*** gtristan has joined #baserock | 05:11 | |
*** franred has joined #baserock | 06:36 | |
*** ctbruce has joined #baserock | 07:25 | |
*** rdale has joined #baserock | 08:01 | |
*** toscalix has joined #baserock | 08:03 | |
rjek | Can I push to a trove over HTTP? | 08:12 |
---|---|---|
*** CTtpollard has quit IRC | 08:26 | |
*** CTtpollard has joined #baserock | 08:34 | |
pedroalvarez | Over https, I think so | 08:36 |
*** edcragg has joined #baserock | 08:37 | |
rjek | pedroalvarez: How does one authenticate? What is the URL scheme? :) | 08:37 |
*** jonathanmaw has joined #baserock | 08:37 | |
pedroalvarez | The url should be listed in the repository page in the trove | 08:38 |
rjek | Isn't that cgit? | 08:38 |
pedroalvarez | Yes | 08:39 |
rjek | Does cgit somehow talk to gitano, then? | 08:39 |
pedroalvarez | And you need to set up the https password for your user... I'm not quite sure about how to do that | 08:39 |
* pedroalvarez hopes rjek doesn't mean 'pushing changes using the Web browser' | 08:40 | |
rjek | No, I mean git push to an http remote | 08:41 |
pedroalvarez | Then, there is no need to have cgit to talk to gitano, I think | 08:42 |
rjek | How would it know that I am allowed to push there? | 08:42 |
rjek | gitano, AFAIK, has no HTTP interface at all, and the only thing that does provide one in a Trove (excluding super secret Lorry) is cgit, but gitano is responsible for all the access control? | 08:44 |
pedroalvarez | Right, I see your point | 08:44 |
pedroalvarez | That's the thing I'm not quite sure how it works | 08:45 |
pedroalvarez | so, lighttpd relies in a /home/git/htpasswd for authentication | 08:48 |
rjek | Perhaps that is generated by gitano; but does that allow pushing or is it pulling only? | 08:49 |
rjek | You'd expect gitano's ACLs to be a bit more complex than what might be expressed in an htpasswd file | 08:50 |
pedroalvarez | that should allow pushing, yes | 08:51 |
pedroalvarez | this is the relevant lighttpd configuration file: http://git.baserock.org/cgit/baserock/baserock/trove-setup.git/tree/ansible/roles/trove-setup/templates/lighttpd/git-httpd.conf | 08:53 |
rjek | Aha, right, so it invokes gitano not git to do the smart HTTP | 08:54 |
rjek | OK, that's interesting to know but sadly means it doesn't solve my problem; I need to test with non-smart :D | 08:55 |
rjek | thanks pedroalvarez | 08:55 |
pedroalvarez | hah | 08:55 |
pedroalvarez | np! | 08:55 |
*** tiagogomes has joined #baserock | 08:57 | |
richard_maw | rjek: you want to push with non-smart HTTP? | 09:06 |
rjek | Sad, isn't it | 09:09 |
*** CTtpollard has quit IRC | 09:09 | |
* richard_maw doesn't remember seeing anything in git to support that | 09:09 | |
robtaylor | its not possible | 09:15 |
rjek | I thought it was, it was just git-receive-pack> | 09:15 |
rjek | ? | 09:15 |
robtaylor | well, you could rsync unpacked objects but... belugh | 09:16 |
robtaylor | nop | 09:16 |
robtaylor | https://git-scm.com/book/en/v2/Git-Internals-Transfer-Protocols | 09:16 |
*** locallycompact has joined #baserock | 09:16 | |
*** gtristan has quit IRC | 09:18 | |
richard_maw | hmm, there _is_ https://git-scm.com/docs/git-http-push with uses webdav | 09:20 |
* richard_maw is no longer so sure you can't push over dumb http | 09:20 | |
robtaylor | ooh, interesting | 09:21 |
rjek | That was the thing yes | 09:21 |
*** CTtpollard has joined #baserock | 09:22 | |
robtaylor | https://github.com/git/git/blob/c2c5f6b1e479f2c38e0e01345350620944e3527f/Documentation/howto/setup-git-server-over-http.txt | 09:25 |
*** CTtpollard has quit IRC | 09:58 | |
*** CTtpollard has joined #baserock | 10:00 | |
*** franred has quit IRC | 10:03 | |
*** franred has joined #baserock | 10:17 | |
*** jonathanmaw has quit IRC | 10:47 | |
*** jonathanmaw has joined #baserock | 10:53 | |
paulsher1ood | can/do system-integration-commands create/move files, or do they just modify files that are already there? | 12:28 |
richard_maw | system integrations were able to create/move files last time I used them | 12:28 |
paulsher1ood | ok thanks | 12:28 |
paulsher1ood | so this means that the final files on a system may not actually be completely aligned with what's in the set of .meta files | 12:30 |
pedroalvarez | yup | 12:38 |
pedroalvarez | I'd say that is not the intended use for system-integration-commands | 12:38 |
pedroalvarez | or a bad practice | 12:38 |
*** bwh___ is now known as bwh | 12:39 | |
persia | +1 to adding/moving files in system-integration being poor practice (same for creating/moving files in ansible for first-boot stuff) | 12:41 |
paulsher1ood | noted. | 12:48 |
*** anahuelamo has quit IRC | 13:46 | |
*** anahuelamo has joined #baserock | 13:46 | |
*** edcragg has quit IRC | 14:32 | |
*** edcragg has joined #baserock | 14:34 | |
*** franred has quit IRC | 15:02 | |
*** ctbruce has quit IRC | 15:23 | |
*** faybrocklebank has quit IRC | 16:04 | |
*** tiagogomes has quit IRC | 16:28 | |
*** anahuelamo_ has joined #baserock | 16:36 | |
*** anahuelamo has quit IRC | 16:36 | |
*** edcragg has quit IRC | 16:39 | |
*** anahuelamo_ has quit IRC | 16:54 | |
*** anahuelamo has joined #baserock | 16:56 | |
*** jonathanmaw has quit IRC | 17:03 | |
*** toscalix has quit IRC | 17:14 | |
*** locallycompact has quit IRC | 17:35 | |
*** gtristan has joined #baserock | 18:20 | |
*** mario-go` has joined #baserock | 19:29 | |
*** gary_perkins has quit IRC | 19:30 | |
*** SotK has quit IRC | 19:30 | |
*** mario-goulart has quit IRC | 19:31 | |
*** rdale has quit IRC | 19:31 | |
*** gary_perkins has joined #baserock | 19:36 | |
*** SotK has joined #baserock | 19:36 | |
*** edcragg has joined #baserock | 19:58 | |
*** gtristan has quit IRC | 20:14 | |
*** mario-go` is now known as mario-goulart | 20:46 | |
*** locallycompact has joined #baserock | 20:54 | |
*** locallycompact has quit IRC | 23:04 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!