[02:40:36] ** bitmonk has left IRC () [05:59:04] [connected at Fri Jan 11 05:59:04 2008] [05:59:04] <> *** Looking up your hostname... [05:59:04] <> *** Checking ident [05:59:04] <> *** No identd (auth) response [05:59:04] <> *** Found your hostname [05:59:04] <> *** Your host is clarke.freenode.net[clarke.freenode.net/6667], running version hyperion-1.0.2b [05:59:06] [I have joined #peak] [05:59:06] ** clarke.freenode.net set the topic to does anyone remember the real topic ? - not really ... [09:14:43] ** apoirier has joined us [11:10:24] ** bitmonk has joined us [13:31:12] ** pje has joined us [13:39:46] ** apoirier has left IRC ("KVIrc 3.2.6 Anomalies http://www.kvirc.net/") [14:02:42] ** bitmonk_ has joined us [14:15:49] ** bitmonk__ has joined us [14:22:41] ** bitmonk has left IRC (Read error: 110 (Connection timed out)) [14:39:24] ** bitmonk_ has left IRC (Connection timed out) [16:01:32] pje is now known as pje_mtg [16:52:13] pje_mtg is now known as pje [18:24:31] zooko is now known as zookonaptime [19:31:30] ** bitmonk has joined us [19:32:46] zookonaptime is now known as zooko [19:45:39] ** bitmonk__ has left IRC (Read error: 110 (Connection timed out)) [19:52:12] ** bitmonk has left IRC (Connection timed out) [20:11:06] Dear pje: see all the green buildslaves: http://allmydata.org/buildbot/waterfall?show_events=false [20:11:14] You can scroll down to see that they were all red a couple of hours ago. [20:11:27] eh? [20:11:39] This is because a couple of hours ago I was using "misc/dependencies/setup.py easy_install --always-unzip misc/dependencies", in order to install a bunch of dependencies unzipped. [20:11:50] oh... bad. [20:11:52] Very bad [20:11:56] I gave up and changed it back to "misc/dependencies/setup.py install", so now it works, but not all of them are unzipped. [20:12:02] You must always run a setup.py in its own directory. [20:12:07] Yes, I've decided that I have no idea what that command was supposed to do. [20:12:11] Or Bad Things Will Happen(tm) [20:12:13] Ah, you don't know yet [20:12:27] that the "setup.py" in question is a tiny package which has install_dependencies on all the other packages. [20:12:46] So it is not the package in question that needs to be installed unzipped (which can be accomplished with zip_safe=False), but rather its dependencies. [20:13:01] I don't care what it is... if it's a distutils or setuptools setup.py, and you run it from a different directory... and it *works*... it's only accidental. [20:13:10] So, now I'm back to either (a) allowing my coworker Rob to go ahead with his "hatch" script which unzips any .egg zip files that it finds, or [20:13:10] ... [20:13:21] Oh, the fact that the CWD wasn't misc/dependencies when I ran it? That's easy to fix. [20:13:44] right, that's what I'm saying. Never ever give a path to a setup.py other than './' [20:13:44] Would you expect cd misc/dependencies && ./setup.py easy_install --always-unzip . [20:13:45] to work? [20:13:50] I'll try that. [20:14:13] You also need --always-copy [20:14:32] Thanks. [20:14:33] Otherwise, if you have an egg sitting there, it'll be on sys.path, so it won't bother copying it (or unzipping it) to your target [20:14:44] Ah. Interesting. [20:14:45] Thanks. [20:15:17] --always-copy tells easy_install not to pay attention to something already being on sys.path, unless it's already in the target location. [20:15:28] it will always copy (or unzip) it to the target [20:15:30] Neat. Thanks for the note. [20:31:25] ** bitmonk has joined us [20:32:20] I would really like to be able to programmatically determine the version numbers of packages at run-time, even if those packages aren't in eggs. [20:32:37] Is the distutils in some near-future version of Python going to produce $pkg.egg-info/PKG_INFO files by default? [20:33:22] 2.5 does, actually [20:33:38] in the form of $pkg.egg-info *files* that contain the PKG-INFO data. [20:33:45] Setuptools recognizes that format as a valid egg. [20:34:21] Hm. [20:34:28] 2.5 distutils generates those files and installs them... unless of course the packaging system (by which I mean boneheaded sysadmins) fights back by removing them. [20:34:59] What's the best programmatic way to discover and parse those files? [20:35:13] I mean, I could write code that examines sys.path, checks for $pkg.egg-info files therein... [20:36:33] Strange.. I'm looking at an Ubuntu Gutsy system -- the latest stable release of Ubuntu. [20:36:37] There is a /var/lib/python-support [20:37:37] Some of the packages therein have accompanying $packagename.egg-info files, others don't. [20:37:39] * zooko investigates. [20:37:42] ** bitmonk_ has joined us [20:39:29] Hm.. Interesting: https://lists.ubuntu.com/archives/ubuntu-motu/2007-May/001617.html [20:40:13] The next release of Ubuntu, out in April, is the new Long Term Support release, and I hope to be in a position of supporting users of my software on that particular version of Ubuntu for many years. [20:47:13] ** bitmonk_ has left IRC (Read error: 104 (Connection reset by peer)) [20:47:26] ** bitmonk_ has joined us [20:52:25] ** bitmonk has left IRC (Read error: 110 (Connection timed out)) [22:01:27] ** pje has left IRC ("Client exiting") [22:34:49] ** bitmonk_ has left IRC () [22:58:38] ** bitmonk has joined us [23:02:15] aaaaa