CentOS 6 release??

[AlanBartlett]

Agreed. If [i]CentOS 5[/i] provides what you require, just stay with it until it reaches EOL. :-D

[scottro]

It will depend upon each machine's purpose and how many machines you're discussing, which are mission critical, and so on.

There will be the advantage of better hardware support, possible performance increases with ext4. There will be the disadvantages of it becoming even more of desktop turned into a server, and more GUI--probably Gnome or KDE--oriented.

The virtualization has improved, but not that drastically. There will, of course, be newer versions of various programs and libraries. Sometimes, one may try to build a package and find it needs, on CentOS 5.x, newer versions of glibc.

It's going to depend upon the needs of the individual and company to generalize so errm, generally, that it means nothing.

[majun]

[quote]
AlanBartlett wrote:
Agreed. If [i]CentOS 5[/i] provides what you require, just stay with it until it reaches EOL. :-D[/quote]
Or even beyond. Our critical database servers are still running SuSE 9.3, which was discontinued years ago. Nobody dares to touch them though :-D

[woboyle]

Well, I really need the additional hardware support found in the newer kernel, so the sooner CentOS 6 is out, the better for me and my clients. In fact, I'd be happy to be a test site for it.

[gerdesas]

[quote]
woboyle wrote:
Well, I really need the additional hardware support found in the newer kernel, so the sooner CentOS 6 is out, the better for me and my clients. In fact, I'd be happy to be a test site for it.[/quote]

I hear that RHEL is shipping now. In fact, I do believe that you can get entitlements 24x7x365, so there will be no delay :-)

[helikaon]

i'm gonna attend the RH conference held in Prague on 18th January. They're gonna introduce RHEL 6 there, very interesting schedule, can't wait ... (also free food)

:-D

[pschaff]

Have fun and give us a report.

The silence on the CentOS 6 front is deafening. :-(

[mickrussom]

I like to at every release point out even a skeletal todo list with target and actual completed dates would be nice.

FreeBSD has a great example:

http://www.freebsd.org/releases/8.2R/schedule.html

http://wiki.freebsd.org/Releng/8.2TODO

Schedule
Action Expected Actual Description
Initial release schedule announcement - 08 October 2010 Release Engineers send announcement email to developers with a rough schedule.
Release schedule reminder - Weekly Release Engineers send reminder announcement e-mail to developers with updated schedule.
Code freeze begins 28 November 2010 29 November 2010 Release Engineers announce that all further commits to the stable/8 branch will require explicit approval. Certain blanket approvals will be granted for narrow areas of development, documentation improvements, etc.
BETA1 03 December 2010 10 December 2010 First beta test snapshot.
releng/8.2 branch 15 December 2010 21 December 2010 Subversion branch created, propagated to CVS; future release engineering proceeds on this branch.
RC1 17 December 2010 26 December 2010 First release candidate.
RC2 07 January 2011 - Second release candidate.
RELEASE build 21 January 2011 - 8.2-RELEASE built.
RELEASE announcement 24 January 2011 - 8.2-RELEASE press release.
Turn over to the secteam 31 January 2011 - RELENG_8_2 branch is handed over to the FreeBSD Security Officer Team in one or two weeks after the announcement.

This seems to be the most basic discipline in multi-contributor open-source projects - a bug list , a todo list, a schedule, something, a loose framework ...

But it seems that those "in the know" about what needs to be done and when it will be done by is a very short list of people.

The danger in this, as stated many times before, that is certain key men are gone, incapacitated or quit, the community still, as far as I can tell, have NO IDEA on how to make a new CentOS happen without these certain key people.

As always , the work is much appreciated, and the end result is great, but its only a matter of time before some catastrophe happens with the project and we have to seek refuge with SL or some other EL rebuild.

I would say to those who like to keep this whole thing black box is that its hard to convince the giving of corporate funded gifts when the project seems like a University Science Project at major release time.

[pschaff]

Another example - the [url=https://www.scientificlinux.org/distributions/rolling.whiteboard]SL Rolling Whiteboard[/url].

[scottro]

FreeBSD, and actually, all the BSDs, I think, are known for their documentation. Linux is also known for its documentation, but not in the same way. The word "infamous" springs to mind.

Trollingly yours......