On Mon, 1 Jun 1998, Patrick Lawrence wrote:
>[e. green] 
> > There are two kinds of stability: program stability, and deployment
> > stability.
> 
> Indeed there are.
> 
> > However, it was obvious from the first that RH 5.0 was not going to be the
> > last in the RH 5 series, and that all support for RH 5.0 would be
> > discontinued as other releases in the RH 5 series were brought out.
> 
> Was there an announcement about this?  Redhat still sends out security
> fixes for 4.2 so I don't see why they won't continue to support 5.0 after
> the 5.1 release.

No announcement, just history. Red Hat discontinued support for Red Hat
4.0 when Red Hat 4.1 came out. In the meantime, they still continued to
support Red Hat 3.0.3. They discontinued support for 3.0.3 only after
"Ping of Death" etc. made it clear that the 1.2 kernel could no longer be
supported.  I suspect that this is going to be true of all of the RH 5.x
series too -- the "final" version (5.2?) will be the only one that
recieves long-term support.

> 
> > Now, personally, I am still running RH 5.0 here in the office on my
> > workstation, and with the latest set of upgrades it is trouble-free and
> > quite pleasant to use. I am running RH 5.1 on my machine at home and
> > having fun. But slobbering over techno-goodies doesn't make them
> > appropriate for use in a commercial setting. There are factors other than
> > technical superiority which explain why an operating system, or particular
> > versions of it, are useful for commercial purposes. 
> 
> Right, but keep in mind that Linux itself is a bleeding-edge operating
> system.  And likewise, for Redhat to stay in front in this arena, they
> need to be bleeding edge in many ways.  Also, the $50 you paid for 5.1 (if

I agree. And yes, I also agree that the $29.95 that I paid for 5.1 at
LinuxExpo is a nominal fee :-). Which is why I am glad that Red Hat
currently supports two versions of Red Hat Linux: A stable, proven
platform (Red Hat 4.2), and a "bleeding edge" platform (Red Hat 5.x). When
Red Hat ramps the revision engine to Red Hat 6.x, I suspect that they will
make the last revision of Red Hat 5.x their "stable" version. 

> building the release.  Supporting 3 or 4 releases back does not make sense

Agreed. Which is why I'm not particularly worried that they will
discontinue support for 5.0 shortly in favor of 5.1, because as long as
they keep around 4.2 support, I have a stable platform for deployment. 
BTW, you mention Caldera or BSDI as not so bleeding edge, but Caldera, in
my experience, has buggier userland tools than Red Hat does. Not
necessarily because Red Hat is so much better at quality control (have
they put the 8 errata items up for download for 5.1 yet?), but because Red
Hat dribbles the errata out as they occur, rather than trying to lump them
into "service packs" or into a new release. 

In short, I don't mind that they constantly rev the product, as long as
they keep around a stable version that I know I can rely on for a long
deployment life. That gives me the best of both worlds -- a bleeding edge
playtoy for the house, and a stable platform for deployment in the field. 
The only thing I ask of Red Hat is that they do not discontinue security
bug fix support for 4.2 until the 5.x platform has achieved its final form
(i.e., when they've announced a 6.x platform :-), at which point we can
start deploying 5.x in the field.

Eric Lee Green   [EMAIL PROTECTED]          Executive Consultants
Systems Specialist               Educational Administration Solutions
             See http://members.tripod.com/~e_l_green


-- 
  PLEASE read the Red Hat FAQ, Tips, Errata and the MAILING LIST ARCHIVES!
http://www.redhat.com/RedHat-FAQ /RedHat-Errata /RedHat-Tips /mailing-lists
         To unsubscribe: mail [EMAIL PROTECTED] with 
                       "unsubscribe" as the Subject.

Reply via email to