date: Wed, 27 Oct 1999 22:06:26 +0800 Subject: [lug] a bit tired of updates In-Reply-To: <99102610472004.00590@morrison.home> Message-ID: <3.0.6.32.19991027220626.007f8750@pop.SoftHome.net> greetings. just to voice an irritating feeling which i feel most linux users share... i'm quite sure we all know sometimes why we're get .rpm files and not .tar.gz or .tgz files... coz .rpm provides dependency checks for one. so .rpm it is. but do you often find out about a new app or a new version of a favourite app and proceed to download the rpm, only to find that you need an rpm of a (set) of libraries which need to be upgraded as well. so you proceed to search and download those rpms required for your new app and install em.. only to find that some other apps are dependent on your old libraries which you are about to replace. and so you proceed to search and download a newer version of that app as well. (unless you choose to break the dependency by forcing the install, and getting your app semi-broken) imho, this is all rather frustrating, esp while following thru many versions of linux distribution (i can't claim i did so far, and i'm not a heavy user too, tell by the mail headers.. heh!). so does anyone have a certain recipe for overcoming this rather irritating thing? (though it only happens once in a month or few) regards, khchan ps: linux is still cool! -------------------------------------------- i don't suffer from stress... i'm a carrier. http://home.backroom.net/~bozo From kirk at fpcc.net Wed Oct 27 08:15:33 1999 From: kirk at fpcc.net (Kirk Rafferty) Date: Wed, 27 Oct 1999 08:15:33 -0600 (MDT) Subject: [lug] a bit tired of updates In-Reply-To: <3.0.6.32.19991027220626.007f8750@pop.SoftHome.net> Message-ID: I don't really think there's a way around what you're describing. Some of the better sites point out dependencies and where you can get them, and that's always nice. Also, if you're sure you have a dependency installed (maybe you cicumvented rpm) and the package complains it needs it, there's always the "--force" option (my favorite "power" command :). RPM at least makes it easy to know what you need. Ever try to install a .tar.gz package and get a bizarre message that it can't find libfoo.so or something? Cripes, then you have to actually go back and read the README.first file! :-) -- _/ Kirk Rafferty - Fairplay Communications - Kirk at fpcc.net - 303.363.8810 _/ _/ Fast Internet access for $14.95/month -- http://www.fpcc.net _/ _/ 10Mb Web Space, 2 Email addresses, your own hostname, Linux shell _/ _/ Fairplay Communications uses Open Source Operating Systems exclusively _/ On Wed, 27 Oct 1999, Chan Kar Heng wrote: > greetings. > > just to voice an irritating feeling which i feel most linux > users share... > > i'm quite sure we all know sometimes why we're get .rpm > files and not .tar.gz or .tgz files... coz .rpm provides > dependency checks for one. > > so .rpm it is. but do you often find out about a new app > or a new version of a favourite app and proceed to download > the rpm, only to find that you need an rpm of a (set) of > libraries which need to be upgraded as well. > > so you proceed to search and download those rpms required > for your new app and install em.. only to find that some > other apps are dependent on your old libraries which you > are about to replace. and so you proceed to search and download > a newer version of that app as well. (unless you choose > to break the dependency by forcing the install, and getting > your app semi-broken) > > imho, this is all rather frustrating, esp while following > thru many versions of linux distribution (i can't claim > i did so far, and i'm not a heavy user too, tell by the mail > headers.. heh!). > > so does anyone have a certain recipe for overcoming this > rather irritating thing? (though it only happens once in > a month or few) > > regards, > khchan > > ps: linux is still cool! > > > > > > > > -------------------------------------------- > i don't suffer from stress... i'm a carrier. > http://home.backroom.net/~bozo > > _______________________________________________ > Web Page: http://lug.boulder.co.us > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug > From jbs at quiotix.com Wed Oct 27 08:18:25 1999 From: jbs at quiotix.com (Jeffrey B. Siegal) Date: Wed, 27 Oct 1999 07:18:25 -0700 Subject: [lug] a bit tired of updates References: <3.0.6.32.19991027220626.007f8750@pop.SoftHome.net> Message-ID: <381709B1.AE54623C@quiotix.com> Chan Kar Heng wrote: > so does anyone have a certain recipe for overcoming this > rather irritating thing? (though it only happens once in > a month or few) This doesn't happen if you stick to the official Red Hat updates (and probably some of the other distributions as well). But the you can't get the latest and greatest until they come out with a new release (every six months or so), since their updates are only for serious problems. From kirk at fpcc.net Wed Oct 27 08:18:29 1999 From: kirk at fpcc.net (Kirk Rafferty) Date: Wed, 27 Oct 1999 08:18:29 -0600 (MDT) Subject: [lug] a bit tired of updates In-Reply-To: Message-ID: Oops, meant to say "--nodeps" instead of "--force" option to rpm. -k On Wed, 27 Oct 1999, Kirk Rafferty wrote: > > I don't really think there's a way around what you're describing. Some > of the better sites point out dependencies and where you can get them, > and that's always nice. Also, if you're sure you have a dependency > installed (maybe you cicumvented rpm) and the package complains it needs > it, there's always the "--force" option (my favorite "power" command :). > > RPM at least makes it easy to know what you need. Ever try to install > a .tar.gz package and get a bizarre message that it can't find libfoo.so > or something? Cripes, then you have to actually go back and read the > README.first file! :-) > > -- > _/ Kirk Rafferty - Fairplay Communications - Kirk at fpcc.net - 303.363.8810 _/ > _/ Fast Internet access for $14.95/month -- http://www.fpcc.net _/ > _/ 10Mb Web Space, 2 Email addresses, your own hostname, Linux shell _/ > _/ Fairplay Communications uses Open Source Operating Systems exclusively _/ > > On Wed, 27 Oct 1999, Chan Kar Heng wrote: > > > greetings. > > > > just to voice an irritating feeling which i feel most linux > > users share... > > > > i'm quite sure we all know sometimes why we're get .rpm > > files and not .tar.gz or .tgz files... coz .rpm provides > > dependency checks for one. > > > > so .rpm it is. but do you often find out about a new app > > or a new version of a favourite app and proceed to download > > the rpm, only to find that you need an rpm of a (set) of > > libraries which need to be upgraded as well. > > > > so you proceed to search and download those rpms required > > for your new app and install em.. only to find that some > > other apps are dependent on your old libraries which you > > are about to replace. and so you proceed to search and download > > a newer version of that app as well. (unless you choose > > to break the dependency by forcing the install, and getting > > your app semi-broken) > > > > imho, this is all rather frustrating, esp while following > > thru many versions of linux distribution (i can't claim > > i did so far, and i'm not a heavy user too, tell by the mail > > headers.. heh!). > > > > so does anyone have a certain recipe for overcoming this > > rather irritating thing? (though it only happens once in > > a month or few) > > > > regards, > > khchan > > > > ps: linux is still cool! > > > > > > > > > > > > > > > > -------------------------------------------- > > i don't suffer from stress... i'm a carrier. > > http://home.backroom.net/~bozo > > > > _______________________________________________ > > Web Page: http://lug.boulder.co.us > > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug > > > > > > _______________________________________________ > Web Page: http://lug.boulder.co.us > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug > From cool at eklektix.com Wed Oct 27 08:25:10 1999 From: cool at eklektix.com (Elizabeth O. Coolbaugh) Date: Wed, 27 Oct 1999 08:25:10 -0600 Subject: [lug] a bit tired of updates In-Reply-To: <381709B1.AE54623C@quiotix.com>; from Jeffrey B. Siegal on Wed, Oct 27, 1999 at 07:18:25AM -0700 References: <3.0.6.32.19991027220626.007f8750@pop.SoftHome.net> <381709B1.AE54623C@quiotix.com> Message-ID: <19991027082510.I4071@net15-cust107.eklektix.com> > This doesn't happen if you stick to the official Red Hat updates (and > probably some of the other distributions as well). But the you can't > get the latest and greatest until they come out with a new release > (every six months or so), since their updates are only for serious > problems. Actually, it just happened to me recently when trying to update my gnome packages using the official updates made available on the web site ... I saw the error message, the rpms did not get updated and I haven't gotten back to try and work out what packages were missing. This is something that I have heard, though not tested, is actually better with the Debian .pkg format. It supposedly tells you what *package* is missing instead of just what library or binary and has options to just tell it to install required dependencies automatically. Again, this isn't from personal experience, but I personally think it should be part of any package management system, to at least give you that capability as an option. Liz Coolbaugh cool at lwn.net From rob at pangalactic.org Wed Oct 27 08:30:42 1999 From: rob at pangalactic.org (Rob Riggs) Date: Wed, 27 Oct 1999 08:30:42 -0600 Subject: [lug] a bit tired of updates References: <3.0.6.32.19991027220626.007f8750@pop.SoftHome.net> Message-ID: <38170C92.43432B89@pangalactic.org> Chan Kar Heng wrote: ... > so .rpm it is. but do you often find out about a new app > or a new version of a favourite app and proceed to download > the rpm, only to find that you need an rpm of a (set) of > libraries which need to be upgraded as well. ... > so does anyone have a certain recipe for overcoming this > rather irritating thing? (though it only happens once in > a month or few) 99% of the time you can overcome this by grabbing the SRPM and rebuilding the RPM from source. The RPM only requires the newer libraries because that is what it was compiled against. However, there are times when you will realize that it is just easier to upgrade your distribution to the latest release. I finally upgraded to Red Hat 6.0 recently, because that is what most of the Gnome developers were using. It got to the point that upgrading the OS was easier than trying to keep up upgrading libraries everytimg I wanted to try out a new application. -- Rob Riggs Technical Staff Tummy.com, Ltd. http://www.tummy.com/ From jbs at quiotix.com Wed Oct 27 08:37:09 1999 From: jbs at quiotix.com (Jeffrey B. Siegal) Date: Wed, 27 Oct 1999 07:37:09 -0700 Subject: [lug] a bit tired of updates References: <3.0.6.32.19991027220626.007f8750@pop.SoftHome.net> <381709B1.AE54623C@quiotix.com> <19991027082510.I4071@net15-cust107.eklektix.com> Message-ID: <38170E15.4F6B31AC@quiotix.com> "Elizabeth O. Coolbaugh" wrote: > Actually, it just happened to me recently when trying to update my > gnome packages using the official updates made available on the > web site ... I saw the error message, the rpms did not get updated > and I haven't gotten back to try and work out what packages were > missing. Did you download *all* of the rpms listed in the errata and update them at the same time (with single invokation of the rpm command referencing all of the rpm's on the command line)? This is critical when updating a complex multi-package facility like GNOME (or XFree86). If you did this, then you've identified a problem that should be reported to Red Hat. I haven't heard of this before. Yeah, the Debian package stuff is better. From wallen at boulder.nist.gov Wed Oct 27 08:43:42 1999 From: wallen at boulder.nist.gov (Wayde Allen) Date: Wed, 27 Oct 1999 08:43:42 -0600 (MDT) Subject: [lug] a bit tired of updates In-Reply-To: <19991027082510.I4071@net15-cust107.eklektix.com> Message-ID: > This is something that I have heard, though not tested, is actually > better with the Debian .pkg format. It supposedly tells you what > *package* is missing instead of just what library or binary and > has options to just tell it to install required dependencies > automatically. That is correct. In general I prefer the Debian package management to Red Hat's. I think there are more available packages in .rpm format though so that becomes a bit of a problem. - Wayde (wallen at boulder.nist.gov) From ruddy at udlug.org Wed Oct 27 08:41:16 1999 From: ruddy at udlug.org (Bob C. Ruddy) Date: Wed, 27 Oct 1999 10:41:16 -0400 (EDT) Subject: [lug] a bit tired of updates In-Reply-To: <38170C92.43432B89@pangalactic.org> Message-ID: Anyone who is tired of upgrading an rpm system should try debian and the apt tool. First the debian package format contains much more information about the packages it depends on. Second there is a package tool called apt. This tool is a front end to the debian package tool. There is a config file for apt inside the config file you give apt a list of urls were it can find the debian packages you want. For example you can say use ftp.debian.org. Then you run 'apt-get update' that updates apt's list of available packages. After that you run 'apt-get upgrade' that will download all the updates for your system and install them. Just two commands and you system is updated. If you want to install another program, say xemacs. 