| @sarya009:matrix.org | Is Python 3.9 and Ubuntu 20.4 compatible with devstack Xena/Yoga? | 07:42 |
|---|---|---|
| -@gerrit:opendev.org- Simon Westphahl proposed: [zuul/zuul] 865809: Ignore commit message in change matchers https://review.opendev.org/c/zuul/zuul/+/865809 | 09:34 | |
| @jim:acmegating.com | ClarkMichael Kelly To expand on what I said last week -- the Gerrit maintainers have recommended the webhook plugin as an alternative to ssh. The implementation of that in Zuul would be much simpler than anything else due to the existing framework used by other drivers. Webhook is likely to be used by a larger number of people, and I think we should keep the number of ways of interacting with Gerrit to a minimum. All of that combined suggests that we should look first to webhook and only consider something else if it provides a significant benefit that can't be achieved by webhook. | 15:55 |
| @jim:acmegating.com | newer gerrit has broken the quickstart job. it looks to be a fairly simple fix which i should be able to make soon, assuming no one else has started on it. | 15:58 |
| @clarkb:matrix.org | corvus: the main drawback to using webhooks is the direction of tcp connection creation. | 16:13 |
| @clarkb:matrix.org | corvus: I think I saw a change pushed to address the new gerrit quickstart problem | 16:13 |
| @jim:acmegating.com | Clark: yes, though that tends to be the main reason people are interested in an alternative to ssh | 16:21 |
| @jim:acmegating.com | Clark: ah yep guillaumec has one at https://review.opendev.org/864828 | 16:21 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 865872: Add Gerrit Verified/Workflow submit reqs to tutorial https://review.opendev.org/c/zuul/zuul/+/865872 | 17:34 | |
| -@gerrit:opendev.org- Zuul merged on behalf of Guillaume Chauvel: [zuul/zuul] 864828: tutorial: Update gerrit configuration to gerrit 3.7 https://review.opendev.org/c/zuul/zuul/+/864828 | 18:45 | |
| -@gerrit:opendev.org- Zuul merged on behalf of Guillaume Chauvel: [zuul/zuul] 865632: tutorial: remove warn parameter from 'shell' ansible module https://review.opendev.org/c/zuul/zuul/+/865632 | 19:18 | |
| -@gerrit:opendev.org- Clark Boylan proposed: [zuul/zuul] 865920: Align tox install_command with tox defaults https://review.opendev.org/c/zuul/zuul/+/865920 | 23:00 | |
| @clarkb:matrix.org | Zuulians ^ is a minor cleanup that came out of me debugging some python3.10 setuptools + virtualenv behavior for openstack tempest. I don't think it is critical, but I do think it should give us slightly cleaner error reporting should zuul run into similar | 23:01 |
| @clarkb:matrix.org | A number of zuul repos have cargo culted this. I wanted to make sure this issomething we think is helpful before making changes for the rest of them. | 23:01 |
| @jim:acmegating.com | Clark: do you remember why we added that? | 23:15 |
| @clarkb:matrix.org | I do not. It has been cargo culted everywhere though | 23:17 |
| @clarkb:matrix.org | https://codesearch.opendev.org/?q=install_command&i=nope&literal=nope&files=tox.ini&excludeFiles=&repos= | 23:17 |
| @jim:acmegating.com | Clark: https://review.opendev.org/c/zuul/zuul/+/47277 is the only concern of relevance i can find | 23:23 |
| @jim:acmegating.com | it seems from your commit msg that is probably no longer a concern? | 23:23 |
| @jim:acmegating.com | (i mean i will totally buy "tox has changed in the last 9 years" ;) | 23:23 |
| @jim:acmegating.com | more or less same as https://review.opendev.org/#/q/I978b906b0ecd6c01c99e08f9d27271a1f31c7a14 (cc fungi for a blast from the past) | 23:26 |
| @clarkb:matrix.org | ya --pre doesn't seem to be in the new default command | 23:30 |
| @clarkb:matrix.org | corvus: https://tox.wiki/en/latest/config.html#conf-install_command | 23:30 |
| @jim:acmegating.com | Clark: yep https://github.com/tox-dev/tox/blob/1.6.0/tox/_config.py#L326 looks very different than https://github.com/tox-dev/tox/blob/6b76e18fcaa7c9610b642555bcb94aab1d37f2b3/src/tox/config/__init__.py#L265 | 23:51 |
| @jim:acmegating.com | that's apparently handled by the `pip_pre` setting now, which is false by default | 23:53 |
| @jim:acmegating.com | so i guess they flipped the defaults somewhere along the line | 23:53 |
| @clarkb:matrix.org | They probably realized no one wanted to test with pre release dependencies by default :) | 23:54 |
| @jim:acmegating.com | AKA break=True | 23:54 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!