| *** tpb has joined #tp | 00:00 | |
| *** ChanServ sets mode: +o tpb | 00:00 | |
| *** guitsaru has joined #tp | 00:27 | |
| *** pschulz01 has joined #tp | 00:56 | |
| pschulz01 | #tp | 00:56 |
|---|---|---|
| mithro | hey people | 01:00 |
| mithro | hey pschulz01 | 01:00 |
| pschulz01 | Might not be here long.. going to look at baby stuff | 01:01 |
| mithro | ~seen niphree | 01:01 |
| tpb | mithro: niphree was last seen in #tp 1 day, 6 hours, 5 minutes, and 56 seconds ago: <niphree> good night all | 01:01 |
| mithro | looks like demo1 crash again | 01:02 |
| pschulz01 | mithro: libtproto2-py ready for packaging? | 01:02 |
| mithro | no | 01:02 |
| pschulz01 | mithro: libtproto-py ? | 01:02 |
| mithro | it probably won't even run on anyone elses computer part from mine | 01:02 |
| mithro | libtproto-py can be packaged | 01:03 |
| pschulz01 | mithro: What supplies tp.netlib, tp.client ? | 01:03 |
| pschulz01 | :-) | 01:03 |
| mithro | libtpproto-py -> tp.netlib | 01:03 |
| mithro | libtpclient-py -> tp.client | 01:03 |
| pschulz01 | Ta. | 01:05 |
| mithro | but for tpclient-pywx-dev you'll want the development branch of tp.client | 01:06 |
| pschulz01 | Ok.. just working on tpproto | 01:08 |
| pschulz01 | You are attempting to install a package to a directory that is not | 01:14 |
| pschulz01 | on PYTHONPATH and which Python does not read ".pth" files from. The | 01:14 |
| pschulz01 | installation directory you specified (via --install-dir, --prefix, or | 01:14 |
| pschulz01 | the distutils default setting) was: | 01:14 |
| pschulz01 | tmp/lib/python2.5/site-packages/ | 01:14 |
| pschulz01 | and your PYTHONPATH environment variable currently contains: | 01:15 |
| pschulz01 | '' | 01:15 |
| pschulz01 | This was using.. | 01:15 |
| pschulz01 | python setup.py install --prefix=tmp | 01:15 |
| pschulz01 | for libtpproto-py | 01:16 |
| mithro | I think you can ignore that warning | 01:23 |
| mithro | as the package will install it to the python-path when you do so | 01:24 |
| pschulz01 | mithro: OK - I used PYTHONPATH to fake it out. | 01:25 |
| pschulz01 | mithro: Bonus.. install tests run | 01:25 |
| mithro | install tests!? | 01:26 |
| pschulz01 | PYTHONPATH="debian/libtpproto-py/lib/python2.5/site-packages" \ | 01:26 |
| pschulz01 | python setup.py install --prefix=debian/libtpproto-py | 01:26 |
| pschulz01 | Well.. it ws trying to compile up stuff. | 01:27 |
| mithro | did you get a package? | 01:35 |
| pschulz01 | http://git.mawsonlakes.org/git/tp/libtpproto-py | 01:36 |
| pschulz01 | Yes.. | 01:36 |
| tpb | Title: Index of /git/tp/libtpproto-py (at git.mawsonlakes.org) | 01:36 |
| pschulz01 | but it didn't work after installing.. tpclient-pywx can't find tp.netlib | 01:36 |
| mithro | hrm.... | 01:36 |
| mithro | what is the contents of the deb? | 01:37 |
| pschulz01 | http://pastebin.org/378 | 01:38 |
| tpb | Title: pastebin - collaborative debugging tool (at pastebin.org) | 01:38 |
| pschulz01 | I have to disappear.. I'll be back later. | 01:38 |
| mithro | pschulz01: I'm afraid that deb is quite wrong :/ | 01:43 |
| mithro | firstly - debs must never contain .pyc's | 01:43 |
| mithro | # | 01:44 |
| mithro | -rw-r--r-- root/root 2362 2007-08-11 14:54 ./lib/python2.5/site-packages/site.py | 01:44 |
| mithro | # | 01:44 |
| mithro | -rw-r--r-- root/root 217 2007-08-11 14:54 ./lib/python2.5/site-packages/easy-install.pth | 01:44 |
| mithro | those two files shouldn't be in the deb | 01:44 |
| mithro | they need to be modified | 01:44 |