'apt-get install xemacs' will get xemacs find the dependacies, download them, and install them as well. You'd think I was a sales person or somthing? :-) Bob On Wed, 27 Oct 1999, Rob Riggs wrote: ->Chan Kar Heng wrote: ->... ->> so .rpm it is. but do you often find out about a new app ->> or a new version of a favourite app and proceed to download ->> the rpm, only to find that you need an rpm of a (set) of ->> libraries which need to be upgraded as well. ->... ->> so does anyone have a certain recipe for overcoming this ->> rather irritating thing? (though it only happens once in ->> a month or few) -> ->99% of the time you can overcome this by grabbing the SRPM ->and rebuilding the RPM from source. The RPM only requires ->the newer libraries because that is what it was compiled ->against. -> ->However, there are times when you will realize that it is ->just easier to upgrade your distribution to the latest ->release. I finally upgraded to Red Hat 6.0 recently, ->because that is what most of the Gnome developers were ->using. It got to the point that upgrading the OS was easier ->than trying to keep up upgrading libraries everytimg I ->wanted to try out a new application. -> ->-- ->Rob Riggs ->Technical Staff ->Tummy.com, Ltd. ->http://www.tummy.com/ -> ->_______________________________________________ ->Web Page: http://lug.boulder.co.us ->Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug -> From ruddy at udlug.org Wed Oct 27 08:42:33 1999 From: ruddy at udlug.org (Bob C. Ruddy) Date: Wed, 27 Oct 1999 10:42:33 -0400 (EDT) Subject: [lug] a bit tired of updates In-Reply-To: Message-ID: On Wed, 27 Oct 1999, Wayde Allen wrote: ->That is correct. In general I prefer the Debian package management to Red ->Hat's. I think there are more available packages in .rpm format though so ->that becomes a bit of a problem. That is true but there is a program that will convert the rpm to a debian package file. It works really well to. Bob From rm at ns.aura.de Wed Oct 27 08:45:14 1999 From: rm at ns.aura.de (Ralf Mattes) Date: Wed, 27 Oct 1999 16:45:14 +0200 (MEST) Subject: [lug] a bit tired of updates In-Reply-To: Message-ID: On Wed, 27 Oct 1999, Bob C. Ruddy wrote: > > > On Wed, 27 Oct 1999, Wayde Allen wrote: > > ->That is correct. In general I prefer the Debian package management to Red > ->Hat's. I think there are more available packages in .rpm format though so > ->that becomes a bit of a problem. > > That is true but there is a program that will convert the rpm to a debian > package file. It works really well to. Hmm, it'll convert the rpm to .deb but it can't create the sophisticated dependency information required for a 'real' Debian packet. Ralf *-------------------------------------------------------------------* | | |== | Ralf Mattes | rm at schauinsland.com |== | Programming, Administration | rm at ns.aura.de |== | | |== *-------------------------------------------------------------------*== ==================================================================== From wallen at boulder.nist.gov Wed Oct 27 08:55:01 1999 From: wallen at boulder.nist.gov (Wayde Allen) Date: Wed, 27 Oct 1999 08:55:01 -0600 (MDT) Subject: [lug] a bit tired of updates In-Reply-To: Message-ID: On Wed, 27 Oct 1999, Bob C. Ruddy wrote: > > > On Wed, 27 Oct 1999, Wayde Allen wrote: > > ->That is correct. In general I prefer the Debian package management to Red > ->Hat's. I think there are more available packages in .rpm format though so > ->that becomes a bit of a problem. > > That is true but there is a program that will convert the rpm to a debian > package file. It works really well to. Yes I know. I've been using Debian for several years now. I finally got around to installing a Red Hat distribution (KRUD) to see how that works out. - Wayde (wallen at boulder.nist.gov) From PederMJ at LOUISVILLE.STORTEK.COM Wed Oct 27 08:58:14 1999 From: PederMJ at LOUISVILLE.STORTEK.COM (Pedersen, Michael J) Date: Wed, 27 Oct 1999 08:58:14 -0600 Subject: [lug] a bit tired of updates Message-ID: <0E22A7366689D31188CB002048406D5164E297@lsv-msg04.stortek.com> Well, there is another option, but it's not a popular one: Debian. Here's what happens, and why Debian's package management is better. RedHat uses the .rpm files. Debian uses .deb files. Differences: RPM files are a totally new format (which is why other distributions use a program named "alien" to read them). RPM files either don't have post-install scripts, or don't do much with them. Post-install scripts allow a program to properly configure itself, which can be a very good thing. DEB files are, as best I can recall, actually .tar.gz files with a different ending (internal format, thought, is .tar.gz). DEB files simply include some extra files to allow Debian's package manager to figure out how to install a program. DEB files have much used post-install scripts, sometimes even overused (explained in a minute). Redhat's strengths lie in the fact that it's installer is very easy (still not as easy as Caldera's, or the new Corel which isn't released yet, but very easy overall). It's RPMs are difficult to use, as you've noticed. System management under RedHat is another one of it's weaknesses (having worked with RedHat for a few years, I can say this. And no, the control-panel is NOT as easy as some of the Debian configurations). Debian's strengths lie in the fact that it is overwhelmingly easy to configure and use, AFTER it's set up. I can do one command (apt-get dist-upgrade), and assuming that I have my configuration information done correctly (which involves typing in a URL into a text file), I will upgrade my entire machine, all in one step. Furthermore, if I need to install a brand new package, I simply type in "apt-get install package" (same configuration restrictions above), and not only will the new package be downloaded and installed, but also it will download and install all dependencies on that package (ie: a gnome upgrade requires an updated library, which would break another package. The other package, the new library, and the new package will all be downloaded and installed in the order it takes to avoid breaking anything). VERY nice command to have. Debian's weaknesses lie in it's installer. While options are being worked on to make the whole install process more painless, it's still very difficult to work out properly. If you are actually picky about what goes on your system, you can spend two hours (or more) picking out which packages to install (last count was over 4000 on 2 CDs), and then the post-install scripts will take 4 or more hours to complete, due to all the questions). And that was what I meant by the post-install scripts sometimes being overused. As a result, my personal preference is Debian. While it took me several days to download all the new packages (over a modem), I was able to do so, and complete an entire system upgrade, without a new CD, without reformatting (something I did EVERY time I upgraded RedHat), without breaking anything. My suggestion? Take a look at other tools which are out there. I believe there are tools available for RedHat which function similarly to apt-get for Debian, but don't know what they are (or where, or how to get them, or ...). Good luck, and feel free to ask questions on the whole process directly to me, if you wish. Michael Pedersen > -----Original Message----- > From: Kirk Rafferty [mailto:kirk at fpcc.net] > Sent: Wednesday, October 27, 1999 8:18 AM > To: lug at lug.boulder.co.us > Subject: Re: [lug] a bit tired of updates > > > > Oops, meant to say "--nodeps" instead of "--force" option to rpm. > > -k > > On Wed, 27 Oct 1999, Kirk Rafferty wrote: > > > > > I don't really think there's a way around what you're > describing. Some > > of the better sites point out dependencies and where you > can get them, > > and that's always nice. Also, if you're sure you have a dependency > > installed (maybe you cicumvented rpm) and the package > complains it needs > > it, there's always the "--force" option (my favorite > "power" command :). > > > > RPM at least makes it easy to know what you need. Ever try > to install > > a .tar.gz package and get a bizarre message that it can't > find libfoo.so > > or something? Cripes, then you have to actually go back > and read the > > README.first file! :-) > > > > -- > > _/ Kirk Rafferty - Fairplay Communications - Kirk at fpcc.net > - 303.363.8810 _/ > > _/ Fast Internet access for $14.95/month -- > http://www.fpcc.net _/ > > _/ 10Mb Web Space, 2 Email addresses, your own hostname, > Linux shell _/ > > _/ Fairplay Communications uses Open Source Operating > Systems exclusively _/ > > > > On Wed, 27 Oct 1999, Chan Kar Heng wrote: > > > > > greetings. > > > > > > just to voice an irritating feeling which i feel most linux > > > users share... > > > > > > i'm quite sure we all know sometimes why we're get .rpm > > > files and not .tar.gz or .tgz files... coz .rpm provides > > > dependency checks for one. > > > > > > so .rpm it is. but do you often find out about a new app > > > or a new version of a favourite app and proceed to download > > > the rpm, only to find that you need an rpm of a (set) of > > > libraries which need to be upgraded as well. > > > > > > so you proceed to search and download those rpms required > > > for your new app and install em.. only to find that some > > > other apps are dependent on your old libraries which you > > > are about to replace. and so you proceed to search and download > > > a newer version of that app as well. (unless you choose > > > to break the dependency by forcing the install, and getting > > > your app semi-broken) > > > > > > imho, this is all rather frustrating, esp while following > > > thru many versions of linux distribution (i can't claim > > > i did so far, and i'm not a heavy user too, tell by the mail > > > headers.. heh!). > > > > > > so does anyone have a certain recipe for overcoming this > > > rather irritating thing? (though it only happens once in > > > a month or few) > > > > > > regards, > > > khchan > > > > > > ps: linux is still cool! > > > > > > > > > > > > > > > > > > > > > > > > -------------------------------------------- > > > i don't suffer from stress... i'm a carrier. > > > http://home.backroom.net/~bozo > > > > > > _______________________________________________ > > > Web Page: http://lug.boulder.co.us > > > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug > > > > > > > > > > > _______________________________________________ > > Web Page: http://lug.boulder.co.us > > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug > > > > > _______________________________________________ > Web Page: http://lug.boulder.co.us > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug > From cool at eklektix.com Wed Oct 27 09:12:00 1999 From: cool at eklektix.com (Elizabeth O. Coolbaugh) Date: Wed, 27 Oct 1999 09:12:00 -0600 Subject: [lug] a bit tired of updates In-Reply-To: <38170E15.4F6B31AC@quiotix.com>; from Jeffrey B. Siegal on Wed, Oct 27, 1999 at 07:37:09AM -0700 References: <3.0.6.32.19991027220626.007f8750@pop.SoftHome.net> <381709B1.AE54623C@quiotix.com> <19991027082510.I4071@net15-cust107.eklektix.com> <38170E15.4F6B31AC@quiotix.com> Message-ID: <19991027091200.K4071@net15-cust107.eklektix.com> Nope :-). In general, I install only the security updates, not all the updates. In addition, I prefer not to install packages I don't need, so I use autorpm to detect packages that have been upgraded, download them automatically and notify me of them. I'm sure the package I need for the gnome updates is in their directory. Installing all of them to find it is unreasonable. Finding the one I need is doable, but just takes time I haven't got today. Liz On Wed, Oct 27, 1999 at 07:37:09AM -0700, Jeffrey B. Siegal wrote: > "Elizabeth O. Coolbaugh" wrote: > > Actually, it just happened to me recently when trying to update my > > gnome packages using the official updates made available on the > > web site ... I saw the error message, the rpms did not get updated > > and I haven't gotten back to try and work out what packages were > > missing. > > Did you download *all* of the rpms listed in the errata and update them > at the same time (with single invokation of the rpm command referencing > all of the rpm's on the command line)? This is critical when updating a > complex multi-package facility like GNOME (or XFree86). If you did > this, then you've identified a problem that should be reported to Red > Hat. I haven't heard of this before. > > Yeah, the Debian package stuff is better. > > _______________________________________________ > Web Page: http://lug.boulder.co.us > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug From robmohr at boulder.earthnet.net Wed Oct 27 09:26:57 1999 From: robmohr at boulder.earthnet.net (Rob Mohr) Date: Wed, 27 Oct 1999 09:26:57 -0600 (MDT) Subject: [lug] Chronology settings,I use Pine. In-Reply-To: <199910261744.LAA00663@feynman.graphics-muse.org> Message-ID: Me too. But I use another tree variety, Pine. Its the only thing I have used since '94. True. But I got rid of the crank telephone and the party line last year:-) (just kidding) eof From jbs at quiotix.com Wed Oct 27 09:28:26 1999 From: jbs at quiotix.com (Jeffrey B. Siegal) Date: Wed, 27 Oct 1999 08:28:26 -0700 Subject: [lug] a bit tired of updates References: <3.0.6.32.19991027220626.007f8750@pop.SoftHome.net> <381709B1.AE54623C@quiotix.com> <19991027082510.I4071@net15-cust107.eklektix.com> <38170E15.4F6B31AC@quiotix.com> <19991027091200.K4071@net15-cust107.eklektix.com> Message-ID: <38171A1A.A731F9F8@quiotix.com> "Elizabeth O. Coolbaugh" wrote: > Installing all of them to find it is unreasonable. I wansn't so much suggesting that you do that as much as clarifying my original response to the original question of "how do I avoid these package conflict problems?" That is: one way to avoid the problems is to install all of the Red Hat-supplied updates, and only the Red Hat-supplied updates. This will avoid conflicts (as well as missing dependencies, etc.), and if it doesn't let them know and they'll fix it. They're pretty good about this sort of thing. But being a "Red Hat slave" has its own set of problems, as both you and I pointed out. From rob at pangalactic.org Wed Oct 27 10:10:15 1999 From: rob at pangalactic.org (Rob Riggs) Date: Wed, 27 Oct 1999 10:10:15 -0600 Subject: [lug] a bit tired of updates References: <0E22A7366689D31188CB002048406D5164E297@lsv-msg04.stortek.com> Message-ID: <381723E7.8976450D@pangalactic.org> Let me clarify some mis-perceptions expressed in Michael's post. "Pedersen, Michael J" wrote: > > RPM files are a totally new format (which is why other distributions use a > program named "alien" to read them). RPMs use the 'cpio' archive format. The reason for the name 'alien' is that it converts packages from foreign environments to be used locally. In fact, 'alien' also converts (rather poorly) from DEB to RPM. What is perceived to be "alien" often depends on one's perspective. (Yes, 'alien' was originally written to convert RPM to DEB, but is has transcended its roots.) > RPM files either don't have post-install scripts, or don't do much with > them. Post-install scripts allow a program to properly configure itself, > which can be a very good thing. RPMs do indeed have pre-install and post-install scripts, and many make good use of them. With many more RPMs available than DEBs, it is quite natural to find more poorly packaged RPMs. Sad, but true. However, the .specs are out there and package maintainers are usually quite happy to take patches. Some of things that Debian has going for it are that there is a much more tight-knit community of package maintainers, there are many more "officially" maintained DEB packages, and there is descent infrastruc- ture in place to report and fix bugs in packages. Many RPM packages do tend to be maintained on a more ad-hoc basis. > DEB files are, as best I can recall, actually .tar.gz files with a different > ending (internal format, thought, is .tar.gz). And, again, RPM is cpio format and can have the components extracted just as easily as a tar file. > DEB files simply include some extra files to allow Debian's package manager > to figure out how to install a program. Both DEB and RPM have very similar facilities from a package management and dependancy checking standpoint. Most of the differences that you will find relate directly to the skill of the package maintainer. > As a result, my personal preference is Debian. While it took me several > days to download all the new packages (over a modem), I was able to do so, > and complete an entire system upgrade, without a new CD, without > reformatting (something I did EVERY time I upgraded RedHat), without > breaking anything. I'm completely confused as to why you reformatted after every install. This system has been running Red Hat since 3.0.3 without ever being reformatted. I've added new drives and migrated the OS over from HD to HD as I upgrade, but don't ever recall having reformatted for an OS upgrade. One of the downsides to RPM that must be addressed now is that there are now *many* RPM based distributions, many having completely different dependancies (as we noted with a recent Mandrake install). This is something that DEBs are not likely to face any time soon. The user must be careful to download RPMs for his/her distribution. Trying to install an RPM created on a Suse system on to a Red Hat system isn't likely to work really well. -- Rob Riggs Technical Staff Tummy.com, Ltd. http://www.tummy.com/ From PederMJ at LOUISVILLE.STORTEK.COM Wed Oct 27 10:36:19 1999 From: PederMJ at LOUISVILLE.STORTEK.COM (Pedersen, Michael J) Date: Wed, 27 Oct 1999 10:36:19 -0600 Subject: [lug] a bit tired of updates Message-ID: <0E22A7366689D31188CB002048406D5164E29C@lsv-msg04.stortek.com> > -----Original Message----- > RPMs use the 'cpio' archive format. The reason for the name 'alien' > is that it converts packages from foreign environments to be used > locally. In fact, 'alien' also converts (rather poorly) from DEB > to RPM. What is perceived to be "alien" often depends on one's > perspective. (Yes, 'alien' was originally written to convert RPM > to DEB, but is has transcended its roots.) Ah, that is something I didn't know. Thank you VERY much. I'll admit to having some poor research into exactly what RPM is, but I did know that it wasn't .tar.gz, which is what a fair number of people do want (I know that I prefer it, simply because I know it better). > RPMs do indeed have pre-install and post-install scripts, and many > make good use of them. With many more RPMs available than DEBs, it > is quite natural to find more poorly packaged RPMs. Sad, but true. > However, the .specs are out there and package maintainers are > usually quite happy to take patches. Well, here's where the MAJOR difference comes in: Debian post-install scripts tend to ask questions. Sometimes, they ask a LOT of them. Sometimes, they make people think that just jumping out a 40th floor window is preferable to answering another question (during an initial install :). Anyway, I've never seen an RPM ask a single question about configuring itself. Despite their sometimes overuse on Debian, I prefer to have that done upfront, rather than reconfigure afterwards. > > DEB files are, as best I can recall, actually .tar.gz files > with a different > > ending (internal format, thought, is .tar.gz). > > And, again, RPM is cpio format and can have the components extracted > just as easily as a tar file. Once you know what the format is, yes, just as easy. Without knowing it, though, it's noticeably more confusing (as evidenced by my previous post). > > DEB files simply include some extra files to allow Debian's > package manager > > to figure out how to install a program. > > Both DEB and RPM have very similar facilities from a package > management and dependancy checking standpoint. Most of the > differences that you will find relate directly to the skill > of the package maintainer. This, again, was a result of my misunderstanding of the differences between the two package formats. As I understood it, the information being extracted was stored in the RPM, as a data block, and not as a separate file inside of an archive of files. As such, my statement should have read something like this (even though the following is wrong): DEB files include extra files inside of them which control the package configuration and installation, while RPM files use a data block which does the same. Again, while wrong, that's the message I was trying to get across (and failed badly at). > I'm completely confused as to why you reformatted after every > install. This system has been running Red Hat since 3.0.3 > without ever being reformatted. I've added new drives and > migrated the OS over from HD to HD as I upgrade, but don't > ever recall having reformatted for an OS upgrade. Well, part of the reason is due to hard drive reconfiguration (ie: repartitioning, replacement, etc). The other part is that RedHat's upgrade process has never been entirely trusted by me. And still a third part is that I didn't believe I could ever muddle through enough of the dependencies to successfully update the configuration of the system without trashing it anyway. And a final part is because I know that Debian, when it encounters a config file already in place, will ask me what to do with it. RedHat, as far as I've ever seen, will simply overwrite it, tossing the old information away entirely. > One of the downsides to RPM that must be addressed now is that > there are now *many* RPM based distributions, many having > completely different dependancies (as we noted with a recent > Mandrake install). This is something that DEBs are not likely > to face any time soon. The user must be careful to download > RPMs for his/her distribution. Trying to install an RPM created > on a Suse system on to a Red Hat system isn't likely to work > really well. That is another issue entirely, and one that I'm sorry to see happen. RedHat was my favorite for a while, but because of the reasons above, I never felt entirely safe when upgrading. And I'm sure most people will agree that upgrading pretty regularly is a nice feeling. That was the reason for looking at other distributions at all. Debian seems very nice to me to maintain and work with, while RedHat seems more difficult. Then again, this may be due to my lack of knowledge of some of the update tools and procedures which should be used under RedHat. For now, I'll stay with Debian. It gives me the flexibility I want, without the headaches of dealing with upgrades in the ways that RedHat does. I did try to upgrade from 5.1 to 6.0 manually, and it sucked. Badly. Anyway, thank you for your information, I found it very helpful. From caldodge at fpcc.net Wed Oct 27 10:48:36 1999 From: caldodge at fpcc.net (Calvin Dodge) Date: Wed, 27 Oct 1999 10:48:36 -0600 Subject: [lug] a bit tired of updates References: <0E22A7366689D31188CB002048406D5164E29C@lsv-msg04.stortek.com> Message-ID: <38172CE4.E168A80F@fpcc.net> "Pedersen, Michael J" wrote: > > a config file already in place, will ask me what to do with it. RedHat, as > far as I've ever seen, will simply overwrite it, tossing the old information > away entirely. That's interesting - every time I've done such an install recently Red Hat has saved such replaced config files with an extension of "rpmsave". A quick "locate rpmsave | grep -v [KS] | wc" just informed me that I have 22 such files on my work computer. Calvin From rm at ns.aura.de Wed Oct 27 10:46:27 1999 From: rm at ns.aura.de (Ralf Mattes) Date: Wed, 27 Oct 1999 18:46:27 +0200 (MEST) Subject: [lug] a bit tired of updates In-Reply-To: <0E22A7366689D31188CB002048406D5164E29C@lsv-msg04.stortek.com> Message-ID: On Wed, 27 Oct 1999, Pedersen, Michael J wrote: > > -----Original Message----- > > RPMs use the 'cpio' archive format. The reason for the name 'alien' > > is that it converts packages from foreign environments to be used > > locally. In fact, 'alien' also converts (rather poorly) from DEB > > to RPM. What is perceived to be "alien" often depends on one's > > perspective. (Yes, 'alien' was originally written to convert RPM > > to DEB, but is has transcended its roots.) > > Ah, that is something I didn't know. Thank you VERY much. I'll admit to > having some poor research into exactly what RPM is, but I did know that it > wasn't .tar.gz, which is what a fair number of people do want (I know that I > prefer it, simply because I know it better). Well, -deb files are NOT tar files ... They are actually .ar files, and i don't see why ar-format would be better known than tar or cpio. Only within the ar-file you can find tar-gz archives of the actuall packet contents and the controlling data. > > RPMs do indeed have pre-install and post-install scripts, and many > > make good use of them. With many more RPMs available than DEBs, it > > is quite natural to find more poorly packaged RPMs. Sad, but true. > > However, the .specs are out there and package maintainers are > > usually quite happy to take patches. > Well, here's where the MAJOR difference comes in: Debian post-install > scripts tend to ask questions. Sometimes, they ask a LOT of them. > Sometimes, they make people think that just jumping out a 40th floor window > is preferable to answering another question (during an initial install :). > Anyway, I've never seen an RPM ask a single question about configuring > itself. Despite their sometimes overuse on Debian, I prefer to have that > done upfront, rather than reconfigure afterwards. Hmm, the absoltute majority of debian packets install without a question. Only some of the 'big' packets have a configuration script (which is BTW independent from the postinstall script. You can (re)configure packets with 'dpkg --configure glumpf'). > > And, again, RPM is cpio format and can have the components extracted > > just as easily as a tar file. > Once you know what the format is, yes, just as easy. Without knowing it, > though, it's noticeably more confusing (as evidenced by my previous post). Where is the difference? You only need to know the internal format if you want to NOT use the packet manager. I doubt i would ever want to unpack a .deb file to install its components by hand (of course i often repack .deb files, but i use .deb format for software distribution). > > > DEB files simply include some extra files to allow Debian's > > package manager > > > to figure out how to install a program. > > > > Both DEB and RPM have very similar facilities from a package > > management and dependancy checking standpoint. Most of the > > differences that you will find relate directly to the skill > > of the package maintainer. [...] Ralf *-------------------------------------------------------------------* | | |== | Ralf Mattes | rm at schauinsland.com |== | Programming, Administration | rm at ns.aura.de |== | | |== *-------------------------------------------------------------------*== ==================================================================== From PederMJ at LOUISVILLE.STORTEK.COM Wed Oct 27 10:50:21 1999 From: PederMJ at LOUISVILLE.STORTEK.COM (Pedersen, Michael J) Date: Wed, 27 Oct 1999 10:50:21 -0600 Subject: [lug] a bit tired of updates Message-ID: <0E22A7366689D31188CB002048406D5164E29D@lsv-msg04.stortek.com> Again, something else I didn't know about. Fortunately, I've never claimed to be omniscient, nor omniscient about RedHat's RPM files :) > -----Original Message----- > From: Calvin Dodge [mailto:caldodge at fpcc.net] > Sent: Wednesday, October 27, 1999 10:49 AM > To: lug at lug.boulder.co.us > Subject: Re: [lug] a bit tired of updates > > > "Pedersen, Michael J" wrote: > > > > a config file already in place, will ask me what to do with > it. RedHat, as > > far as I've ever seen, will simply overwrite it, tossing > the old information > > away entirely. > > That's interesting - every time I've done such an install recently Red > Hat has saved such replaced config files with an extension of > "rpmsave". > A quick "locate rpmsave | grep -v [KS] | wc" just informed me that I > have 22 such files on my work computer. > > Calvin > > _______________________________________________ > Web Page: http://lug.boulder.co.us > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug > From PederMJ at LOUISVILLE.STORTEK.COM Wed Oct 27 11:07:32 1999 From: PederMJ at LOUISVILLE.STORTEK.COM (Pedersen, Michael J) Date: Wed, 27 Oct 1999 11:07:32 -0600 Subject: [lug] a bit tired of updates Message-ID: <0E22A7366689D31188CB002048406D5164E29E@lsv-msg04.stortek.com> > -----Original Message----- > From: Ralf Mattes [mailto:rm at ns.aura.de] > Sent: Wednesday, October 27, 1999 10:46 AM > To: 'lug at lug.boulder.co.us' > Subject: RE: [lug] a bit tired of updates > > > On Wed, 27 Oct 1999, Pedersen, Michael J wrote: > Well, -deb files are NOT tar files ... They are actually .ar > files, and > i don't see why ar-format would be better known than tar or cpio. Only > within the ar-file you can find tar-gz archives of the actuall packet > contents and the controlling data. Well, at this point, I'm starting to have to wonder if my entire message was invalid, and needs to be retracted. Thank you, though, I do appreciate the correction. > Hmm, the absoltute majority of debian packets install without > a question. > Only some of the 'big' packets have a configuration script > (which is BTW > independent from the postinstall script. You can > (re)configure packets with > 'dpkg --configure glumpf'). Well, I should have known that, but for some reason never thought about it. And you are right, the majority of them install, no questions asked. But enough of them insist on asking questions that it can drive a person crazy. Especially when apt/dpkg can't find a way to ask all the questions first, or last, but instead will install a bunch of packages, then come up to one that asks questions, then a bunch more with no questions, etc. I know, it's a silly thing, but I do wish it did that. It would be a LOT easier to handle a major installation then. > Where is the difference? You only need to know the internal format > if you want to NOT use the packet manager. I doubt i would ever want > to unpack a .deb file to install its components by hand (of course > i often repack .deb files, but i use .deb format for software > distribution). I have seen quite a few people who want to extract an rpm onto a Debian (or other) system, and they have to either have alien, or know the format. Having alien is considerably easier, but some would rather extract it by hand. Plus, as was mentioned in another message, RPM is not a truly universal format anymore. An RPM made for Suse might not work on Mandrake or on RedHat. Being able to extract it by hand is a useful utility to have. Debian doesn't have that issue (AFAIK) yet, but it could happen in the future. From dennis at bullamanka.com Wed Oct 27 10:32:36 1999 From: dennis at bullamanka.com (Dennis Keller) Date: Wed, 27 Oct 1999 10:32:36 -0600 Subject: [lug] XMMS Problems with RH6.1 - Follow up Message-ID: <38172924.1D8C38DB@bullamanka.com> Hi all- I posted a question to the list earlier in the month about XMMS not working correctly after the RH6.1 upgrade with the enlightenment wm. The symptoms were that it would simply skip through a song without playing it. I've found the solution and wanted to share it. (It's no secret, I just didn't figure it out until now). The problem is with ESD. Apparently the esd that gets installed with RH6.1 isn't as compatible with XMMS .9.5.1 as the one in RH6.0. The solution is to upgrade to ESound 0.2.15 and install the XMMS ESD plugin from the following page: http://www.xmms.org/rpmplugins.html Once they are installed, simply choose preferences in XMMS and change the ouput plugin to your new esd output plugin. Hopefully this will help someone who has run into this problem! Dennis Keller dennis at bullamanka.com From andrew at NETdelivery.com Wed Oct 27 13:12:52 1999 From: andrew at NETdelivery.com (Andrew Diederich) Date: Wed, 27 Oct 1999 13:12:52 -0600 Subject: [lug] Java timezone question Message-ID: <7114B83CE30FD31183CD00805FBE772D4EE8E8@BERLIN> Okay, this is certainly off topic, but here goes. I'm helping out a buddy who has a Java question, but I don't write code, and some of you do, so I decided to use the list for my own nefarious purposes. I'm looking for a call in Java 1.1 to get the timezone off an NT box. We can get the current system time, but I guess it doesn't get the offset from GMT with it. Since we store all timestamps in our database on GMT we need to figure out how far away we are from GMT. Our current solution is to enter the current timezone in an ini file, with the dates/times to change to and from daylight savings, but that is a hack and prone to errors. Any ideas? -- Andrew Diederich | This email was made andrew at netdelivery.com | with 100% recycled | electrons. From kenw at ihs.com Wed Oct 27 13:32:23 1999 From: kenw at ihs.com (kenw at ihs.com) Date: Wed, 27 Oct 1999 13:32:23 -0600 Subject: [lug] a bit tired of updates In-Reply-To: ; from Kirk Rafferty on Wed, Oct 27, 1999 at 08:15:33AM -0600 References: <3.0.6.32.19991027220626.007f8750@pop.SoftHome.net> Message-ID: <19991027133223.A24592@adamselene.ihs.com> * Kirk Rafferty (kirk at fpcc.net) [991027 13:16]: > > I don't really think there's a way around what you're describing. Some > of the better sites point out dependencies and where you can get them, > and that's always nice. Also, if you're sure you have a dependency > installed (maybe you cicumvented rpm) and the package complains it needs > it, there's always the "--force" option (my favorite "power" command :). For example, I just did a clean install of SuSE 6.2 and tried to install the new gnome/enlightenment (from the web.) Turns out you have to have *both* libjpeg.s0.6 and libjpeg.so.62 installed. Had to work around that one. In addition, it complained that it needed netpbm (which was installed) so I just hit the power button (--force :-) to get it working. For me, this is the first time I've tried going strictly with RPMs because of this problem. Of course, I had to do a clean install because I tried to upgrade to the new gnome/enlightment without RPMs and totally screwed up my shared libs so that 1/2 the stuff wouldn't work. I find myself in the unique position of seeing both sides of the issue at the same time right now :-) -- Ken Weinert kenw at ihs.com From mjhammel at graphics-muse.org Wed Oct 27 14:31:04 1999 From: mjhammel at graphics-muse.org (Michael J. Hammel) Date: Wed, 27 Oct 1999 14:31:04 -0600 (MDT) Subject: [lug] rsync compile on RH5.2 Message-ID: <199910272031.OAA10866@feynman.graphics-muse.org> Does anyone else have rsync compiled on a RH5.2 box? The compile succeeds but the link fails with this: main.o: In function `main': main.o(.text+0xf33): undefined reference to `ap_signal' main.o(.text+0xfdf): undefined reference to `ap_signal' main.o(.text+0xfeb): undefined reference to `ap_signal' main.o(.text+0xff7): undefined reference to `ap_signal' main.o(.text+0x1003): undefined reference to `ap_signal' main.o(.text+0x1012): more undefined references to `ap_signal' follow These come from calls to signal: signal(SIGUSR1, sigusr1_handler); But the call is fine. Some header file is mucked up, but I can't find ap_signal in any of the files in rsync, nor can I find it in any header file under /usr/lib. -- Michael J. Hammel | It isn't pollution that's harming the environment. The Graphics Muse | It's the impurities in our air and water that are mjhammel at graphics-muse.org | doing it. http://www.graphics-muse.com Former U.S. Vice-President Dan Quayle From mjhammel at graphics-muse.org Wed Oct 27 11:44:12 1999 From: mjhammel at graphics-muse.org (Michael J. Hammel) Date: Wed, 27 Oct 1999 11:44:12 -0600 (MDT) Subject: [lug] Font installation question In-Reply-To: from "Gary Masters" at Oct 27, 99 07:19:16 am Message-ID: <199910271744.LAA06432@feynman.graphics-muse.org> Thus spoke Gary Masters > Doesn't RedHat 6.0 use the font server (xfs)? Will it work correctly Don't know. I run 5.2. > to have part of the fonts accessed via xfs and others via a more > typical FontPath directive? The X server doesn't care. You tell can tell it to look in any number of places or query any number of daemons for font information. > If the answer to the second question is no, perhaps it would work out > better to add that font to the xfs configuration, and restart xfs. Dunno. I don't know what the advantage to using xfs is over using the X server directly, unless xfs supports TrueType fonts (I use xfstt for that). -- Michael J. Hammel | Knowledge comes, The Graphics Muse | but wisdom lingers. mjhammel at graphics-muse.org | Tennyson, "Locksley Hall" http://www.graphics-muse.com From mjhammel at graphics-muse.org Wed Oct 27 12:00:53 1999 From: mjhammel at graphics-muse.org (Michael J. Hammel) Date: Wed, 27 Oct 1999 12:00:53 -0600 (MDT) Subject: [lug] a bit tired of updates In-Reply-To: <38170C92.43432B89@pangalactic.org> from "Rob Riggs" at Oct 27, 99 08:30:42 am Message-ID: <199910271800.MAA06453@feynman.graphics-muse.org> Thus spoke Rob Riggs > However, there are times when you will realize that it is > just easier to upgrade your distribution to the latest > release. I finally upgraded to Red Hat 6.0 recently, > because that is what most of the Gnome developers were > using. It got to the point that upgrading the OS was easier > than trying to keep up upgrading libraries everytimg I > wanted to try out a new application. Since GNOME is a fairly young set of tools, you'll find you're going to be updating it relatively often to get the full features it promises for the long term. KDE has similar problems, but is a little more mature. Other applications have similar issues. What you end up doing is finding the tools that need to updated along with GNOME (GTK for example) or KDE and put those under their own directories under /usr/local (which is on its own partition). Then you install a new distribution, remove the GNOME/GTK, etc rpms that the install put on the system and continue on your merry way. Upgrades are then done from source into your /usr/local directories. I haven't done an GNOME build yet, but the GTK builds are pretty easy. If you install from source you often aren't required to get the latest foobar.so library - it'll work with what you have. The dependencies of RPMs can make life rough for users of rapidly evolving application-level tools. For example, I placed gtk, gtk-engines, gimp-1.0 and gimp-1.1 under /usr/local. Then I install a new RH installation. After the install finishes, I use rpm - to remove the gtk RPMs and update the /etc/ld.so.conf to include my /usr/local installations (actually, I just save a copy of ld.so.conf and copy it over the newly installed one. The only exception I currently have for this rule is Perl, since it liked upgrading itself in place, under the system directories. The bad news is I can't remember all the modules I've added, so if I ever upgrade the OS a lot of things are gonna break. *sigh* The process is moderately painless once you get used to it. The problem I have is that the stupid C libraries change on the whim of madmen, breaking commercial packages. No matter what anyone says, Applix is a much better word processor than AbiWord, StarOffice, WordPerfect or anything else out there. Until I can afford to upgrade Applix, I have to stick with RH 5.2. The version I have doesn't work on 6.x. It bugs me because similar problems has shown up a number of times in the past: migrating from a.out to elf, from libc5 to libc6, from glibc2.0 to glibc2.2. Arrggghhh! -- Michael J. Hammel | The Graphics Muse | Chinese Proverb: mjhammel at graphics-muse.org | Man who leaps from tall building, jumps to http://www.graphics-muse.com conclusion. From tromey at cygnus.com Wed Oct 27 14:32:26 1999 From: tromey at cygnus.com (Tom Tromey) Date: Wed, 27 Oct 1999 13:32:26 -0700 Subject: [lug] rsync compile on RH5.2 In-Reply-To: <199910272031.OAA10866@feynman.graphics-muse.org> References: <199910272031.OAA10866@feynman.graphics-muse.org> Message-ID: <199910272032.NAA04639@ferrule.cygnus.com.