| opendevreview | Merged openstack/oslotest master: Add note about requirements lower bounds https://review.opendev.org/c/openstack/oslotest/+/1013597 | 01:04 |
|---|---|---|
| *** mhen_ is now known as mhen | 02:14 | |
| opendevreview | Adam Harwell proposed openstack/tooz master: redis: Translate errors of group operations https://review.opendev.org/c/openstack/tooz/+/1007611 | 02:27 |
| opendevreview | Adam Harwell proposed openstack/tooz master: redis: Read groups without WATCH https://review.opendev.org/c/openstack/tooz/+/1007408 | 02:27 |
| opendevreview | Adam Harwell proposed openstack/tooz master: redis: Retry operations in one bounded layer https://review.opendev.org/c/openstack/tooz/+/1007940 | 02:27 |
| opendevreview | Adam Harwell proposed openstack/tooz master: redis: Recover after primary role changes https://review.opendev.org/c/openstack/tooz/+/1003880 | 02:27 |
| rm_work | tkajinam: really appreciate the merges and your reviews! I fixed that set and I can give you a list again if you want of the ones I think are review-priority for me | 03:19 |
| opendevreview | Takashi Kajinami proposed openstack/tooz master: redis: Refactor unit tests https://review.opendev.org/c/openstack/tooz/+/1013627 | 04:36 |
| opendevreview | Adam Harwell proposed openstack/tooz master: redis: Read groups without WATCH https://review.opendev.org/c/openstack/tooz/+/1007408 | 04:38 |
| opendevreview | Adam Harwell proposed openstack/tooz master: redis: Retry operations in one bounded layer https://review.opendev.org/c/openstack/tooz/+/1007940 | 04:38 |
| opendevreview | Adam Harwell proposed openstack/tooz master: redis: Recover after primary role changes https://review.opendev.org/c/openstack/tooz/+/1003880 | 04:38 |
| rm_work | Adjusted ^^ | 04:40 |
| opendevreview | Michael Still proposed openstack/futurist master: Do not import eventlet until it is needed https://review.opendev.org/c/openstack/futurist/+/1012147 | 08:14 |
| opendevreview | Michael Still proposed openstack/oslo.messaging master: Find patched threading without importing eventlet https://review.opendev.org/c/openstack/oslo.messaging/+/1012179 | 08:27 |
| opendevreview | Michael Still proposed openstack/oslo.service master: Import eventlet only for the WSGI server https://review.opendev.org/c/openstack/oslo.service/+/1012180 | 08:28 |
| opendevreview | Michael Still proposed openstack/oslo.utils master: Do not import eventlet in eventletutils until used https://review.opendev.org/c/openstack/oslo.utils/+/1012175 | 08:28 |
| opendevreview | Daniel Bengtsson proposed openstack/oslo.service master: backend: Require explicit backend selection https://review.opendev.org/c/openstack/oslo.service/+/1012683 | 10:31 |
| opendevreview | Gary T. Giesen proposed openstack/etcd3gw master: Add user/password authentication https://review.opendev.org/c/openstack/etcd3gw/+/1013860 | 12:00 |
| tkajinam | stephenfin, hi would be nice if you have time to check https://zuul.opendev.org/t/openstack/build/a94193559253433da40006a13c6629e7 | 12:31 |
| tkajinam | it looks like a mypy error caused by https://review.opendev.org/c/openstack/oslo.config/+/1012466 . just in case you have any idea about the problem. | 12:32 |
| *** M00SE7 is now known as M00SE | 12:53 | |
| *** haleyb|out is now known as haleyb | 13:03 | |
| opendevreview | Takashi Kajinami proposed openstack/oslo.config master: Explicitly accept types.ConfigType https://review.opendev.org/c/openstack/oslo.config/+/1013883 | 13:33 |
| rm_work | tkajinam: I looked a little and had my agent check and it agrees and wanted to make a CR, but then you beat me to it ^^ 😛 | 13:38 |
| rm_work | though the solution it made was a bit different | 13:38 |
| opendevreview | Takashi Kajinami proposed openstack/oslo.config master: Explicitly accept types.ConfigType https://review.opendev.org/c/openstack/oslo.config/+/1013883 | 13:43 |