| mithro | (by the package install process) | 01:44 |
| mithro | http://wiki.debian.org/DebianPythonFAQ | 01:47 |
| tpb | Title: DebianPythonFAQ - Debian Wiki (at wiki.debian.org) | 01:47 |
| mithro | try the "Python eggs" section? | 01:47 |
| mithro | http://python-modules.alioth.debian.org/python-central_howto.txt | 01:48 |
| tpb | <http://ln-s.net/ofc> (at python-modules.alioth.debian.org) | 01:48 |
| TBBle | Didn't I push package scripts for tpproto-py and tpclient-py? | 03:00 |
| mithro | TBBle: yes, kind of | 03:01 |
| mithro | TBBle: I don't think we ever successfully built debs from them? | 03:02 |
| TBBle | Kind of? Were they broken? | 03:02 |
| TBBle | Sorry I dropped off the earth after that, BTW. Project's ramping up, again, I've just checked my email properly for the first time in a fortnight. | 03:02 |
| TBBle | Weird. I'd built them here before pushing them. *updates his local copy to see what's up* | 03:03 |
| mithro | TBBle: yeah, but we are now trying to build "from git" | 03:05 |
| mithro | rather then real packaged versions | 03:05 |
| TBBle | Yeah, that's what I pushed. What I was supposed to do next was make them work with the released tarballs. | 03:05 |
| mithro | TBBle: oh? | 03:05 |
| TBBle | I just build a python-tp-netlib package from git. | 03:05 |
| TBBle | Uh, built | 03:05 |
| mithro | oh? | 03:06 |
| TBBle | It doesn't handle updating the changelog string, mind you. Prolly should, but that wouldn't be too hard. I'm guessing from discussion earlier someone's already got something that throws a git commit string into the version number. | 03:06 |
| mithro | TBBle: how does one build this magical deb? | 03:06 |
| mithro | btw, do you know if BigWorld is in need of any Python Programmers? /me is in the market for a job :P | 03:07 |
| TBBle | Go into the directory. dpkg-buildpackage -us -uc -b -rfakeroot | 03:07 |
| TBBle | mithro: Actually, they very well may be... I remember something being emailed around, but I don't remember what sort of specific programmers they were looking for, if any. | 03:08 |
| TBBle | If you want, email me at work (I'll /query you the address) and I'll give you the details. (Sorry, they're in my work email, don't have them here) | 03:08 |
| mithro | TBBle: i've been meaning to finish get my Resume in order | 03:10 |
| TBBle | Oh, I see, there's code in the version.py file to fetch the git revision... There ought to be an easier way, though. Also, that code doesn't produce numerically-increasing version numbers for later git-commits. | 03:10 |
| mithro | TBBle: no | 03:11 |
| TBBle | No? | 03:11 |
| mithro | TBBle: the version.py in tpclient-pywx has some stuff which might be useful | 03:12 |
| mithro | tim@ultraslim:~/oss/tp/tpclient-pywx-dev$ python version.py | 03:12 |
| mithro | 0.2.99+0.3.0.20070811 (git fdeb5cc90d34b26943a738172fade7d44142112c) | 03:12 |
| mithro | I can add similar stuff to libtpproto-py if it helps | 03:13 |
| TBBle | (I assume the 'master' branch is the right one for these, BTW...) | 03:13 |
| TBBle | Ah, that must be on the other branch? | 03:14 |
| mithro | yeah, the development branch | 03:14 |
| mithro | hows this look? | 03:15 |
| mithro | tim@ultraslim:~/oss/tp/libtpproto-py$ python ./tp/netlib/version.py | 03:15 |
| mithro | 0.2.3+0.2.4.20070811 (git 9c5e10eaef0b2cc809b0f6006f8cfbad8371e9b6) | 03:15 |
| TBBle | That's not bad. Of course, it'll still need a little munging for Debian. | 03:16 |