> >>>>> "Michael" == Michael J Hammel writes: Michael> But the call is fine. Some header file is mucked up, but I Michael> can't find ap_signal in any of the files in rsync, nor can I Michael> find it in any header file under /usr/lib. The headers are in /usr/include, but I assume that is just a typo here. My advice is: don't rely on grep to find the symbol for you. It could be pasted together via cpp, in which case you would never find it. Instead look at the output of "gcc -E" and use that to help you find where it is coming from. If you think it is coming from the compiler itself (doubtful in this case), use "gcc -S" and read the assembly. T From kevin at scrye.com Wed Oct 27 14:36:16 1999 From: kevin at scrye.com (Kevin Fenzi) Date: 27 Oct 1999 14:36:16 -0600 Subject: [lug] rsync compile on RH5.2 In-Reply-To: mjhammel@graphics-muse.org's message of "27 Oct 1999 14:27:22 -0600" References: <199910272031.OAA10866@feynman.graphics-muse.org> Message-ID: >>>>> "Michael" == "Michael J Hammel" writes: Michael> Does anyone else have rsync compiled on a RH5.2 box? The Michael> compile succeeds but the link fails with this: Michael> main.o: In function `main': main.o(.text+0xf33): undefined Michael> reference to `ap_signal' main.o(.text+0xfdf): undefined Michael> reference to `ap_signal' main.o(.text+0xfeb): undefined Michael> reference to `ap_signal' main.o(.text+0xff7): undefined Michael> reference to `ap_signal' main.o(.text+0x1003): undefined Michael> reference to `ap_signal' main.o(.text+0x1012): more undefined Michael> references to `ap_signal' follow Michael> These come from calls to signal: Michael> signal(SIGUSR1, sigusr1_handler); Michael> But the call is fine. Some header file is mucked up, but I Michael> can't find ap_signal in any of the files in rsync, nor can I Michael> find it in any header file under /usr/lib. -- Michael poking around on my 5.2 box the only include refrence I see for it is in /usr/include/apache/ap_config.h, where it's: #define signal(s,f) ap_signal(s,f) That file is owned by apache-devel. I have no idea why rsync would want to use something from the apache-devel package, but perhaps this helps... kevin -- Kevin Fenzi MTS, tummy.com, ltd. http://www.tummy.com/ KRUD - Kevin's Red Hat Uber Distribution From jafo at tummy.com Wed Oct 27 23:54:39 1999 From: jafo at tummy.com (Sean Reifschneider) Date: Wed, 27 Oct 1999 23:54:39 -0600 Subject: [lug] rsync compile on RH5.2 In-Reply-To: <199910272031.OAA10866@feynman.graphics-muse.org>; from Michael J. Hammel on Wed, Oct 27, 1999 at 02:31:04PM -0600 References: <199910272031.OAA10866@feynman.graphics-muse.org> Message-ID: <19991027235439.J883@tummy.com> On Wed, Oct 27, 1999 at 02:31:04PM -0600, Michael J. Hammel wrote: >Does anyone else have rsync compiled on a RH5.2 box? The compile succeeds >but the link fails with this: I've just put the RPM that I'm using on my 5.2 box on "ftp.tummy.com:/pub/tummy/RPMS". Note that as far as I recall, this was simply a re-compile (using "rpm --rebuild") of the source RPM for 6.(1? 0?). Sean -- Rocky: "Do you know what an A-Bomb is?" Bullwinkle: "Of course. ``A Bomb'' is what some people call our show." Sean Reifschneider, Inimitably Superfluous URL: HP-UX/Linux/FreeBSD/BSDOS scanning software. From mjhammel at graphics-muse.org Thu Oct 28 09:33:59 1999 From: mjhammel at graphics-muse.org (Michael J. Hammel) Date: Thu, 28 Oct 1999 09:33:59 -0600 (MDT) Subject: [lug] rsync compile on RH5.2 In-Reply-To: from "Kevin Fenzi" at Oct 27, 99 02:36:16 pm Message-ID: <199910281534.JAA13159@feynman.graphics-muse.org> Thus spoke Kevin Fenzi > poking around on my 5.2 box the only include refrence I see for it is > in /usr/include/apache/ap_config.h, where it's: > > #define signal(s,f) ap_signal(s,f) > > That file is owned by apache-devel. I have no idea why rsync would > want to use something from the apache-devel package, but perhaps this > helps... Hmmm. I doubt its including that, but one never knows till one checks. Thanks for the tip. -- Michael J. Hammel | Outside of the killings, Washington has one of The Graphics Muse | the lowest crime rates in the country." mjhammel at graphics-muse.org | Mayor Marion Barry, Washington, D.C. http://www.graphics-muse.com From mjhammel at graphics-muse.org Thu Oct 28 09:36:05 1999 From: mjhammel at graphics-muse.org (Michael J. Hammel) Date: Thu, 28 Oct 1999 09:36:05 -0600 (MDT) Subject: [lug] rsync compile on RH5.2 In-Reply-To: <199910272032.NAA04639@ferrule.cygnus.com.> from "Tom Tromey" at Oct 27, 99 01:32:26 pm Message-ID: <199910281536.JAA13170@feynman.graphics-muse.org> Thus spoke Tom Tromey > The headers are in /usr/include, but I assume that is just a typo > here. It was a typo. > My advice is: don't rely on grep to find the symbol for you. It could > be pasted together via cpp, in which case you would never find it. > Instead look at the output of "gcc -E" and use that to help you find > where it is coming from. I'll try that and see if it helps. > If you think it is coming from the compiler itself (doubtful in this > case), use "gcc -S" and read the assembly. Ick. No, I don't do assembly anymore. Gave it up for sanity sake. If I have to go that deep, I punt and look for another tool, or write it myself. -- Michael J. Hammel | Outside of the killings, Washington has one of The Graphics Muse | the lowest crime rates in the country." mjhammel at graphics-muse.org | Mayor Marion Barry, Washington, D.C. http://www.graphics-muse.com From mjhammel at graphics-muse.org Thu Oct 28 09:37:34 1999 From: mjhammel at graphics-muse.org (Michael J. Hammel) Date: Thu, 28 Oct 1999 09:37:34 -0600 (MDT) Subject: [lug] rsync compile on RH5.2 In-Reply-To: <19991027235439.J883@tummy.com> from "Sean Reifschneider" at Oct 27, 99 11:54:39 pm Message-ID: <199910281537.JAA13181@feynman.graphics-muse.org> Thus spoke Sean Reifschneider > I've just put the RPM that I'm using on my 5.2 box on > "ftp.tummy.com:/pub/tummy/RPMS". Note that as far as I recall, this > was simply a re-compile (using "rpm --rebuild") of the source > RPM for 6.(1? 0?). Thanks. I'll grab that if I can't get it to build properly today. I really dislike using RPMs. Just old fashioned. :-) But it'll be easier to use that than find another tool. -- Michael J. Hammel | Outside of the killings, Washington has one of The Graphics Muse | the lowest crime rates in the country." mjhammel at graphics-muse.org | Mayor Marion Barry, Washington, D.C. http://www.graphics-muse.com From cross at lithos.colorado.edu Thu Oct 28 13:14:22 1999 From: cross at lithos.colorado.edu (Jason Cross) Date: Thu, 28 Oct 1999 13:14:22 -0600 (MDT) Subject: [lug] off subject: Sun problem. Message-ID: <199910281914.NAA17915@lithos.colorado.edu> Hi guys, This is not exactly a Linux problem, but I need some help, and you guys are the only ones I have left. I do seismology experiments, in which I bring back large numbers of SCSI disks from field instruments, and read them off onto an ultra 10. I'm having a heck of a problem getting the disks to read. It seems to be some intermittent hardware problem (maybe with the SCSI or PCI buses?). Anyway, the people who are supposed to be my tech support don't know whats going on (their words, not mine). They are going to try to ship me a Sparc 5, but I'd rather keep the ultra 10 if I can get it to work. If worse comes to wors, I can use a Linux laptop with a scsi card and dump it to the ultra via NFS, but that's a long way around (the Linux laptop doesn't have any problems!!) Anyway, I can sometimes use the disks (usually just one), and then I can't use any others, and an error message comes up on the console window complaining about "reserved phase" errors. When it works, I can do a probe-scsi-all on the setup screen, and it will finhe disk, however, when it doesn't work I'll get the following error: > probe-scsi-all /pci at 1f,0/pci at 1/scsi at 1,1 Script interrupt: Reserved phase > I switch disks pretty much at will, but I was given to understand that this was O.K., as we aren't mounting them or anything, a program more or less just does a dd, and dumps the data off of them. Then I plug in the next one and do the same. The tech support people gave me a new power supply for the disks, saying that the problem was probably that the external power supply for my disks needed to be grounded to the computer case. That didn't fix it. Anybody have any clues? I'd be very greatful for anything to try!! -Jason Crosswhite jason at newberry.uoregon.edu . From wallen at boulder.nist.gov Thu Oct 28 13:25:42 1999 From: wallen at boulder.nist.gov (Wayde Allen) Date: Thu, 28 Oct 1999 13:25:42 -0600 (MDT) Subject: [lug] off subject: Sun problem. In-Reply-To: <199910281914.NAA17915@lithos.colorado.edu> Message-ID: On Thu, 28 Oct 1999, Jason Cross wrote: > The tech support people gave me a new power supply for the disks, saying that > the problem was probably that the external power supply for my disks needed > to be grounded to the computer case. That didn't fix it. > > Anybody have any clues? I'd be very greatful for anything to try!! The first question that comes to my mind is whether or not the SCSI bus is properly terminated? Depending on your drive, or the jumper settings on the drive, it will either be internally terminated or externally terminated. Consequently it makes a difference how you connect the device to your SCSI controller. Could this be it? - Wayde (wallen at boulder.nist.gov) From gilmorea at verglas.Colorado.EDU Thu Oct 28 13:29:49 1999 From: gilmorea at verglas.Colorado.EDU (Andrew Gilmore) Date: Thu, 28 Oct 1999 13:29:49 -0600 Subject: [lug] off subject: Sun problem. In-Reply-To: <199910281914.NAA17915@lithos.colorado.edu>; from Jason Cross on Thu, Oct 28, 1999 at 01:14:22PM -0600 References: <199910281914.NAA17915@lithos.colorado.edu> Message-ID: <19991028132949.D8023@verglas.Colorado.EDU> Hey Jason, you working with Anne? At any rate, did you say that the laptop has no problems reading the drives? Hardware and cabling is of course, the other main issue to consider. What kind of SCSI card does the ultra 10 have in it? I know the Ultra 5's (one under my desk) don't have SCSI unless it's added on? What cables are you using to hook these up? Andrew On Thu, Oct 28, 1999 at 01:14:22PM -0600, Jason Cross wrote: > Hi guys, > > This is not exactly a Linux problem, but I need some help, and you > guys are the only ones I have left. > > I do seismology experiments, in which I bring back large numbers of > SCSI disks from field instruments, and read them off onto an ultra 10. > > I'm having a heck of a problem getting the disks to read. It seems to be > some intermittent hardware problem (maybe with the SCSI or PCI buses?). > > Anyway, the people who are supposed to be my tech support don't know > whats going on (their words, not mine). They are going to try to ship me > a Sparc 5, but I'd rather keep the ultra 10 if I can get it to work. > > If worse comes to wors, I can use a Linux laptop with a scsi card and dump it > to the ultra via NFS, but that's a long way around (the Linux laptop doesn't > have any problems!!) > > Anyway, I can sometimes use the disks (usually just one), and then I can't use any others, and an error message comes up on the console window complaining about > "reserved phase" errors. > > When it works, I can do a probe-scsi-all on the setup screen, and it will finhe disk, however, when it doesn't work I'll get the following error: > > > probe-scsi-all > /pci at 1f,0/pci at 1/scsi at 1,1 > Script interrupt: Reserved phase > > > > I switch disks pretty much at will, but I was given to understand that this > was O.K., as we aren't mounting them or anything, a program more or less just > does a dd, and dumps the data off of them. Then I plug in the next one and > do the same. > > The tech support people gave me a new power supply for the disks, saying that > the problem was probably that the external power supply for my disks needed > to be grounded to the computer case. That didn't fix it. > > Anybody have any clues? I'd be very greatful for anything to try!! > > -Jason Crosswhite > jason at newberry.uoregon.edu > . > > _______________________________________________ > Web Page: http://lug.boulder.co.us > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug From acesouth at frii.com Thu Oct 28 13:40:50 1999 From: acesouth at frii.com (David Willcox) Date: Thu, 28 Oct 1999 13:40:50 -0600 Subject: [lug] Losing Clock setting Message-ID: <007801bf217c$6b9a5140$189e11d8@davewill> Does anyone know why I keep losing my date & time settings in Linux 5.0 and 5.2 versions. I set it using this format: # date 102701351999 But on rebooting it defaults to something way off Thanks ahead for any help, Dave Willcox :-) -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Daivd