[00:29:05] --- Russ has become available [00:40:57] The Debian build slave is screwed - it has been for a while. I suspect that the whole VM needs reinstalled. [01:00:09] --- Russ has left: Disconnected [01:21:36] --- haba has become available [04:02:49] Can anyone think of a platform we support that isn't matched by alpha_dux*|sgi_*|sun*_5*|rs_aix*|*linux*|hp_ux11*|ia64_hpux*|*[nof]bsd*|*_darwin_* ? [04:05:59] --- jaltman has left: Disconnected [04:20:36] --- haba has left [04:54:58] --- haba has become available [04:59:13] --- reuteras has left [05:27:41] --- meffie has become available [06:13:23] alpha_dux probably doesn't build anymore [06:13:52] and also, sure. windows. [06:36:48] lol, commit msg on 5025. [06:56:45] --- deason/gmail has become available [07:38:59] hmm, when installing the src.rpm: error: unpacking of archive failed on file /afs/bx.psu.edu/user/phalenor/rpmbuild/SOURCES/CellServDB.2010-12-13;4e2050f6: cpio: MD5 sum mismatch [08:36:25] can we remove the whole series of echos in the rpm postinstall about configuring the cache? I can't think of any other rpm package that spits out stuff like that during install [08:42:40] phalenor: Which platform are you unpacking the RPM on? [08:42:55] My suspicion is that Derrick may have created it on something with a 'new' rpm. [08:44:40] i did. centos 6 [08:44:49] Yeah. Don't do that. [08:44:59] i guess that was foolish. [08:45:21] yeah, this is 5. I guess 6 switches to sha-something? [08:45:25] I think the cut off is around Fedora 12, and RHEL6 [08:45:33] Indeed. And isn't backwards compatible. Go figure. [08:46:16] I think you can work round it if you suppress the hash checks, but that probably isn't a good plan. [08:46:22] --nomd5 [08:46:27] Yeah. [08:46:36] have to do that all the time when backporting fedora packages co rhel5 [08:47:01] I just have a RHEL5 compatible RPM that understands SHA1 hashes. [08:47:10] It's the only way that the build farm could continue to function. [08:47:12] anyone that's going to be rebuilding from src.rpm probably knows about that switch, so I wouldn't worry for now [08:47:43] Yeah, not hugely bothered from pre7 (rather not have to respin the builds because the SRPM has changed), but we should try and avoid it for final. [08:47:43] package seems to have built fine, about to reboot into that plus a new kernel shortly [08:52:43] Cool. Let us know how you get on! [08:53:22] With regards to printing that message, at one point quite a few RPMs gave you hints about further configuration required by the package at the end of their installs. I think with the rise of yum, and GUI installers, there's less of it around. [09:01:01] I'm sure there were some at one point that would give you messages like that, but now it just looks like an error or something at first glance [09:01:27] and suggesting to use pam_afs.so probably isn't what most people want [09:02:23] Yeah. That's definitely true. [09:02:31] The spec file is in git. Feel free to submit a patch. [09:02:52] I just might. [09:03:15] Please do. [09:11:42] pushed to master [09:49:40] --- haba has left [10:40:27] (The zone.mit.edu cell has been running pre7 since last night, apparently without issue.) [10:58:52] --- summatusmentis has left [12:46:42] --- rra has become available [13:16:13] --- meffie has left [13:58:57] --- mfelliott has left [14:33:57] I hope to get one of our web/imap/ftp servers upgraded this weekend [14:34:14] those are running pre5+patch right now, apparently without issue [15:12:56] --- mfelliott has become available [15:26:52] --- deason/gmail has left [15:30:06] --- summatusmentis has become available [16:33:02] 1.6.0pre7 builds should be popping out of the RPM buildfarm any time now. [17:35:16] --- rra has left: Disconnected [17:56:25] --- Russ has become available [18:06:12] --- Russ has left [19:58:21] --- jaltman/FrogsLeap has become available [20:06:11] any chance of getting dev.mit.edu to run pre7 or at least the server with project/winaftest/... on it? [20:07:03] it would be to see the rx improvements over the wan. [20:54:35] dev.mit.edu is slated to die in the forseeable future. [20:55:26] How much space does winafstest actually need? [21:10:02] Also, project.winafstest -c dev is a replicated volume.