| TBBle | Hmm. Isn't this what update-debian-changelog in the debian/ tree does? | 03:18 |
| TBBle | (Of python-pywx, I mean) | 03:18 |
| CIA-23 | mithro libtpclient-py-development * r999072b36328 /tp/client/ (cache.py media.py threads.py): Removed a lot of excess prints. | 03:22 |
| CIA-23 | mithro libtpproto-py * r8a6bd385a875 /tp/netlib/version.py: Added more versioning information. | 03:22 |
| CIA-23 | mithro libtpclient-py-development * r6ccfe8d757b7 /tp/client/__init__.py: Register a _ function if one doesn't exist. | 03:22 |
| CIA-23 | mithro libtpclient-py-development * r4cec2d6ca0c9 /tp/client/version.py: Added more versioning information. | 03:22 |
| mithro | TBBle: update-debian-changelog seems to generate an updated changelog | 03:24 |
| TBBle | Yeah. Which is what you do if you want the package to have a different version. | 03:24 |
| mithro | TBBle: so just run a script like that first? | 03:28 |
| TBBle | Yeah, something like that. | 03:30 |
| TBBle | Beh. If I from tp.netlib.version import version,version_str, I don't get the .git part, just 0.2.3 | 03:31 |
| TBBle | Oh, wait, that's what I get for skimming. | 03:31 |
| *** zipola has joined #tp | 03:33 | |
| TBBle | Of course, if the file is not installed in a .git directory, version_git and target version won't be defined. Would be good to have an else: which sets those to None, so a script in debian/ can just import those, and if it sees the None, know not to touch the changelog since we're a release tarball. | 03:34 |
| mithro | python-tp-netlib_0.2.3+0.2.4.20070811-1auto1_all.deb | 03:35 |
| TBBle | Hmm. Maybe git-describe would be a better result? | 03:35 |
| mithro | git-describe? | 03:36 |
| mithro | it doesn't use any git commands | 03:36 |
| mithro | just reads the .git directory directly | 03:37 |
| TBBle | Yeah, but it could. If we can run git-describe, and it doesn't return 1, then it gives a much more useful versioning string. | 03:37 |
| mithro | TBBle: it does? | 03:37 |
| CIA-23 | mithro libtpproto-py * r03707ba1998a / (3 files in 3 dirs): Add support for auto-building. | 03:38 |
| TBBle | Yeah. | 03:38 |
| mithro | seems all we need to do is | 03:38 |
| mithro | cd debian | 03:39 |
| TBBle | git-describe my current libtpptorot-py: LIBTPPROTO_0_2_2-72-g8a6bd38 | 03:39 |
| TBBle | That's 'tag'-'commits since tag'-g'commit sha1' | 03:39 |
| mithro | python update-debian-changelog | 03:39 |
| mithro | cd .. | 03:39 |
| TBBle | So you _know_ it's always increasing, unless someone moves the tag. | 03:39 |
| mithro | dpkg-buildpackage -us -uc -b -rfakeroot | 03:39 |
| TBBle | Yeah, that's what I'm saying. I'm suggesting tp/netlib/version.py call git-describe to get the git version, rather than manually parsing around in .git | 03:40 |
| mithro | hrm... | 03:40 |
| TBBle | Well, I guess that assumes the tags are increasing... | 03:40 |
| mithro | no debian directory in the development branch? | 03:41 |
| TBBle | I dunno. I'm working in master, I dunno if I committed one to development or not. | 03:42 |
| TBBle | git-describe also has the advantage that if you build your package again the next day, the version number doesn't change. ^_^ | 03:44 |
| TBBle | Of course, I've never ``d in python before. | 03:45 |
| mithro | ``d ? | 03:47 |
| mithro | he he, this is a little confusing | 03:50 |
| mithro | both our debian packagers are a "Paul" :P | 03:50 |
| CIA-23 | tim libtpclient-py-development * r8da6e49c7f78 /debian/ (changelog compat control copyright rules watch): | 03:53 |
