[lug] Opinions on E*Trade dumping Sun for Linux?

rm at fabula.de rm at fabula.de
Fri Feb 8 09:24:35 MST 2002


On Fri, Feb 08, 2002 at 08:32:48AM -0700, Bryan Field-Elliot wrote:
> As I mentioned, I love Java and have used it successfully on probably
> 15+ projects over the last 3 or so years. But in asking myself why I
> love Java (whether because of the language grammar or because of the
> rich libraries which come with it standard, as well as J2EE), I found a
> roughly 50/50 split. This is what got me interested in the .Net CLI -
> that people's work on libraries and other ready-made components can be
> done independent of language, and used equally effectively by all
> languages which compile to the CLI. At least, that's how the theory
> goes, and the article you referenced on javalobby tries to argue that
> this is largely a myth. Personally, I don't know, as I haven't looked
> deeply into .Net yet.
> 
> As for Microsoft sneaking a patent in there or something to otherwise
> upset future cross-platform development (e.g. Mono) -- I think they
> would have to patent something new, rather than something already
> submitted to the ECMA. In which case, Mono and other OS implementations
> fork away from Microsoft, having already gotten (for free) what they
> mostly wanted in the first place -- the CLI specs.

True, but the CLI specs are only a small part of the "plot". 
And, given the fact that probably the majority of .Net programmers will
be MS followers, where would forking away from Microsoft leave Mono? In
the proprietary niche section. The main, if not only, reason for Java and
.Net _is_ the unified code base you gain (even if it's only on paper. My
scheme programs run on _far_ more platforms then my java code). 

> Again, I'm just playing devil's advocate here. I would never give .Net a
> second glance if it only ran on Windows, but now that projects like Mono
> are well off the ground, it deserves a look in my opinion.

I wouldn't call 'we still compile under windows' well off ground ;-)
My fear is that all this tied important mind resources - and that's some-
thing the free software comunity doesn't have in abundance. If MS fails
in a product that might be o.k. for them. Spending two years on a product
that's doomed because of copyright/patent issues is a major showstopper.

> The topic which triggered this discussion in the first place was Sun
> losing market share, and I maintain that they may in fact lose a lot of
> developer mindshare (over the mid term) if .Net lives up to it's hype
> (which admittedly wouldn't be consistent with Microsoft's track record).
> But what do I know, I'm just a developer interested in new and exciting
> tools.

What's _so_ exiting about it. I saw a comparison with Java and the technology
didn't look that much more advanced. I didn't even understand most of the
Java bytecode hype - Lisp had it, Pascal had it, Scheme has it. If you want
to see network transparent code done right, have a look at Kali Scheme ;-)
I run code on a Intel Linux, mk68 Linux and a Mac PPC, all in the same
(virtual) address/naming space an can even send closures from one box to
the other ...

Another thing i still have problems with: the claim that the VM is better
suited to serve different 'front end' languages. I'd agree that the Java
VM is especially bad to be used by other front ends, but from looking at
the little infos i have so far i don't see that the CLI is so much better.
And then, as was allready mentioned: aren't we fooled by this? Is it really
the Fortran/Haskell/Perl _syntax_ we need or isn't the speed/type safety/
code mess  what we want?

 Ralf 
> Bryan
> 
> 
> 
> 
> On Wed, 2002-02-06 at 22:26, Ed Hill wrote:
> 
>     On Wed, 2002-02-06 at 20:58, Bryan Field-Elliot wrote:
>     > 
>     > I'm curious if anyone in this group has played with .Net, either via
>     > Microsoft or via the Mono project?  I love Java and have used it
>     > exclusively for the last three years, but the .Net CLI architecture is
>     > admittedly compelling. 
>     
>     
>     Huh?  Compelling?  What exactly do you gain?
>     
>     As a Java programmer, you loose the security model and the platform
>     independence.  So no more developing on Linux and running on (almost)
>     anything.  And if you're a C++ coder, with the castrated ".NET managed
>     C++" you loose multiple inheritance, dynamic typing, and generics
>     (b-bye, templates!).  And I doubt that any Mono/Ximian implementation
>     would fix these problems because the CLS/CLR (AFAIK) simply doesn't have
>     the design or expressive power to support them.
>     
>     So before anyone gets carried away with unbridled optimism such as:
>     
>     http://mail.gnome.org/archives/gnome-hackers/2002-February/msg00031.html
>     http://eiffel.com/doc/manuals/technology/bmarticles/sd/dotnet.html
>     
>     they should learn about some of the non-trivial down-sides:
>     
>     http://www.javalobby.org/clr.html
>     
>     
>     And keep in mind the unchanging goal of Microsoft: enslaving you to
>     their platform.  But don't take it personally!  Its just a business
>     model...  ;-)
>     
>     Ed
>     
>     ps - I'm afraid Alan Cox's assessment of Mono:
>     
>            "Be assured that the day they decide you are a nuisance 
>             the VM will acquire a patented neat feature that kills 
>             you off.  Just ask the Samba people."
>     
>          will come true.  For all Miguel knows, many such claims are 
>          already submitted and slithering their way through the US PTO.
>     
>     
>     -- 
>     Edward H. Hill III, PhD
>     Post-Doctoral Researcher   |  Email:       ed at eh3.com, ehill at mines.edu
>     Division of ESE            |  URL:         http://www.eh3.com
>     Colorado School of Mines   |  Phone:       303-273-3483
>     Golden, CO  80401          |  Fax:         303-273-3311
>     Key fingerprint = 5BDE 4DA1 66BE 4F7B BC17  3A0C 932B 7266 1E76 F123
>     
>     



More information about the LUG mailing list