Fedora Core 6 release date pushed back
| From: | Jesse Keating <jkeating-AT-redhat.com> | |
| To: | fedora-devel-list-AT-redhat.com, fedora-test-list-AT-redhat.com, List for Fedora Package Maintainers <fedora-maintainers-AT-redhat.com>, fedora-announce-list-AT-redhat.com | |
| Subject: | Fedora Core 6 release date slip | |
| Date: | Fri, 6 Oct 2006 15:35:13 -0400 |
We regret to announce a slip of the Fedora Core 6 release schedule. A few issues are still present that we would like to see fixed before we release. - Possible ext3 corruption bug - Installs with 256megs of ram stall - Package ordering issues on multilib platforms (x86_64, ppc64) - SELinux issue with updating kernels on ppc platforms - iscsi based installations not functional There are obviously other issues and bugs still open, but these are the ones that are really "blocking" the release. To give enough time to fix these issues, we've extended the release date 6 days to Tuesday, Oct 17th. Freezes are still in place (even more so now). Your extra careful testing of rawhide over the next few days would greatly be appreciated. Keep an eye on http://fedoraproject.org/wiki/Core/Schedule for any changes. -- Jesse Keating Release Engineer: Fedora -- fedora-test-list mailing list fedora-test-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-test-list
Posted Oct 9, 2006 15:26 UTC (Mon)
by abatters (✭ supporter ✭, #6932)
[Link] (5 responses)
Posted Oct 9, 2006 16:12 UTC (Mon)
by jwb (guest, #15467)
[Link] (4 responses)
Posted Oct 9, 2006 16:15 UTC (Mon)
by zdzichu (subscriber, #17118)
[Link] (3 responses)
Posted Oct 9, 2006 18:41 UTC (Mon)
by JoeBuck (subscriber, #2330)
[Link] (2 responses)
Posted Oct 10, 2006 0:42 UTC (Tue)
by xoddam (guest, #2322)
[Link] (1 responses)
Posted Oct 11, 2006 5:29 UTC (Wed)
by davej (subscriber, #354)
[Link]
(And yes, it affects 2.6.18 [and probably still .19rc1] too).
Work is ongoing to get to the bottom of this.
Posted Oct 10, 2006 3:46 UTC (Tue)
by jd (guest, #26381)
[Link]
It'd be nice if RPM was cleaned up - it's a bugbear when you accumulate too many packages and have large updates. We're talking tens of thousands of packages and a couple of thousand updates at a time. The database accumulates cruft and requires regular --rebuilddb'ing to stop RPM hanging.
Also, if manually partitioning, just offer all the filesystems. It's easier, it lets me mix-n-match, and it's much more in line with the ethos of an OS that offers choices openly and freely.
I find Fedora Core to be overall one of the better Linux distributions, I've used Red Hat's stuff since it first came out and I'm impressed with the overall quality, but it really shouldn't be so easy for me to find problems with it. (Fedora's maintainers can take pride in the fact that I've found far more flaws in every other OS and most other Linux distros I've ever used, and in far less time, but the RPM bug is almost an antique and the other problems are careless.)
I would be concerned if Fedora Core 6 were to be released prior to Red Hat being confident that it'll install well on common machines. It would be far, far better for Fedora to verify "problem" machines are working as well as they should, even if it means adding a further day or four from the extended deadline. A quality release that's a day or so late will be remembered for being a quality release. A brown-paper-bag release that's on time will only ever be a brown paper bag, no matter how quickly the errors are corrected.
Posted Oct 11, 2006 11:24 UTC (Wed)
by Lovechild (guest, #3592)
[Link]
I've run FC6 since the cycle was opened and it's an absolutely amazing release: faster, blinger, more secure and naturally it adds new meaning to the word stability.
Better a little late than a tad likely to eat your data in the default configuration.
Dave Jone's blog entry about ext3 corruption
ext3 corruption
It is difficult to understand his blog post. Is he saying there's an ext3 corruption bug in the stock 2.6.18 kernel?ext3 corruption
Probably yes. Just in time, when SUSE thinks about switching to ext3...ext3 corruption
It's not clear whether he's saying the bug is in .18 or in the patch that is queued for .18.1.
ext3 corruption
The bug is in 2.6.18 and should be fixed by the patch, but for some ext3 corruption
reason the patch doesn't work in the presence of whatever other changes
are in the Fedora kernel, which implies it's not a correct fix.
It seems to be a completely unrelated bug to the one fixed by the aforementioned patch, that only manifests when you use a filesystem with a different block size to the CPU page size. Somehow I created a filesystem with 1K blocks on my testbox instead of the default 4K.ext3 corruption
I can install it. The last version I could install on my Dell (Dimension 2400) was test release 1. For test release 2, X rejected the graphics card. For 3, the installer segfaulted. C'mon, guys, I've never known any Fedora Core - even in the test releases - to be as hard to get going.
Just so long as...
The ext3 bug is fairly hard to trigger under normal conditions, it can be triggered easily on only a few machines and only under stresstesting conditions. The fact that the Fedora Project are pushing back the release to resolve issues like this shows a true dedication to quality.Fedora Core 6 release date pushed back