| CIA-23 | Debian packaging goodness. | 03:53 |
| CIA-23 | Kindly produced by Paul Hampson aka TBBle. | 03:53 |
| CIA-23 | (cherry picked from commit c394cb09bff8a76168b0684498c2c858569715f2) | 03:53 |
| CIA-23 | Paul.Hampson libtpclient-py-development * rb4bca9b1237a /debian/rules: | 03:53 |
| CIA-23 | Remove workaround for setup.cfg left next to setup.py | 03:53 |
| CIA-23 | (cherry picked from commit abbadb9a84716d5eac3522f22ef0622c3ec97cf9) | 03:53 |
| CIA-23 | Paul.Hampson libtpclient-py-development * r0b04ebf09944 /debian/changelog: | 03:53 |
| CIA-23 | Update Debian packaging to current version number | 03:53 |
| CIA-23 | (cherry picked from commit 1e2c8a007c8b75ae8a7bffaca70b14ef6e23cb7c) | 03:53 |
| CIA-23 | Paul.Hampson libtpclient-py-development * re3c4279c5ec5 /debian/ (changelog control): | 03:53 |
| CIA-23 | Build-depend on python-all, to ensure runtimes are present | 03:53 |
| CIA-23 | (cherry picked from commit a6bd007ec7cddc689e941078824040fa56382d13) | 03:53 |
| CIA-23 | mithro libtpclient-py-development * r303119404a90 / (4 files in 3 dirs): Updated to make debian packaging working. | 03:53 |
| TBBle | ``d, like in bash. In perl, you can also $str = `git-describe`;. I dunno if python has such a syntactic sugar.... | 03:54 |
| TBBle | OK. So I should play with the development branch, when I play with this stuff? | 03:55 |
| mithro | TBBle: sure! | 03:56 |
| TBBle | OK. | 03:56 |
| TBBle | Ah. tpproto-py doesn't have a development branch. ^_^ | 03:56 |
| mithro | yeah | 03:57 |
| mithro | libtpproto-py doesn't have a dev branch | 03:57 |
| mithro | but both libtpclient-py and tpclient-pywx do | 03:57 |
| mithro | TBBle: you are probably right about the GIT-version - I don't like the look of the tag however? | 03:57 |
| TBBle | Well, they _are_ your tags... | 03:58 |
| TBBle | You could easily trim off the stuff before the first -, I guess. | 03:58 |
| TBBle | Or rather, before the second-last -. | 03:58 |
| TBBle | It'll be .*-\d+-g[0-9a-f]+ as a regex. | 03:59 |
| TBBle | Unless you happen to be on the tag exactly. | 03:59 |
| TBBle | I wonder if git-tags are forbidden from having - in them... | 03:59 |
| TBBle | Doesn't seem like it. | 04:03 |
| TBBle | Ah, popen is the thing. | 04:07 |
| mithro | TBBle: yeah, I think you would want to reformat the tags a little | 04:08 |
| TBBle | I'm just throwing something together, I'll get you to have a look at it before I do anything with it. | 04:11 |
| TBBle | Hmm. This highlights the fact that 0.2.3 wasn't every tagged, or if it was, git-remote fetch isn't grabbing it. | 04:12 |
| * mithro ponders if 0.2.3 was ever released | 04:17 | |
| TBBle | Hehe. Turns out git-describe produces something which is actually treeish. | 04:18 |
| TBBle | So I'd argue for including it literally, since you can trivially git-diff <git-commit-description> to see what's changed since that commit. | 04:18 |
| mithro | okay | 04:20 |
| mithro | but what about in the package name? | 04:20 |
| mithro | 0.2.2 is the current "released" version of libtpproto-py | 04:21 |
| CIA-23 | mithro libtpproto-py * rdfe9a4f729de /tp/netlib/version.py: Fixed the version number now we also have a "development version". | 04:23 |
| TBBle | True. But the point is if someone's ferreting away on their own version, and puts their own tag in a tree, you want to know about it in any package they roll. | 04:27 |
| TBBle | Hmm. This also has the issue that once it's packaged and installed, it loses its .git versioning. Does that matter? | 04:30 |