| opendevreview | Takashi Kajinami proposed openstack/oslo.config master: Revert "typing: Loosen types for Opt.type" https://review.opendev.org/c/openstack/oslo.config/+/1013888 | 13:48 |
| tkajinam | I kind of feel like the revert ^^^ may be much simpler ... | 13:48 |
| tkajinam | rm_work, assuming you found that in ironic mypy check the change touches a few other places not directly related to the failure there (but they potentially cause similar problems in other places, I guess | 13:50 |
| rm_work | Yep basically | 13:56 |
| rm_work | It wanted to change it in ironic but also realized that Oslo might be better since as you say it could affect others | 13:56 |
| rm_work | Though the only actual solution it came up with was kinda meh | 13:59 |
| rm_work | I’ve not seen Claude get stumped like this in a while, it is kind of impressive 😂 | 14:02 |
| stephenfin | tkajinam: Okay, that was needed for e.g. https://review.opendev.org/c/openstack/keystoneauth/+/1012482/1/keystoneauth1/loading/opts.py#110 but it's clearly not cooked. Let's revert for now and I'll take a second spin through this | 14:04 |
| stephenfin | cardoe: jfyi ☝️ | 14:04 |
| tkajinam | ahh ok I got it. | 14:07 |
| tkajinam | there are many interesting handling in keystone repos... :-( | 14:07 |
| tkajinam | stephenfin, once CI reports its result then I'll merge the revert and create a new release. | 14:07 |
| tkajinam | stephenfin, thanks for checking it. | 14:07 |
| stephenfin | I've approved it (on the basis that it won't merge if CI fails) | 14:07 |
| stephenfin | Will ack the release change if you push it though | 14:08 |
| tkajinam | thx | 14:09 |
| stephenfin | fwiw, this is a very intersting issue 😅 I don't really understand why that's failing. Once to figure out later | 14:13 |
| stephenfin | sounds like it stumped Claude too though (CC rm_work), so at least it's not just me 😅 | 14:14 |
| tkajinam | yeah. the error indicates that mypy can't expand the return type properly but idk why it can't and how we can more educate it better in a simple way | 14:22 |
| opendevreview | Daniel Bengtsson proposed openstack/oslo.service master: backend: Require explicit backend selection https://review.opendev.org/c/openstack/oslo.service/+/1012683 | 14:23 |
| tkajinam | maybe we eventually need the union approach to accept both ConfigType and Callable (though it looks just redundant ...) | 14:23 |
| *** haleyb is now known as haleyb|out | 14:39 | |
| opendevreview | Merged openstack/tooz master: Drop leftover of typing-extensions https://review.opendev.org/c/openstack/tooz/+/1006302 | 17:03 |
| opendevreview | Merged openstack/tooz master: trivial: Use positive check https://review.opendev.org/c/openstack/tooz/+/1000829 | 18:37 |
| cardoe | ironic/conf/mdns.py:31: error: Argument "type" to "Opt" has incompatible type "oslo_config.types.Dict[str]"; expected "Callable[..., dict[Never, Never] | None] | None" [arg-type] | 19:06 |
| cardoe | ironic/conf/pxe.py:209: error: Argument "type" to "Opt" has incompatible type "oslo_config.types.Dict[str]"; expected "Callable[..., dict[Never, Never] | None] | None" [arg-type] | 19:07 |
| cardoe | is now expected? | 19:07 |
| frickler | cardoe: iiuc https://review.opendev.org/c/openstack/oslo.config/+/1013888 should fix this? cf. the discussion earlier today | 19:38 |
| frickler | we could also revert https://review.opendev.org/c/openstack/requirements/+/1012723 in the meantime if needed | 19:39 |
| cardoe | yeah I was doing too much context switching that makes sense | 20:00 |
| opendevreview | Merged openstack/oslo.config master: Revert "typing: Loosen types for Opt.type" https://review.opendev.org/c/openstack/oslo.config/+/1013888 | 20:03 |
| *** elodilles is now known as elodilles_pto | 20:24 | |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!