[lug] The future of Linux support

[Cory Dekker]

Hi all,

Rather than extend the existing "MS vs. Linux" discussion, I would like to
launch a related, but distinctly different, discussion about support
philosophies, in general.  In the Linux world, the OS, and most related
tools/utilities, are basically free.  Companies, like Red Hat, that
"package" an installation, can only honestly charge for their added value to
the installation process and for the effort the "bundling" of the selected
tools, utilities, documentation, and "support".  Where do you see the future
of Linux "packaging" companies, like Red Hat, and their support services,
headed?

Should companies like Red Hat be
1) trying to create a minimal-cost product that only covers their packaging,
distribution, and minimal support (only on tried-and-proven system
configurations) services, effectively abandoning "needy" customers?
2) trying to create an after-sale support organization that charges
(probably outrageously) by the hour for product support, over and above
anything associated with the "tried-and-proven" configurations, or for more
extensive problems beyond simply installation?
3) trying to create a more expensive product that provides comprehensive
support, rolling the anticipated/potential support organization costs into
the basic cost of the product?
4) both #1 and #3; and if so, would anyone except maybe Corporate America
(who isn't currently buying Linux, in mass) actually buy #3?
5) trying to create a more comprehensive set of supported
("tried-and-proven") configurations, rolling the cost of such efforts to
address less commonly used configurations, into the basic cost of the
product, thus forcing the general buyer to fund service costs for the more
obscure customers?
6) trying to create a more focused set of supported common configurations,
abandoning less common configuration users to either go-it-alone, or pay
extra to utilize after-sale support?

I know first-hand the futility (and expense) of trying to deal with
Micro$oft support ("yep... that's a known bug... it will be fixed in Windows
3000... upgrade then.")   I also know all of the joys and horrors of RH
support... especially in the not-so-mainstream configuration world.  And, I
understand the limited market that exists for #2 (I was one of the early RH
Certified consultants.... it hasn't really proven to be of that much value,
yet).  However, as Linux continues to move towards the mainstream, I think
that things are going to change.  I'm curious to hear what direction(s)
folks think those changes are going to head towards.  There are really only
a couple basic philosophy questions here...
- Target only the main-stream markets/configurations, ... or ... charge
more, to fund aiming at wider markets/configurations?
- Increase QA costs, time-to-market, and thus product cost, ... or ... trim
and cut corners, save, and get there faster and cheaper?
- Bundle service fees in the basic product cost, ... or ... charge extra
only to/for those who need the service(s)?

I realize that the answers to these questions have plagued Corporate America
for years, but as Open-Source based solutions become more and more
commercially viable, will the answers... or even the questions... be
changed?

                                                -Cory