| TBBle | Actually, it loses the version-targetted version too. | 04:30 |
| TBBle | You could always create nicer tags for the same revisions, if the name of the tag is an issue. | 04:31 |
| TBBle | Of course, I don't know that the nicer tag would override the older tag in git-describe. | 04:32 |
| mithro | let me try this... | 04:33 |
| TBBle | Fascinating. Putting version.py in ~ and running ~/version.py makes it claim it's installed to / | 04:38 |
| mithro | ? | 04:43 |
| mithro | oh | 04:43 |
| mithro | I know why | 04:43 |
| TBBle | Yeah, the more I think about this, the less actually useful I think this versioning is. It's lost as soon as it's installed. | 04:52 |
| TBBle | Beh. And os.popen doesn't eat stderr for you. And the existence of a .git is no guarantee that git-describe exists. And you don't want to be doing that at runtime anyway. | 04:55 |
| TBBle | I think the version in version.py should prolly have been an API/bugfix version, rather than a release version. Beh. Versioning is hard. | 04:57 |
| mithro | hold on a sec | 04:58 |
| mithro | This version of this deb will be 0.2.2+0.2.3+git.74.gdfe9a4f-1auto1 | 04:59 |
| mithro | how does that look? | 04:59 |
| mithro | in the changelog it will have the output from gitdescribe | 04:59 |
| mithro | it also checks that the tp.netlib.version and git tags agree otherwise it bails | 05:00 |
| TBBle | Yeah, that works. | 05:01 |
| TBBle | Goin' to the shops, back later | 05:02 |
| TBBle | (Although that's what I said in June...) | 05:02 |
| *** pschulz01_ has joined #tp | 05:06 | |
| *** pangvisor has left #tp | 05:07 | |
| mithro | hey pschulz01 | 05:09 |
| mithro | I would recommend reading the logs | 05:09 |
| *** pschulz01 has quit IRC | 05:09 | |
| *** pschulz01_ has quit IRC | 05:09 | |
| *** pschulz01 has joined #tp | 05:10 | |
| mithro | as TBBle was about, he did the original python packages | 05:11 |
| pschulz01 | mithro: Ok.. | 05:11 |
| mithro | pschulz01: he also found "git-describe" | 05:12 |
| mithro | which gives us a continually incrementing version related to the git repository | 05:12 |
| mithro | http://www.thousandparsec.net/~irc/logs/%23tp.2007-08-11.log.html#t2007-08-11T01:48:57 | 05:13 |
| tpb | <http://ln-s.net/og1> (at www.thousandparsec.net) | 05:13 |
| pschulz01 | I was figuring that there should have been something like this done before. | 05:13 |
| mithro | pschulz01: so I've updated your "update-change-log" | 05:14 |
| mithro | and using TBBle's stuff it seems to build correct packages | 05:14 |
| pschulz01 | mithro: I'm still learning | 05:14 |
| mithro | pschulz01: that is okay | 05:15 |
| mithro | btw just to be interesting TBBle is a Paul too :P | 05:15 |
| CIA-23 | mithro libtpproto-py * r5423f44482e7 / (debian/update-debian-changelog tp/netlib/version.py): Use the "git-describe" version for the deb packages. | 05:15 |
| mithro | pschulz01: give that a look | 05:15 |
| CIA-23 | mithro libtpclient-py-development * r2796f651f567 / (3 files in 2 dirs): Use the "git-describe" version for the deb packages. | 05:24 |
| CIA-23 | mithro tpclient-pywx-development * r5bdb77859333 / (debian/update-debian-changelog version.py): Use the "git-describe" version for the deb packages. | 05:39 |
| pschulz01 | mithro: | 05:41 |
| pschulz01 | paul@flik:~/Desktop/git/tp/libtpproto-py/debian$ ./update-debian-changelog | 05:41 |
| pschulz01 | Currently in /home/paul/Desktop/git/tp/libtpproto-py/debian | 05:41 |