Willcox.vcf Type: text/x-vcard Size: 562 bytes Desc: not available URL: From shawd at hex.net Thu Oct 28 14:10:50 1999 From: shawd at hex.net (Doug Shaw) Date: Thu, 28 Oct 1999 14:10:50 -0600 Subject: [lug] Losing Clock setting In-Reply-To: <007801bf217c$6b9a5140$189e11d8@davewill> Message-ID: <3.0.6.32.19991028141050.00870450@pop3.hex.net> >Does anyone know why I keep losing my date & time settings in >Linux 5.0 and 5.2 versions. >I set it using this format: # date 102701351999 >But on rebooting it defaults to something way off If you have compiled your own kernel, there is a configuration option for power management under 'General Setup' that indicates that the system clock stores the time in Greenwich Mean Time, not in your local time zone setting. Doug From terrybn at ustman.com Thu Oct 28 14:00:10 1999 From: terrybn at ustman.com (Terry Branaman) Date: Thu, 28 Oct 1999 14:00:10 -0600 Subject: [lug] Losing Clock setting In-Reply-To: <007801bf217c$6b9a5140$189e11d8@davewill> Message-ID: <000201bf217f$0bf2fbe0$394eaecf@bigmama> If the computer is not fairly new, make sure that the system's battery is OK. A low battery can cause the clock to be off, as well as loss of CMOS settings. -----Original Message----- From: lug-admin at lug.boulder.co.us [mailto:lug-admin at lug.boulder.co.us]On Behalf Of David Willcox Sent: Thursday, October 28, 1999 01:41 PM To: LINUX User group Subject: [lug] Losing Clock setting Does anyone know why I keep losing my date & time settings in Linux 5.0 and 5.2 versions. I set it using this format: # date 102701351999 But on rebooting it defaults to something way off Thanks ahead for any help, Dave Willcox :-) -------------- next part -------------- An HTML attachment was scrubbed... URL: From rodmur at maybe.org Thu Oct 28 14:03:53 1999 From: rodmur at maybe.org (Dale Harris) Date: Thu, 28 Oct 1999 14:03:53 -0600 Subject: [lug] off subject: Sun problem. In-Reply-To: <199910281914.NAA17915@lithos.colorado.edu>; from cross@lithos.colorado.edu on Thu, Oct 28, 1999 at 01:14:22PM -0600 References: <199910281914.NAA17915@lithos.colorado.edu> Message-ID: <19991028140353.A29231@maybe.org> On Thu, Oct 28, 1999 at 01:14:22PM -0600, Jason Cross elucidated: > Hi guys, > > This is not exactly a Linux problem, but I need some help, and you > guys are the only ones I have left. > > I do seismology experiments, in which I bring back large numbers of > SCSI disks from field instruments, and read them off onto an ultra 10. > > I'm having a heck of a problem getting the disks to read. It seems to be > some intermittent hardware problem (maybe with the SCSI or PCI buses?). > Do you see any kind of SCSI errors in the logs? Like sense key, or ASC/ASCQ errors. -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Dale Harris GPG key: 372FBD57 http://www.maybe.org/ M.A.Y.B.E. = Maybe is an Altruistic Yet Bohemian Enigma From caldodge at fpcc.net Thu Oct 28 14:04:30 1999 From: caldodge at fpcc.net (Calvin Dodge) Date: Thu, 28 Oct 1999 14:04:30 -0600 Subject: [lug] Losing Clock setting References: <007801bf217c$6b9a5140$189e11d8@davewill> Message-ID: <3818AC4E.7CD66548@fpcc.net> > David Willcox wrote: > > Does anyone know why I keep losing my date & time settings in Linux > 5.0 and > 5.2 versions. > > I set it using this format: # date 102701351999 > > But on rebooting it defaults to something way off Unlike DOS and Windows, Linux system clock settings are not automatically plugged into the BIOS clock. You have to use a command to tell the computer to do just that - "/sbin/hwclock --systohc" should do the trick. Calvin From rodmur at maybe.org Thu Oct 28 14:09:36 1999 From: rodmur at maybe.org (Dale Harris) Date: Thu, 28 Oct 1999 14:09:36 -0600 Subject: [lug] Losing Clock setting In-Reply-To: <3.0.6.32.19991028141050.00870450@pop3.hex.net>; from shawd@hex.net on Thu, Oct 28, 1999 at 02:10:50PM -0600 References: <007801bf217c$6b9a5140$189e11d8@davewill> <3.0.6.32.19991028141050.00870450@pop3.hex.net> Message-ID: <19991028140936.B29231@maybe.org> On Thu, Oct 28, 1999 at 02:10:50PM -0600, Doug Shaw elucidated: > >Does anyone know why I keep losing my date & time settings in > >Linux 5.0 and 5.2 versions. > > >I set it using this format: # date 102701351999 > > >But on rebooting it defaults to something way off > > If you have compiled your own kernel, there is a configuration option for > power management under 'General Setup' that indicates that the system clock > stores the time in Greenwich Mean Time, not in your local time zone setting. > > Doug Also look in your /etc/init.d (hmm, not sure if that is /etc/rc.d/init.d or /sbin/init.d on Redhat) scripts for a clock or hwclock command. It will be setting your system clock from your CMOS, most likely. And your CMOS is probably running on UTC, or it could just be way off, or both. -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Dale Harris GPG key: 372FBD57 http://www.maybe.org/ M.A.Y.B.E. = Maybe is an Altruistic Yet Bohemian Enigma From cross at lithos.colorado.edu Thu Oct 28 14:41:23 1999 From: cross at lithos.colorado.edu (Jason Cross) Date: Thu, 28 Oct 1999 14:41:23 -0600 (MDT) Subject: [lug] off subject: Sun problem. Message-ID: <199910282041.OAA18495@lithos.colorado.edu> > The first question that comes to my mind is whether or not the SCSI bus is > properly terminated? I believe they are terminated internally. They work fine out in the field when I plug them into the digitizer (no terminator required), and they work fine on the Linux laptop without me ever adding termination. Thanks for the thought though. -Jason From carl.wagner at level3.com Thu Oct 28 16:32:24 1999 From: carl.wagner at level3.com (Carl Wagner) Date: Thu, 28 Oct 1999 16:32:24 -0600 Subject: [lug] off subject: Sun problem. References: <199910282041.OAA18495@lithos.colorado.edu> Message-ID: <3818CEF8.1966959B@level3.com> Jason, Is there any way that you have double terminated one end? Is it possible that you have a bad SCSI cable or terminator in your Sun? Carl. Jason Cross wrote: > > The first question that comes to my mind is whether or not the SCSI bus is > > properly terminated? > > I believe they are terminated internally. They work fine out in the field when > I plug them into the digitizer (no terminator required), and they work fine on the Linux laptop without me ever adding termination. > > Thanks for the thought though. > > -Jason > > _______________________________________________ > Web Page: http://lug.boulder.co.us > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug From cross at lithos.colorado.edu Thu Oct 28 17:37:13 1999 From: cross at lithos.colorado.edu (Jason Cross) Date: Thu, 28 Oct 1999 17:37:13 -0600 (MDT) Subject: [lug] Sun problem resolved. Message-ID: <199910282337.RAA20258@lithos.colorado.edu> > > Jason, > > Is there any way that you have double terminated one end? Is it possible that you have > a bad SCSI cable or terminator in your Sun? of the cable in epoxy, and the cable degraded over time due to being plugged in and unplugged so many times. Thanks for everyone's help. I'm replacing the cable, and all seems well now. -Jason Crosswhite From socket at peakpeak.com Thu Oct 28 20:00:20 1999 From: socket at peakpeak.com (socket at peakpeak.com) Date: 28 Oct 1999 20:00:20 -0600 Subject: [lug] A Big Red Switch... Message-ID: <86zox3ncy3.fsf@peakpeak.com> Perhaps someone here could help me with a minor issue, something in the realm of hardware and switches and device drivers and such. A 486 in the corner houses a 56k modem, masqueraded to our other computers over 10Mbit ethernet. The problem is about deciding how to connect or disconnect to my ISP without needing to inform the (possibly sleeping, possibly surfing) roommate of the change. On the one hand, we could continue doing what we're doing (telnet into the router and issue 'pon' or 'poff' commands, a debian thing) or put some work into automoting the task somewhat (by doing some sort of diald or new-pppd thing that I haven't really been able to get working)... ...or we could do the *really* geeky thing and make a Big Red Switch that we can put in a lego case and mount it to the wall. The switch would be connected directly to the router, probably through the serial port, and I would make a device driver to watch activity on the switch and bring up or down the connection appropriately. This is what we've decided to do. (despite the obvious complexity of the plan) Here's the problem: I've taken the Computers as Components class at CU, which taught me a good deal about serial I/O, polling, interrupts, handlers, etc. but completely failed to do it in a context of useful computing. I could solve this task under DOS with Borland 3, but I don't even know where to begin to learn about hardware interaction in the Linux kernel source to make a device driver, or (assuming the existing serial kernel code is sufficient, which it likely is) how I'd go about writing a program that calls on available resources to do this. Basically, what I'm asking is this: Can somebody refer me to documentation for people familiar with C and the *concepts* of hardware I/O handling, so I could learn about creating a Linux handler (module, daemon, whatever) for a Big Red Switch? Also, any hints on the proper mechanism of the switch or associated software would be very welcome. We're really in the design stage at the moment. Thanks muchly! -- Chris Riddoch socket at peakpeak.com Will provide pseudo-insightful commentary for food http://www.peakpeak.com/~socket From jafo at tummy.com Thu Oct 28 21:07:36 1999 From: jafo at tummy.com (Sean Reifschneider) Date: Thu, 28 Oct 1999 21:07:36 -0600 Subject: [lug] A Big Red Switch... In-Reply-To: <86zox3ncy3.fsf@peakpeak.com>; from socket@peakpeak.com on Thu, Oct 28, 1999 at 08:00:20PM -0600 References: <86zox3ncy3.fsf@peakpeak.com> Message-ID: <19991028210736.S883@tummy.com> On Thu, Oct 28, 1999 at 08:00:20PM -0600, socket at peakpeak.com wrote: >documentation for people familiar with C and the *concepts* of >hardware I/O handling, so I could learn about creating a Linux >handler (module, daemon, whatever) for a Big Red Switch? Connect both BRSs to DTR and DCD. Then you want a program which does something like: #include #include #include #include #include int fd = open("/dev/ttyS0", O_RDONLY), ret; int state; // 0 if line is down, 1 if up state = 0; while (1) { ioctl(fd, TIOCMGET, &ret); if ((ret & TIOCM_CAR) && !state) { system("pon"); state = 0; } if (!(ret & TIOCM_CAR) && state) { system("poff"); state = 1; } sleep(1); } Sean -- It is far from clear whether "good intentions plus stupidity" or "evil intentions plus intelligence" have wrought more harm in the world. --- Domer, The Logic of Failure Sean Reifschneider, Inimitably Superfluous URL: HP-UX/Linux/FreeBSD/BSDOS scanning software. From jason at newberry.uoregon.edu Thu Oct 28 22:45:24 1999 From: jason at newberry.uoregon.edu (Jason A. Crosswhite) Date: Thu, 28 Oct 1999 21:45:24 -0700 Subject: [lug] AARGGG! ADSL and NAT. References: <199910281537.JAA13181@feynman.graphics-muse.org> Message-ID: <38192664.362375BA@newberry.uoregon.edu> Hello all, Sorry to bother you all again. Seems to be a busy day for me and computer problems. Here's the story: I've got an external ADSL modem (Cisco 675) connected to a hub in my house, to which me and my room mate's computers are connected. We use NAT and DHCP services over a PPP to USWEST. Everything is working hunky dorey, except for one thing. I can't figure out how to set the DISPLAY environment variable on a unix machine I'm working on in campus so that graphics displays on my home machine. I guess I just don't understand NAT well enough. I though that each IP in my SOHO network was given as a port for my one real global IP which my 675 owns (well on port for each protocol). I logged onto my 675 and did a "show nat", got the IP and some port numbers, but no luck. Can't even telnet in. I have to be doing something brain dead, but what? I'm sure somebody must know the answer here. If its something obvious, and I missed a HOWTO somewhere (I looked!), then feel free to beat me over the head. I just want to work from home!!! Thank in advance. -- -Jason Crosswhite jason at newberry.uoregon.edu http://darkwing.uoregon.edu/~jacrossw Rm 220, Cascade Building University of Oregon From jafo at tummy.com Thu Oct 28 22:49:16 1999 From: jafo at tummy.com (Sean Reifschneider) Date: Thu, 28 Oct 1999 22:49:16 -0600 Subject: [lug] AARGGG! ADSL and NAT. In-Reply-To: <38192664.362375BA@newberry.uoregon.edu>; from Jason A. Crosswhite on Thu, Oct 28, 1999 at 09:45:24PM -0700 References: <199910281537.JAA13181@feynman.graphics-muse.org> <38192664.362375BA@newberry.uoregon.edu> Message-ID: <19991028224916.W883@tummy.com> On Thu, Oct 28, 1999 at 09:45:24PM -0700, Jason A. Crosswhite wrote: >I've got an external ADSL modem (Cisco 675) connected to a hub in my I presume you mean "RADSL", since that's the variety of DSL offered around here. :-) >house, to which me and my room mate's computers are connected. We use >NAT and DHCP services over a PPP to USWEST. Everything is working hunky >dorey, except for one thing. I can't figure out how to set the DISPLAY >environment variable on a unix machine I'm working on in campus so that >graphics displays on my home machine. If you're using NAT, you