CentOS - May 2005 - Re: Demonizing generic Linux issues as Fedora Core-only issues -- WAS: Hi, Bryan

From: Les Mikesell <lesmikesell at gmail.com>
> I'm still wondering about that... If anyone except Linus himself
> even suggested that changing kernel interfaces in a way that would
> break device drivers was a good thing, I can't imagine the reaction.
> I could see that the changes through 2.4 were improving things, but
> is there anything that is measurably better in 2.6 (at least compared
> to the RH-patched 2.4)?  I've been too busy trying to make some
> firewire drives work as well as they did on FC1 to notice any other
> changes.
Agreed.  But even Red Hat stands on the shoulders of Linus and the
kernel developers.  No matter how many people Red Hat puts on the
kernel, there are still things outside of Red Hat's control.

If you don't like what does or doesn't go into RHEL/CentOS, then you
need to get involved with Fedora Core and make a case of what should.
And one way to do that is to donate time.  Because if it's not proven
in Fedora Core to Red Hat's tastes, it doesn't go into RHEL/CentOS.

So the most effective way you can influence RHEL, and therefore
CentOS, is to get involved with Fedora Core.  Red Hat is _not_ going
to "force" something that it can't get to work with Fedora Core
into
RHEL arbitrarily.  So if you don't see something in RHEL/CentOS as
standard, you need to see why it didn't make it into Fedora Core.

Once again, I will remind people that several people here _are_ indeed
complaining about what doesn't come with RHEL, and not aware that
it _does_:  
A)  Not work in Fedora Core, and failure after failure keeps it out
B)  And is, subsequently, not going into RHEL where SLAs are involved
C)  Does _not_ come with pretty much _all_ other, similar distros

Red Hat Enterprise Linux, like SuSE Linux Enterprise Server, are _not_
about features.  Both 18 month releases are subsets of their 6 month
revisions (even locales not supported in SLA are often removed).  You
are not paying for R&D and inclusion in the "Enterprise" flavor
(although
the funds _do_ pay for extensive GPL development, R&D and other
integration attempts in Fedora Core and SuSE Linux).
> Or with the logic of asking customers to pay extra to get something with
> features removed...
Again, the people who actually _pay_ for RHEL/SLES are _not_ paying for
features.  ;->

You are complaining about CentOS not coming with things standard, and
then assinging blame for that to Red Hat, because you believe they are
supposedly paid and should have more features.  I can only label this as
"ignorance" of how RHEL (as well as SLES) is developed, because anyone
can go out to Bugzilla and see the detail of why something is not included.

If you want the most features, then Fedora Core + { FE+Lorg, DAG, etc...}
is your baby, maybe CentOS + additives if you don't mind waiting a year
later.
I think it's great that CentOS is trying to be the best of both worlds: 
SLA-
quality with added features.  But it really is "demonizing" when all I
see is
people making complaint after complaint about RHEL not having something
when many, many, _many_ Fedora Development manhours were made trying
to get exactly what you wanted working.

At some point Red Hat has to say "this doesn't work, even if it works
in
2-3 months, it still won't have time to 'bake' in usage, so
we're dropping
it from consideration for this version of RHEL."
> I haven't seen anyone demonizing RedHat here.  I see some people
> reporting painful experiences,
But people are going beyond reporting painful experiences.  They are
making conclusions and get "upset" when I give the technical details
of
why something wasn't included.  If you care enough to want something
I assume people are open enough to understand why it wasn't included.

Microsoft is the company that doesn't do something unless their is
at least 250,000 users.  Red Hat, SuSE and others often try to do
everything they can, even for maybe only 1,000 users, in their 6 month
revisions.  But at some point, if something really doesn't work, has
security issues and just isn't something they want to stake their time
and money on in a SLA (which can get _very_expensive_ for software
company), they have to drop it.

After all, after a thousand man-hours of attempted integration for even
their Fedora Core release, are they really going to sign themselves up
for yet another thousand-plus hours and nagging re-fix after re-fix when
maybe only a couple hundred licenses of RHEL are actually going to hold
them to it?

Just remember what CentOS is based on, and how much work not only
goes into that, but into making sure the distro works as advertised
because SLAs are involved and that can get really costly for Red Hat
if it ships something that is not well tested.  Especially if only a few
companies want it.
> but I don't think anyone expects perfection and the best thing is to
> learn from them.
If people here should understand one thing from the Windows world,
it is that features and quality can be mutually exclusive -- especially
the farther you move away from the economies of scale.

I just find it humorous that several people who use CentOS, a free
redistribution of RHEL (where Red Hat doesn't see a dime), are not
complaining about Linus & co. for the kernel, not complaining about
the CIPE team and their lack of movement on kernel 2.6, not
volunteering to even look at the Bugzilla reports to find out what
issues were blantantly repeatable, let alone other distros that have
the same issues ...

But blaming solely Red Hat for this issue (and this is just one example).

But I expect that because some people just blame Microsoft for things
outside of their control too.  Sometimes the only person to blame is
the lack of interest by anyone.

If that isn't ignorance, I don't know what is.



--
Bryan J. Smith   mailto:b.j.smith at ieee.org

On Wed, 2005-05-25 at 09:43 -0500, Bryan J. Smith  wrote:
> If you don't like what does or doesn't go into RHEL/CentOS, then
you
> need to get involved with Fedora Core and make a case of what should.
> And one way to do that is to donate time.  Because if it's not proven
> in Fedora Core to Red Hat's tastes, it doesn't go into RHEL/CentOS.
> 
> So the most effective way you can influence RHEL, and therefore
> CentOS, is to get involved with Fedora Core.  Red Hat is _not_ going
> to "force" something that it can't get to work with Fedora
Core into
> RHEL arbitrarily.  So if you don't see something in RHEL/CentOS as
> standard, you need to see why it didn't make it into Fedora Core.
> 
Agreed that getting stuff into FC is the way to get them tested and into
RHEL proper ... then they will be standard in the base CentOS product.