| pschulz01 | Traceback (most recent call last): | 05:41 |
| pschulz01 | File "/home/paul/Desktop/git/tp/libtpproto-py/debian/update-debian-changelog", line 58, in <module> | 05:41 |
| pschulz01 | tag, commits, gitcommit = gitdescribe.strip().split('-') | 05:41 |
| pschulz01 | ValueError: need more than 2 values to unpack | 05:41 |
| mithro | hrm.... | 05:41 |
| pschulz01 | Freshly cloned | 05:42 |
| mithro | can you tell me what "git-describe" outputs in your directory? | 05:42 |
| pschulz01 | LIBTPPROTO_0_2_2-g5423f44 | 05:42 |
| CIA-23 | mithro tpclient-pywx-development * rc100f4b2c467 /requirements.py: Make the version requirement match the development version. | 05:43 |
| mithro | git --version | 05:43 |
| mithro | my guess is 1.4 right? | 05:43 |
| pschulz01 | git version 1.4.4.2 | 05:43 |
| pschulz01 | bad? | 05:43 |
| mithro | yeah | 05:45 |
| mithro | that version doesn't do the right thing | 05:46 |
| pschulz01 | Hmm.. lot's of packages to upgrade. git-core gitk included. | 05:48 |
| mithro | I guess I should probably check git-version? | 05:51 |
| pschulz01 | mithro: git describe will not give sequential versions though.. will it? (Not that I can see) | 05:51 |
| pschulz01 | mithro: Incoming version - 1.5.2.3 | 05:53 |
| mithro | yes it will | 05:56 |
| pschulz01 | bbs.. Dr.Who on. | 05:56 |
| mithro | pschulz01: the new gits output | 05:57 |
| mithro | <tag>-<number of commits since tag>-<commit short sha1> | 05:57 |
| pschulz01 | Ahh.. good :-) | 05:58 |
| mithro | pschulz01: gf over, bblr | 06:02 |
| TBBle | It'd be nice to have the update-debian-changelog catch the failure to import debian_bundle and report a useful message about needing python-debian installed. | 06:03 |
| TBBle | Ah. It needs to add '.' to sys.path as well, as stuff in the debian/ directory should expect to be able to be run as debian/blah. (Ideally it'd be available as "debian/rules update-changelog" as well which then calls the script.) | 06:21 |
| pschulz01 | TBBle: Howdy.. | 06:45 |
| * JLP wakes up | 06:46 | |
| JLP | ahoy people | 06:46 |
| pschulz01 | I was thinking that update-debian-changelog was not something inside the debian build process... (eg. not in debian/rules) | 06:47 |
| pschulz01 | unless there was some way of separating 'auto' builds from regular release builds. | 06:48 |
| pschulz01 | debian_bundle? | 06:48 |
| mithro | I guess the last trick, is how to get the dependencies right, the latest deb of libtpclient-py should depend on the latest deb of libtpproto-py | 06:58 |
| mithro | etc | 06:58 |
| mithro | be back later | 06:59 |
| TBBle | No, I'm not saying you make it part of the debian build process, but you do make it runnable as a rule in debian/rules, which reduces the number of places people have to look to work out how to run stuff. | 07:37 |
| TBBle | mithro: For interdependancies, you're going to have to do it by hand, or track when relevant APIs change. Sounds like versioning-nightmare to me, frankly. | 07:38 |
| TBBle | pschulz01: When you do "from debian_bundle.changelog import Changelog, Version" wrap it in a try: ... except ImportError: print "You need to install the python-debian package"; sys.exit() | 07:40 |
| TBBle | Or similar. | 07:40 |
| *** zipola has joined #tp | 07:42 | |
| TBBle | And add a rule to debian/rules (which is a Makefile) for "update-changelog" which just runs the changelog update script. Or even better, runs the changelog update script and then runs dpkg-buildpackage -us -uc -b -rfakeroot or similar. Then of course it'd be called "build-git-package" or something. | 07:45 |