don't. You could set up something on your NAT machine so that the incoming X connections are forwarded to your internal machine, but your best bet would probably be to use SSH to connect to CSU (check, but I think licensing would allow it for SSH version 1.2 and older), and have it proxy the X connection for you. If the SSH licensing bites you, you could use the dxpc software to set up a similar tunnel (and you get REALLY good compression over it too). This is probably the preferred way to do it, except that it doesn't offer encryption. It DOES offer GREAT X compression though. Sean -- "I was just thinking of the immortal words of Socrates, when he said ``I drank what?''" -- Chris Knight, _Real_Genius_ Sean Reifschneider, Inimitably Superfluous URL: HP-UX/Linux/FreeBSD/BSDOS scanning software. From jason at newberry.uoregon.edu Fri Oct 29 00:46:55 1999 From: jason at newberry.uoregon.edu (Jason A. Crosswhite) Date: Thu, 28 Oct 1999 23:46:55 -0700 Subject: [lug] AARGGG! ADSL and NAT. References: <199910281537.JAA13181@feynman.graphics-muse.org> <38192664.362375BA@newberry.uoregon.edu> <19991028224916.W883@tummy.com> Message-ID: <381942DF.316597CB@newberry.uoregon.edu> > I presume you mean "RADSL", since that's the variety of DSL offered > around here. :-) I'll try to be more careful next time, but no promises!! :-) > If the SSH licensing bites you, you could use the dxpc software to set up > a similar tunnel (and you get REALLY good compression over it too). This > is probably the preferred way to do it, except that it doesn't offer > encryption. It DOES offer GREAT X compression though. Thanks! That turned the trick. Didn't even know this existed. Much faster this way too. Sure wish it had better image compression though! Text works really nicely though. -- -Jason Crosswhite jason at newberry.uoregon.edu http://darkwing.uoregon.edu/~jacrossw Rm 220, Cascade Building University of Oregon From wallen at boulder.nist.gov Fri Oct 29 08:19:14 1999 From: wallen at boulder.nist.gov (Wayde Allen) Date: Fri, 29 Oct 1999 08:19:14 -0600 (MDT) Subject: [lug] take my job, please... (fwd) Message-ID: ---------- Forwarded message ---------- Date: Thu, 28 Oct 1999 21:34:07 -0600 From: Charles Morrison To: meta at lug.boulder.co.us Reply-To: meta at lug.boulder.co.us Subject: [BLUGMETA] take my job, please... Greetings, I recently took a new job and committed myself to try to help find a replacement for the old job. As no one has applied for this position yet, I volunteered to post it to a few lists where some very linux (and that other OS) literate people would see it. I'd appreciate if you would post this to your LUG lists. The job is with the Weld County Health Department and listed as a health data specialist or some such thing. I'll cut the beaurocratic labels and explain it. The job is composed of the following duties. 1) Maintain an existing Linux / Apache / MySQL web-database server. This server supplies health related data, via dynamic web searches, to the public, local and state health departments. It utilizes CGI's built on compiled C code. This server has proven to be bullet-proof and is VERY trouble free. You can look at the public part of it at http://cohid.gen.co.us 2) Migrate the above to a NT4 / IIS / ASP / SQL Server 7 environment located at the Colorado Dept. of Public Health & Environment (CDPHE). This involves both the database end (largely finished) and the web server / page end (largely undone). It involves working closely with IT personnel from CDPHE - Yes they are very nice people :-) , and working to programming standards that they determine. This is really more of a partnership than I make it sound. They are exploring this territory too, and this project is part of their exploration. 3) The structure of the job is unique. You would be responsible to yourself and the committee that governs the project. That committee is made up of data users and suppliers from county health departments and CDPHE. While the job is administered (salary, benefits, etc ) through Weld County, It is really the committee that has authority over the project. If you have the equipment, a great deal of work could be done at home (I did about 50/50). All MS software is supplied by the CDPHE. There is a private office at WCHD in Greeley, complete with Windows PCs and Linux PCs. 4) Additional duties may include light maintenance of a couple MS Access timekeeping systems (a few minutes a month, if that) and first line support of office computer glitches. Weld County Info Services has locked down all the win95 pcs pretty tight, so there is very little you can do for most things except call IS. This is a high profile position in it's own way and could be a great stepping stone for the right person. The current funding (Government job, remember???) is through 2000. A good number of people see this project as unique and valuable, so it's possible that funding may continue beyond 2000. That may depend largely on you. Salary for 1999 was $36,000/yr. The budget for 2000 appears to be for up to $10,000 more for the right person. That may be negotiable. The position closes Wed. Nov. 2, 1999. For more info you can email Gaye Akin at gakin at co.weld.co.us or myself at cmorrison at info2000.net -- Chuck Morrison Sr. Systems Engineer VA Linux Inc. _______________________________________________ Meta mailing list Meta at lug.boulder.co.us http://lists.lug.boulder.co.us/mailman/listinfo/meta From byelen at atlas.csd.net Fri Oct 29 09:52:31 1999 From: byelen at atlas.csd.net (Bruce Yelen) Date: Fri, 29 Oct 1999 09:52:31 -0600 (MDT) Subject: [lug] AARGGG! ADSL and NAT. In-Reply-To: <38192664.362375BA@newberry.uoregon.edu> Message-ID: Hi Jason, You need to configure the NAT (Network Address Translation) on your router to directect any incoming packets for the port you're using for your X display. If you're setting your DISPLAY parm to XXX.XXX.XXX.XXX:0, the port would be 6000. It's normally the display ID (ie :n) plus 600. (..xxx:1 would be 6001). Regards, Bruce On Thu, 28 Oct 1999, Jason A. Crosswhite wrote: > Hello all, > > Sorry to bother you all again. Seems to be a busy day for me and > computer problems. > > Here's the story: > > I've got an external ADSL modem (Cisco 675) connected to a hub in my > house, to which me and my room mate's computers are connected. We use > NAT and DHCP services over a PPP to USWEST. Everything is working hunky > dorey, except for one thing. I can't figure out how to set the DISPLAY > environment variable on a unix machine I'm working on in campus so that > graphics displays on my home machine. > > I guess I just don't understand NAT well enough. I though that each IP > in my SOHO network was given as a port for my one real global IP which > my 675 owns (well on port for each protocol). I logged onto my 675 and > did a "show nat", got the IP and some port numbers, but no luck. Can't > even telnet in. > > I have to be doing something brain dead, but what? I'm sure somebody > must know the answer here. If its something obvious, and I missed a > HOWTO somewhere (I looked!), then feel free to beat me over the head. I > just want to work from home!!! > > Thank in advance. > > -- > -Jason Crosswhite > > jason at newberry.uoregon.edu > http://darkwing.uoregon.edu/~jacrossw > Rm 220, Cascade Building > University of Oregon > > _______________________________________________ > Web Page: http://lug.boulder.co.us > Mailing List: http://lists.lug.boulder.co.us/mailman/listinfo/lug > From llornkcor at llornkcor.com Fri Oct 29 11:14:24 1999 From: llornkcor at llornkcor.com (llornkcor at llornkcor.com) Date: Fri, 29 Oct 99 11:14:24 Mountain Daylight Time Subject: [lug] halloween linux humor Message-ID: <199910291714.MAA03379@charon.host4u.net> This appeared on Linux Today- got a kick out of it. http://criswell.bizland.com/fall99/10-30.html worth a chuckle LP From jafo at tummy.com Fri Oct 29 11:18:04 1999 From: jafo at tummy.com (Sean Reifschneider) Date: Fri, 29 Oct 1999 11:18:04 -0600 Subject: [lug] AARGGG! ADSL and NAT. In-Reply-To: ; from Bruce Yelen on Fri, Oct 29, 1999 at 09:52:31AM -0600 References: <38192664.362375BA@newberry.uoregon.edu> Message-ID: <19991029111804.E883@tummy.com> On Fri, Oct 29, 1999 at 09:52:31AM -0600, Bruce Yelen wrote: >to directect any incoming packets for the port you're using for your X >display. If you're setting your DISPLAY parm to XXX.XXX.XXX.XXX:0, the >port would be 6000. It's normally the display ID (ie :n) plus 600. >(..xxx:1 would be 6001). Note that would be a string addition, not integer. :-) Sean -- "All I'm saying is that when I'm around you I find myself showing off, which is the idiots version of being interesting." -- _LA_Story_ Sean Reifschneider, Inimitably Superfluous URL: HP-UX/Linux/FreeBSD/BSDOS scanning software. From neal at bighorn.dr.lucent.com Fri Oct 29 11:33:54 1999 From: neal at bighorn.dr.lucent.com (Neal McBurnett) Date: Fri, 29 Oct 1999 11:33:54 -0600 Subject: [lug] ftp-only user ids without sendmail, pop, global read, etc. Message-ID: <9910291133.ZM5625@feynman.dr.lucent.com> I want to allow a bunch of minimally-privileged users to update stuff on a web site with ftp. Each user should not have read or write permission outside their own directory or else they could read files elsewhere on the server that are protected via http .htaccess files. We don't want to affect other services on the machine or permission issues. E.g. these users should not be able to login, receive email, retrieve email, etc. FTP-only users would still need a way to change their passwords. I know from the httpd side (apache) how to prevent them from doing cgis and server-side-includes (anything else I've forgotten for the moment?) I found some info on configuring wu-ftp (we are using FTP Version wu-2.6.0(1) Wed Oct 20 10:51:40 MDT 1999) at the WU FTP FAQ. E.g. information on the "no-telnet" desire is at http://www.cetis.hvu.nl/~koos/wu-ftpd-faq.html#QA59 But that mentions an issue with people being able to receive email. I don't know if they would be able to read it or not via pop. If so that seems like a bigger problem. If not it would still seem a good idea to turn off mail delivery for these users. The fact that /etc/ftponly is in /etc/shells would also seem to mean that this user might even be able to execute commands thru email via the right config files in their home directory. Is that (or can it be) turned off in the sendmail config file? I've heard of one way to set up ftp to restrict access to only one directory, but it involves chroot and copies of the ftp bin area. Is there a way without all those silly copies (which would then show up on the web....)? For some visions of other new and future ways for distributed maintenance of web sites (WebDAV, CVS, DELTA-V) see this recent article: http://www.webtechniques.com/archives/1999/10/whitehead/ But I think that although some of those things are available now, they don't really address our need to provide simple, standard safe, well-tested tools for novice users.... Cheers, Neal McBurnett 303-538-4852 Denver Bell Labs / Lucent Technologies http://bcn.boulder.co.us/~neal/ (with PGP key) From wallen at boulder.nist.gov Fri Oct 29 11:42:50 1999 From: wallen at boulder.nist.gov (Wayde Allen) Date: Fri, 29 Oct 1999 11:42:50 -0600 (MDT) Subject: [lug] Forwarded mail.... Message-ID: ---------- Forwarded message ---------- Date: Fri, 29 Oct 1999 11:38:56 -0400 From: Bill Holliker We are in the process on starting up a 4 week Web Administrator School in Houston, Tx, San Diego, CA, Denver, CO, and St. Louis, MO. We are in need of Linux, Web, JavaScript, Perl, and Java instructors. More information about the school can be found at http://www.webmasterschool.com. If any of your members might be interested in teaching, please have them contact me at billh at trainix.com .. thanks From acesouth at frii.com Fri Oct 29 11:54:36 1999 From: acesouth at frii.com (David Willcox) Date: Fri, 29 Oct 1999 11:54:36 -0600 Subject: [lug] Digiboard drivers and Linux Message-ID: <003101bf2236$bc0229e0$109e11d8@davewill> Has anyone had any experience about Digiboard PC/xe or any knowledge about how to access these intelligent multi-serial boards with Linux. I'm told that the drivers for these boards are located in the kernel and have my kernel settings correct. I could really use some help regarding even a hint of how to set Linux to see these boards. I also know that Linux looks at address 200 but let's say from my older OS using Xenix I had these boards setting selectable from an install disk 5 1/4" .Ancient I know but it worked fine. I even callled Digiboard and they told me to set rc.serial so first port is set at 200 for these 8-port boards. I had no luck after doing all these changes and I have about 10 Pawnshops depending on me to get their boards working with Linux real soon like yesterday. There must be someplace besides /usr/src/linux/include/config or someplace in that area to configure these. How about linuxconf? thanks ahead for any help concerning this, dave willcox -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Daivd Willcox.vcf Type: text/x-vcard Size: 562 bytes Desc: not available URL: From cross at lithos.colorado.edu Fri Oct 29 12:25:28 1999 From: cross at lithos.colorado.edu (Jason Cross) Date: Fri, 29 Oct 1999 12:25:28 -0600 (MDT) Subject: [lug] AARGGG! ADSL and NAT. Message-ID: <199910291825.MAA11291@lithos.colorado.edu> Bruce Yelen wrote: > > Hi Jason, > > You need to configure the NAT (Network Address Translation) on your router > to directect any incoming packets for the port you're using for your X > display. If you're setting your DISPLAY parm to XXX.XXX.XXX.XXX:0, the > port would be 6000. It's normally the display ID (ie :n) plus 600. > (..xxx:1 would be 6001). Thanks. I think the cisco router allows me to do this easily. I'll do it when I get home. f all this operating system bigotry!! -Jason From jafo at tummy.com Fri Oct 29 12:35:27 1999 From: jafo at tummy.com (Sean Reifschneider) Date: Fri, 29 Oct 1999 12:35:27 -0600 Subject: [lug] AARGGG! ADSL and NAT. In-Reply-To: <199910291825.MAA11291@lithos.colorado.edu>; from Jason Cross on Fri, Oct 29, 1999 at 12:25:28PM -0600 References: <199910291825.MAA11291@lithos.colorado.edu> Message-ID: <19991029123527.H883@tummy.com> On Fri, Oct 29, 1999 at 12:25:28PM -0600, Jason Cross wrote: >Thanks. I think the cisco router allows me to do this easily. I'll do >it when I get home. By "router" what was meant was the machine you're running NAT on -- presumably your Linux machine acting as a gateway... Sean -- A computer lets you make more mistakes faster than any invention in human history -- with the possible exceptions of handguns and tequila. -- Mitch Ratcliffe Sean Reifschneider, Inimitably Superfluous URL: HP-UX/Linux/FreeBSD/BSDOS scanning software. From bdoctor at nda.com Sun Oct 31 21:26:19 1999 From: bdoctor at nda.com (bdoctor at nda.com) Date: Sun, 31 Oct 1999 21:26:19 -0700 Subject: [lug] Digiboard drivers and Linux References: <003101bf2236$bc0229e0$109e11d8@davewill> Message-ID: <381D166B.8140CE9D@nda.com> I have experience with some older Digi hardware, which worked well, after the driver was sorted, as well as newer hardware, the PC XEM unit, which has an external module for it's serial ports, rather than an octopus cable. Works extremely well! ftp.digi.com is your best resource. All devices are /dev/ttyD. The included RC script should give a good indication as to what goes where. Of course, you'll need the binaries, from the FTP site. This script is designed for a cloned machine, one which may or may not have the Digi device in it, as such it will configure a new machine. Assumes that: The device is working properly (We have PCI devices, very easy) The kernel supports it The modules/kernel are all installed and synced I've not come across any IRQ conflicts, or I/O conflicts, etc. The PCI model is wonderfully easy, compared to the older ISA models, which were very difficult to setup. If you need additional assistance, let me know, I'll do what I can :) I don't do code pageants. #!/bin/sh # Start/stop script by Brad Doctor # This should do the trick for a properly configured digi card. . /etc/rc.d/init.d/functions MODULE_LOCATION=/lib/modules/`uname -r`/misc/epca.o MODULE_INIT_BIN=/usr/local/bin/digiDload case "$1" in start) echo -n "Bringing up Digiboard XEM Driver... " if grep epca /proc/modules > /dev/null ; then echo "The Digi epca driver is already loaded!" exit 0 fi if [ -f $MODULE_LOCATION ] ; then /sbin/insmod $MODULE_LOCATION else echo "Could not find $MODULE_LOCATION, abort" exit 0 fi if [ -f $MODULE_INIT_BIN ] ; then echo "Initializing the driver: " $MODULE_INIT_BIN fi if [ $? != 0 ] ; then echo "Error intializing the card!" else /bin/dmesg | tail -1 fi ;; stop) echo "The machine must be rebooted to remove the module." ;; status) echo "Checking /proc/modules: " grep epca /proc/modules echo "" echo "Checking /proc/epca: " if [ -f /proc/epca ] ; then cat /proc/epca else echo "/proc/epca does not exist, this is wrong." fi ;; *) echo "Usage: $0 [start|stop|status]" ;; esac -brad > David Willcox wrote: > > Has anyone had any experience about Digiboard PC/xe or any > knowledge about how to access these intelligent > multi-serial boards with Linux. I'm told that the drivers > for these boards are located in the kernel and have my > kernel settings correct. I could really use some help > regarding even a hint of how to set Linux to see these > boards. I also know that Linux looks at address 200 but > let's say from my older OS using Xenix I had these boards > setting selectable from an install disk 5 1/4" .Ancient I > know but it worked fine. I even callled Digiboard and they > told me to set rc.serial so first port is set at 200 for > these 8-port boards. I had no luck after doing all these > changes and I have about 10 Pawnshops depending on me to > get their boards working with Linux real soon like > yesterday. There must be someplace besides > /usr/src/linux/include/config or someplace in that area to > configure these. How about > linuxconf? > thanks ahead for any help concerning this, > dave willcox > > -- Brad Doctor Manager of System and Network Administration NDA, An Interliant Company www.nda.com 303-825-4124 - - - - - The profession of book writing makes horse racing seem like a solid, From gmasters at devcg.denver.co.us Sat Oct 30 18:56:41 1999 From: gmasters at devcg.denver.co.us (Gary Masters) Date: Sat, 30 Oct 1999 18:56:41 -0600 (MDT) Subject: [lug] Chronology settings,I use Pine. In-Reply-To: Message-ID: On Wed, 27 Oct 1999, Rob Mohr wrote: > Me too. But I use another tree variety, Pine. Its the only thing I have > used since '94. True. But I got rid of the crank telephone and the > party line last year:-) (just kidding) > eof > I too use Pine. I can't get used to versions above 3.96, however. So I keep the 3.96 source code and just keep compiling it on everything I end up having to read mail on. I won't tell you of my withdrawal symptoms when I had to give up VAX mail about three years ago. ;-) -- Gary Masters gmasters at devcg.denver.co.us From gmasters at devcg.denver.co.us Sat Oct 30 19:58:24 1999 From: gmasters at devcg.denver.co.us (Gary Masters) Date: Sat, 30 Oct 1999 19:58:24 -0600 (MDT) Subject: [lug] ftp-only user ids without sendmail, pop, global read, etc. In-Reply-To: <9910291133.ZM5625@feynman.dr.lucent.com> Message-ID: On Fri, 29 Oct 1999, Neal McBurnett wrote: [SNIP] > I've heard of one way to set up ftp to restrict access to only > one directory, but it involves chroot and copies of the ftp > bin area. Is there a way without all those silly copies (which > would then show up on the web....)? > ProFtpd can be configured to prevent users from changing above their home directories. And it isn't necessary to provide /bin, /lib, /etc, etc., in the user directories. As early as two releases ago, ProFtpd was riddled with security problems, but the current release [1.2.0pre8] claims to be better. All configuration of ProFtpd is done in the file proftpd.conf. The syntax is similar to Apache configuration files. I have had good luck with it, with the exception of frequent updates to keep ahead of the security flaws. You can get more information at http://www.proftpd.org. -- Gary Masters gmasters at devcg.denver.co.us From dajo at privatei.com Sun Oct 31 17:58:33 1999 From: dajo at privatei.com (David Outteridge) Date: Sun, 31 Oct 1999 17:58:33 -0700 Subject: [lug] Oracle uses Linux Message-ID: <199911010058.RAA00817@Blackhand.Anthrax> From matthew.w.mcillece at lmco.com Sun Oct 31 17:57:24 1999 From: matthew.w.mcillece at lmco.com (McIllece, Matthew W) Date: Sun, 31 Oct 1999 17:57:24 -0700 Subject: [lug] VC to X in 2 minutes 42 seconds Message-ID: <98B98E951BA0D1119A590000F8045A47053ECA00@emss02m05.ems.lmco.com> > When I do a simple "xinit" from a virtual console, it consistently takes 2 > minutes and 42 seconds for the X desktop to show up. This seems like an > unreasonable amount of time just to start the X server, and I remember a > time months ago when it only took about 5-10 seconds to start X. I > imagine I did something that caused this, but don't remember what it was. > When I installed KRUD last month I configured for an anticipated network > connection, but my NIC isn't plugged into a network yet. Could this be > the problem? I've looked through numerous HOWTO's, READMEs, and FAQs in > search of a reason and haven't found one. Anyone have a clue? > > Running an AMDK6-2-350 (Oc'd to 400), KRUD Red Hat 6.0 > > X reports: > XFree86 Version 3.3.3.1 > Operating System: Linux 2.0.36 i586 > > "uname -a" reports: > Linux (none) 2.2.5-22 > > CPU states (BEFORE doing "xinit"): 0.0% user, 0.3% system, 0.0% nice, > 99.6% idle > > largest CPU process reported by top BEFORE doing "xinit" is: > top 0.3 %CPU 0.7 %MEM > > largest memory user reported by "ps -aux | sort +5n" BEFORE doing "xinit" > is: > xfs 1.5 %MEM > > CPU states (AFTER doing "xinit"): 0.3% user, 0.1% system, 0.0% nice, 99.4% > idle > > largest CPU process reported by top AFTER "xinit" is: > top 0.5 %CPU 0.7 %MEM > > largest memory user reported by "ps -aux | sort +5n" AFTER "xinit" is: > /usr/X11R6/bin/Xw 1.8 %MEM > > Services started at bootup: > apmd > atd > crond > gnomehack (started, but not running) > gpm > inet > keytable (started, but not running) > linuxconf (started, but not running) > lpd > network (started, but not running) > pcmcia (started, but not running) > portmap > random (started, but not running) > rinet > sound (started, but not running) > sshd > xfs > > /etc/hosts: > 127.0.0.1 localhost localhost.localdomain > ABC.DEF.GHI.JKL > > > (where ABC.DEF.GHI.JKL is an actual IP address inside the network I > anticipated being connected to) From matthew.w.mcillece at lmco.com Sun Oct 31 17:57:24 1999 From: matthew.w.mcillece at lmco.com (McIllece, Matthew W) Date: Sun, 31 Oct 1999 17:57:24 -0700 Subject: [lug] VC to X in 2 minutes 42 seconds Message-ID: <98B98E951BA0D1119A590000F8045A47053ECA00@emss02m05.ems.lmco.com> > When I do a simple "xinit" from a virtual console, it consistently takes 2 > minutes and 42 seconds for the X desktop to show up. This seems like an > unreasonable amount of time just to start the X server, and I remember a > time months ago when it only took about 5-10 seconds to start X. I > imagine I did something that caused this, but don't remember what it was. > When I installed KRUD last month I configured for an anticipated network > connection, but my NIC isn't plugged into a network yet. Could this be > the problem? I've looked through numerous HOWTO's, READMEs, and FAQs in > search of a reason and haven't found one. Anyone have a clue? > > Running an AMDK6-2-350 (Oc'd to 400), KRUD Red Hat 6.0 > > X reports: > XFree86 Version 3.3.3.1 > Operating System: Linux 2.0.36 i586 > > "uname -a" reports: > Linux (none) 2.2.5-22 > > CPU states (BEFORE doing "xinit"): 0.0% user, 0.3% system, 0.0% nice, > 99.6% idle > > largest CPU process reported by top BEFORE doing "xinit" is: > top 0.3 %CPU 0.7 %MEM > > largest memory user reported by "ps -aux | sort +5n" BEFORE doing "xinit" > is: > xfs 1.5 %MEM > > CPU states (AFTER doing "xinit"): 0.3% user, 0.1% system, 0.0% nice, 99.4% > idle > > largest CPU process reported by top AFTER "xinit" is: > top 0.5 %CPU 0.7 %MEM > > largest memory user reported by "ps -aux | sort +5n" AFTER "xinit" is: > /usr/X11R6/bin/Xw 1.8 %MEM > > Services started at bootup: > apmd > atd > crond > gnomehack (started, but not running) > gpm > inet > keytable (started, but not running) > linuxconf (started, but not running) > lpd > network (started, but not running) > pcmcia (started, but not running) > portmap > random (started, but not running) > rinet > sound (started, but not running) > sshd > xfs > > /etc/hosts: > 127.0.0.1 localhost localhost.localdomain > ABC.DEF.GHI.JKL > > > (where ABC.DEF.GHI.JKL is an actual IP address inside the network I > anticipated being connected to) From winrip at diac.com Sun Oct 31 18:09:02 1999 From: winrip at diac.com (winrip) Date: Sun, 31 Oct 1999 20:09:02 -0500 Subject: [lug] Losing Clock setting References: <007801bf217c$6b9a5140$189e11d8@davewill> Message-ID: <381CE82D.316D31D4@diac.com> David Willcox wrote: > Does anyone know why I keep losing my date & time settings in Linux > 5.0 and5.2 versions. I set it using this format: # date > 102701351999 But on rebooting it defaults to something way off Thanks > ahead for any help, Dave Willcox :-) I had the same problem with an ATX Mother board... I had to turn off APM in the bios. Granted the machine doesn't go into power save anymore and doesn't auto power down, but it keeps time. Hope it helps Bill. -------------- next part -------------- An HTML attachment was scrubbed... URL: From jason at newberry.uoregon.edu Fri Oct 29 15:09:30 1999 From: jason at newberry.uoregon.edu (Jason Crosswhite) Date: Fri, 29 Oct 1999 15:09:30 -0600 Subject: [lug] AARGGG! ADSL and NAT. References: <199910291825.MAA11291@lithos.colorado.edu> <19991029123527.H883@tummy.com> Message-ID: <381A0D0A.84E6B56D@newberry.uoregon.edu> > By "router" what was meant was the machine you're running NAT on -- > presumably your Linux machine acting as a gateway... Nope. The CISCO 675 actually runs NAT, and I'm using that. I believe it allows me to reassign the global incoming IP and port to whatever internal IP and port I want. So I was hoping to just have the router change the port to the correct one that X wanted on my Linux box. Am I doing something stupid? I was reading the operating system manual for the 675 last night, and I'm pretty sure it would work. -Jason