OMSA for Red Hat 9?

[Brian A. Seklecki]

> You've never encountered a "legacy machine" before?
>

Generally ISV businesses have a legal and moral obligation to their 
clients to ensure that they're running a vendor-supported system. 
Migration planning is normally contractually obligated.

In the Cathedral and the Bazaar, E.S.R. makes a lot of good references to 
the value of vendor-support and commercial software in his 
software-as-a-service sections.

The value of commercial software drops to $0 when the vendor discontinues 
service.  RH9 was essentially commercial software in that aspect.  ISVs 
inherit the cost-overhead of backporting patches/bug/security-fixes to 
vendor unsupported branches.

This is where CentOS comes in.

Look on the bright side: Cisco and Net-SNMP are all maintaining three 
generations/branches of supported code. >:}

What Redhat is Redhat doing?  It's hard to say.  They're pretty tight 
lipped about it.  I mean, what we're really talking about is a Libc 
version, a toolchain version set, and the kernel.  But in general it's a 
problem with trying to coordinate linux ditro releng with thousands of 
other projects releng.

It's just not very ISV-friendly.  It's one of the best arguments 
anti-Linux pundents have right now.  Maintaining a long-standing 
stable/legacy release like IOS 12.1, FreeBSD 4.x, Net-SNMP 5.1.x, Apache 
1.3.x, HFUX 11, etc is very important.  The RedHat's and PHP's out there 
just don't get it yet.

We're way off-topic at this point, anyway.  Release engineering theory is 
a topic for another forum

~BAS

> I'm not real happy about them.  Heck, there are still a couple of Red