Team Chat Logs
February 9, 2011
| 2011 1 | ||||||
|---|---|---|---|---|---|---|
| Mo | Tu | We | Th | Fr | Sa | Su |
| 1 | 2 | 3 | 4 | 5 | 6 | |
| 7 | 8 | 9 | 10 | 11 | 12 | 13 |
| 14 | 15 | 16 | 17 | 18 | 19 | 20 |
| 21 | 22 | 23 | 24 | 25 | 26 | 27 |
| 28 | ||||||
| << | < | > | >> | |||
Select Date
| [04:13:33.870676] | <jormon> | hey all, does anyone have any experience using bitten/build status to also support buliding/testing custom pre-commit diffs |
| [04:13:36.407957] | <jormon> | ? |
| [04:17:57.554480] | <osimons> | jormon: bitten is really not made for running arbitrary builds by developers as part of develop/commit cycle. |
| [04:19:56.060231] | <jormon> | osimons: thanks for the info... i saw this article on the wiki but it's very old. it looks like it describes a simple enough protocol that one could build a custom master that would accept diff uploads and checked in revisions to apply them to and delegate that work with custom generated build receipies for slaves. is this currently the communication protocol used between master and slave? |
| [04:19:56.455809] | <jormon> | http://bitten.edgewall.org/wiki/MasterSlaveProtocolHttp |
| [04:20:56.760225] | <osimons> | jormon: yes, but.... bitten ONLY works against actual revisions in the repository - each build is tied to a revision |
| [04:21:34.554761] | <osimons> | you could of course build and then discard results |
| [04:21:52.802112] | <jormon> | osimons: but technically i could cheat and add a custom step to the dynamically generated build recipe that would do a fetch the changes from the custom master |
| [04:22:34.169874] | <osimons> | jormon: of course, bitten no longer sends the source to slave - the recipe should contain all details of getting the source, building and reporting |
| [04:23:53.021466] | <jormon> | okay, good to know. i don't think i have the man hours to contribute right now, but i'll consider it in the future! if so, i would probably build as a simple ror app and publish on github. thanks again for the info |
| [04:24:55.670260] | <jormon> | i suppose to be consistent i should use pylons or something like that... too many languages floating around the interwebs these days |
| [04:25:33.914298] | <osimons> | jormon: with ticket:301 (using xi:include for recipes) your recipe can be in the source code + you can just run the slave locally to read the xml instead of fetching it from a master |
| [04:27:05.508258] | <osimons> | (oops that turned out confusing - building locally is always an option, but xi:include would allow you to split up responsibility) |
| [04:28:58.186420] | <jormon> | well, there are already have recipes in the source tree that are used with bitten for post-commit builds, so I would probably just recycle those and insert a custom build step into them. building locally also means that I can't take advantage of having lots of slaves to do the building for me, or i'd be reimplementing something bitten already does for me. |
| [04:29:22.155390] | <osimons> | jormon: re-reading the http protocol that is not updated - it is the page for the original proposal many moons ago... it has elements of reality, but cannot be used as-is |
| [04:29:58.519264] | <osimons> | jormon: found the wiki page for building locally; http://bitten.edgewall.org/wiki/BittenRunSlaveLocally |
| [04:30:53.180620] | <osimons> | no doubt you can adapt a variation of local build to run the checks you need before committing - and thereby triggering a full build |
| [04:31:42.601375] | <osimons> | heh. that wiki page actually uses the words "pre-commit validation" - never noticed that ;-) |
| [04:31:46.888303] | <jormon> | osimons: yeah that's really interesting, good idea. i can even make it a script that copies everything to an actually build slave we already have set up so it doesn't occupy local resouces on the developer's machine |
| [04:38:00.267479] | <jormon> | osimons: i think that turns out to be exactly what i needed... thanks again for your help - g'day |
| [04:38:16.161503] | <osimons> | no probs. good luck, jormon |
| [04:39:18.508653] | <jormon> | osimons: actually, while i have you on the horn, is there a simple way in the master http api for me to grab a list of all available build slaves? (for use as potential targets in my script?) if this script works out, i'd be happy to contribute it back |
| [04:41:06.033234] | <osimons> | uh. not that i can remember, but i'm quite sure there is an open ticket for it - others have wanted the master to list status of available slaves. "work in progress", but maybe you can find some working patch somewhere? |
| [04:42:25.385796] | <osimons> | jormon: got it. http://bitten.edgewall.org/ticket/179 |
| [04:44:00.270693] | <osimons> | a start at least - look forward to your updated patch that unites this with your needs ;-) |
| [04:44:56.073651] | <jormon> | osimons: maybe if the people i'm working for decide to prioritize this custom build script thing i can work on the ticket while i'm at it. :) thanks again for your help |
| [05:38:33.031268] | <falkb> | Hi, what happened to the release? :) There is one blocker ticket again. Though I have Py2.5 and cannot help with 2.6 |
| [05:47:30.047798] | <Hodgestar> | falkb: I fell off the releasing bandwagon. |
| [05:48:53.169002] | <Hodgestar> | Let me check the patch quickly (I wrote it but I need to test it). |
| [06:25:37.046445] | <falkb> | you must hurry because demoniac #666 is lurking and when it arrives, bitten 0.6 must have been released to prevent the gates of hell from opening... ;) |
| [06:30:01.846114] | <Hodgestar> | Hehe. |
| [06:30:58.357751] | <Hodgestar> | Annoyingly, I can't reproduce the original bug under Python 2.6.6 on Ubuntu. I suspect that the Ubuntu maintainers may patched Python. |
| [06:31:18.850846] | <Hodgestar> | I will have to investigate more tomorrow. |
| [06:32:48.745950] | <falkb> | Can't the bug reporter test your patch? |
| [23:52:02.276744] | <Hodgestar> | falkb: Ke did. |