<snip>
> Again, the people who actually _pay_ for RHEL/SLES are _not_ paying for
> features.  ;->
> 
Agreed ... RHEL is not the place for experimentation, only well working
and proven things should go in there.  Which is why I think SELinux is a
little premature in this version RHEL.

<snip>
> If you want the most features, then Fedora Core + { FE+Lorg, DAG,
> etc...} is your baby, maybe CentOS + additives if you don't mind
> waiting a year later.
This is not really true. We will probably never have everything that FC
has as added features ... but CentOS-4 has several added features and we
have been out for only a 3 months (so, not required to wait a year).  We
even have some features (mysql compiled postfix, NX/freeNX, and soon
MySQL-Administrator / MySQL-Query Builder) that are not yet part of FC
or FC Extras at all.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL:
<http://lists.centos.org/pipermail/centos/attachments/20050525/52df9453/attachment-0004.sig>

On Wed, 2005-05-25 at 09:43, Bryan J. Smith  wrote:
> From: Les Mikesell <lesmikesell at gmail.com>
> > I'm still wondering about that... If anyone except Linus himself
> > even suggested that changing kernel interfaces in a way that would
> > break device drivers was a good thing, I can't imagine the
reaction.
> > I could see that the changes through 2.4 were improving things, but
> > is there anything that is measurably better in 2.6 (at least compared
> > to the RH-patched 2.4)?  I've been too busy trying to make some
> > firewire drives work as well as they did on FC1 to notice any other
> > changes.
> 
> Agreed.  But even Red Hat stands on the shoulders of Linus and the
> kernel developers.  No matter how many people Red Hat puts on the
> kernel, there are still things outside of Red Hat's control.
Yes, but... whose choice was it to ship 2.6 with lots of broken
and omitted stuff when 2.4 works better for many things?  Just
because a developer writes some code and tacks on a higher
version number doesn't mean it's ready for prime time.
> Once again, I will remind people that several people here _are_ indeed
> complaining about what doesn't come with RHEL, and not aware that
> it _does_:  
> A)  Not work in Fedora Core, and failure after failure keeps it out
> B)  And is, subsequently, not going into RHEL where SLAs are involved
> C)  Does _not_ come with pretty much _all_ other, similar distros
Well, an occasional rant is therapeutic - or at least cheaper than
smashing the equipment that no longer works after you upgrade the
software.
> You are complaining about CentOS not coming with things standard, and
> then assinging blame for that to Red Hat, because you believe they are
> supposedly paid and should have more features.  I can only label this as
> "ignorance" of how RHEL (as well as SLES) is developed, because
anyone
> can go out to Bugzilla and see the detail of why something is not included.
I'm questioning the sensibility of shipping a kernel in what should be
an upgrade with many things that don't work at all that worked in
the previous version.  As you point out, these aren't surprises.
While I hate to complain about free software, I think that user's
experiences are relevant and should be reported even if they are
painful.  
> If you want the most features, then Fedora Core + { FE+Lorg, DAG, etc...}
> is your baby, maybe CentOS + additives if you don't mind waiting a year
later.
> I think it's great that CentOS is trying to be the best of both worlds:
SLA-
> quality with added features.  But it really is "demonizing" when
all I see is
> people making complaint after complaint about RHEL not having something
> when many, many, _many_ Fedora Development manhours were made trying
> to get exactly what you wanted working.
I'm not sure I believe that in the case of CIPE, since the 1.6 version
specifically addresses the changes in the 2.6 kernel and was available
well before FC3 or RHEL4 releases which did not include it again. If
I were to speculate about intentions as you suggest above, I'd guess
that someone at RedHat read the one negative review published about
CIPE, decided it was no longer a selling point, and walked away from
the integration they had done before.  But of course I wouldn't
speculate about something like that...
> I just find it humorous that several people who use CentOS, a free
> redistribution of RHEL (where Red Hat doesn't see a dime), are not
> complaining about Linus & co. for the kernel, not complaining about
> the CIPE team and their lack of movement on kernel 2.6, not
> volunteering to even look at the Bugzilla reports to find out what
> issues were blantantly repeatable, let alone other distros that have
> the same issues ...
It's a complicated system, and it isn't immediately obvious that other
distros have exactly the same problems.   If you follow
www.distrowatch.com for a few months you'll see that they each
keep trying to improve things in different ways. 
> But blaming solely Red Hat for this issue (and this is just one example).
Switching distros should always be an option - that's one of the
big reasons for using open source.  But, in spite of what I consider
an occasional bad choice, Fedora/RH seem to be among the best and
I do give the fact that they've been responsible for getting buggy
software in front of a vast number of people credit for it eventually
being fixed.  If you remember the state of free software before
RH 4.0 was released you will know what I mean.  I do sometimes wonder
how things would have turned out if someone had built an equally
easy to install CD based on freebsd first, though.  
> But I expect that because some people just blame Microsoft for things
> outside of their control too. 
Hmmm, just who would you blame for WindowsME?
> Sometimes the only person to blame is
> the lack of interest by anyone.
> 
> If that isn't ignorance, I don't know what is.
Who would you blame for the upgrade that made interfaces not
start if the hardware address in the configs didn't match
the NIC?  That was bad news for my remote machines where all
the drives had been cloned and the IP addresses set before
shipping.  I'm not positive about this, but I think it happened
at about the same time across Fedora/RH/Centos updates instead
of following the scenario you mentioned for testing things that
are supposed to change behavior.

-- 
  Les Mikesell
   lesmikesell at gmail.com