| pschulz01 | TBBle: Wouldn't it be beter in a top level Makeifle? | 08:04 |
| TBBle | No, it's a rule that relates only to the Debian packaging, so it goes in debian/rules. | 08:51 |
| TBBle | You could put a rule in the toplevel makefile to call that rule if you wanted, but I think that sort of thing pollutes top-level makefiles, personally. | 08:52 |
| *** zipola has quit IRC | 09:24 | |
| *** zipola has joined #tp | 10:16 | |
| * mithro is back | 10:17 | |
| * mithro assumes that both TBBle and pschulz01 have disappeared | 10:17 | |
| mithro | TBBle: I think for development versions it would be enough to just say the latest libtpclient-py depends on the latest libtpproto-py | 10:18 |
| mithro | the problem with that is how do you find out the latest version of the other package? | 10:20 |
| TBBle | If by latest you mean latest released, that's easy. If you want the actualy current git HEAD, that's harder. | 10:20 |
| TBBle | git-peek-remote or something might help there. | 10:20 |
| mithro | TBBle: I mean git HEAD | 10:20 |
| mithro | we need a git-describe which works on a remote repository | 10:21 |
| TBBle | git-ls-remote might be a good start. | 10:22 |
| TBBle | Then you've got to turn that into something that matches git-describe's output... | 10:24 |
| mithro | TBBle: the hard part is getting the number of commits part | 10:26 |
| *** zipola has quit IRC | 10:27 | |
| *** zipola has joined #tp | 10:31 | |
| TBBle | This might be something worth chasing up on a git-users mailing list. All the stuff must be there to query, but I suspect it's currently internal to git-fetch | 10:31 |
| mithro | TBBle: we could just cheat and look in a directory somewhere for the latest deb? | 10:39 |
| mithro | Obviously "releases" would have fixed numbers | 10:39 |
| mithro | it's not like the development versions are to end up in Debian | 10:41 |
| TBBle | Actually, I was just thinking your bot for monitoring commits could auto-generate the git-describe result of the libtpproto-py head and put it someone people building libtpclient-py can get it. | 11:21 |
| TBBle | Hmm. Bedtime for me, methinks. | 11:22 |
| mithro | TBBle: have a good night | 11:30 |
| *** JLP has quit IRC | 11:55 | |
| *** JLP has joined #tp | 11:59 | |
| *** JLP has quit IRC | 12:42 | |
| *** greywhind has joined #tp | 14:15 | |
| *** guitsaru has quit IRC | 14:51 | |
| mithro | ~seen niphree | 15:58 |
| tpb | mithro: niphree was last seen in #tp 1 day, 21 hours, 2 minutes, and 51 seconds ago: <niphree> good night all | 15:58 |
| *** pschulz01 has quit IRC | 20:13 | |
| *** pschulz01 has joined #tp | 20:15 | |
| CIA-23 | fr33.em4il tpserver-cpp-rfts * rba617173c20e /modules/games/rfts/ (planet.cpp planet.h productionorder.cpp productionorder.h): | 21:02 |
| CIA-23 | production bug fixes. | 21:02 |
| CIA-23 | production should be working. just needs a couple fixes, tweaking, etc | 21:02 |
| CIA-23 | fr33.em4il tpserver-cpp-rfts * r400f321cc073 /modules/games/rfts/ (planet.cpp productionorder.cpp rfts.cpp rftsturn.cpp): | 21:02 |
| CIA-23 | Production bug fixes. Added planet RP creation on odd turns. Production is | 21:02 |
| CIA-23 | in and working (but barely tested). Just needs some tweaks, and fixes. | 21:02 |
| xdotx | wow those updates seem vaguely similar ~~ | 21:03 |
| *** DystopicFro has joined #tp | 22:15 | |
Generated by irclog2html.py 2.5 by Marius Gedminas - find it at mg.pov.lt!