| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2018-08-07 | |||
| 14:07:16 | mordred | yes - that's correct | |
| 14:07:38 | Shrews | mordred: and the set and immediately delete the seen_flag | |
| 14:07:39 | mordred | the alias_flag is the thing we need to check to see if it's already been set | |
| 14:07:53 | mordred | well, the getattr may not seem like a call into a recursion stack ... but it is | |
| 14:08:38 | mordred | so setting the flag before the getattr is a way for the check in the getattr call to know it shouldn't itself make a subsequent getattr call | |
| 14:09:29 | mordred | Shrews: http://git.openstack.org/cgit/openstack/openstacksdk/tree/openstack/object_store/v1/container.py#n44 and http://git.openstack.org/cgit/openstack/openstacksdk/tree/openstack/object_store/v1/container.py#n51 | |
| 14:09:40 | mordred | Shrews: are the example case that make this whole thing go batty | |
| 14:10:19 | mordred | count is aliased to object_count and object_count is aliased to count - if _neither_ are set, without some sort of something, the infinite recursion happens | |
| 14:10:51 | mordred | Shrews: (also, this took me literally the entire work day yesterday to wrap my head around, so if you're even partially following it on a first coffee, you're doing great) | |
| 14:12:34 | Shrews | mordred: ok, follow the alias thing now. but: setattr(instance, seen_flag, True); value = ; delattr(instance, seen_flag) | |
| 14:12:46 | Shrews | don't get the set-then-delete | |
| 14:13:42 | Shrews | oh! | |
| 14:13:47 | Shrews | ok, coffee just kicked in | |
| 14:13:52 | mordred | \o/ | |
| 14:13:53 | Shrews | mordred: duh | |
| 14:14:21 | mordred | Shrews: should we add a comment there to help future-us understand what the heck is going on? | |
| 14:14:51 | Shrews | meh, i've been away for a week so i'm just slow. up to you | |
| 14:15:15 | mordred | writing comments is hard | |
| 14:15:40 | Shrews | yah | |
| 14:16:14 | mordred | adriant: if you get a sec when your around, could you check out https://review.openstack.org/#/q/topic:resource-dict - and most importantly make sure it doesn't break you? | |
| 14:16:43 | mordred | it's one of the last 2 breaking API changes on the list before a 1.0 | |
| 14:17:52 | Shrews | mordred: it totally not apparent (until you pasted those container lines) how that test was testing recursion. maybe a comment in that test? | |
| 14:18:13 | mordred | Shrews: ++ | |
| 14:25:31 | mordred | Shrews: ok in that case, I'm adding a comment to the method too | |
| 14:25:55 | Shrews | k. sorry | |
| 14:27:04 | mordred | no - it's a good point - the chances we understand this in a months' time are pretty low | |
| 14:27:46 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Fix to_dict recursion issues with circular aliases https://review.openstack.org/589250 | |
| 14:27:46 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Rename Resource get and update to not clash with dict https://review.openstack.org/589193 | |
| 14:27:47 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Add computed attribute type and location to base resource https://review.openstack.org/589194 | |
| 14:27:47 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Remove special handling of stacks https://review.openstack.org/589195 | |
| 14:27:48 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Make resource a dict subclass usable by shade layer https://review.openstack.org/530835 | |
| 14:27:58 | mordred | Shrews: kk. updated | |
| 14:45:28 | openstackgerrit | Merged openstack/python-openstackclient master: Fix broken gate jobs https://review.openstack.org/587005 | |
| 15:17:21 | mordred | \o | |
| 15:17:23 | mordred | gah | |
| 15:17:25 | mordred | \o/ | |
| 15:17:28 | mordred | is what I meant | |
| 15:25:57 | openstackgerrit | Monty Taylor proposed openstack/openstackclient master: Update openstackclient-check-plugins to honor constraints https://review.openstack.org/589465 | |
| 17:19:42 | mordred | Shrews: yay https://review.openstack.org/#/c/589250/ is green! :) | |
| 17:20:28 | Shrews | mordred: +Ad | |
| 17:47:29 | PagliaccisCloud | anyone got a +2 to spare? https://review.openstack.org/#/c/584484/ | |
| 18:30:37 | mordred | PagliaccisCloud: I got AJaeger to leave a comment - but it's an update to a translated string which should go through the translations process. the base string seems to have already been fixed in the codebase. however, translation string updates are only proposed when there has been a change in the translations | |
| 18:44:40 | frickler | mordred: dtroyer: most of the rechecks I did seem to have passed, the others look like genuine issues to me. would be great if you could put https://review.openstack.org/516701 onto your review list | |
| 19:46:40 | openstackgerrit | Merged openstack/openstacksdk master: Fix to_dict recursion issues with circular aliases https://review.openstack.org/589250 | |
| 19:46:41 | openstackgerrit | Merged openstack/python-openstackclient master: Fix missing trailing spaces in network help messages https://review.openstack.org/588616 | |
| 20:13:56 | openstackgerrit | Monty Taylor proposed openstack/openstackclient master: Update openstackclient-check-plugins to honor constraints https://review.openstack.org/589465 | |
| 20:15:27 | mordred | dtroyer, frickler: ^^ that should fix theissues tosky was seeing with openstackclient-check-plugins on stable/queens (and the one in the stack before it, about stestr, is needed in that repo anyway) | |
| 20:16:16 | mordred | dhellmann: ^^ you might find that patch interesting as well - it is a patch thatuses the siblings logic outside of a tox context | |
| 20:55:44 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: DNM Testing pep8 line comments patch https://review.openstack.org/589635 | |
| 21:17:40 | mordred | Shrews: easy review if you have a sec: https://review.openstack.org/#/c/589212/ | |
| 22:10:42 | openstackgerrit | Merged openstack/python-openstackclient master: Add dns-domain support to Network object https://review.openstack.org/516701 | |
| 23:05:53 | openstackgerrit | Merged openstack/openstacksdk master: Stop calling get_all_types when service-type is None https://review.openstack.org/589212 | |
| 23:54:48 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Stop using the -consumer devstack jobs https://review.openstack.org/589675 | |
| 23:56:49 | mordred | Shrews: ^^ I think openstacksdk is sufficiently integrated with devstack that we shoudl do that | |
| #openstack-sdks - 2018-08-08 | |||
| 01:26:35 | adriant | mordred: will look! | |
| 01:27:21 | adriant | mordred: sorry I haven't had much time to look at anything SDK related. Been deep within billing related stuff for the last while. | |
| 09:14:50 | openstackgerrit | yanpuqing proposed openstack/python-openstackclient master: Add error message when network resources use "set" "unset" command https://review.openstack.org/579515 | |
| 11:09:20 | openstackgerrit | Yushiro FURUKAWA proposed openstack/osc-lib master: Revert "Fixes a bug with issubclass() being called on an instance" https://review.openstack.org/586617 | |
| 12:58:05 | Shrews | mordred: any idea what these failures are about? http://logs.openstack.org/75/589675/1/check/osc-functional-devstack-tips/cc9ccf7/testr_results.html.gz | |
| 13:00:11 | mordred | Shrews: I do not - osc's gate tests were broken until yesterday | |
| 13:21:31 | mordred | Shrews: got a sec for this one? https://review.openstack.org/#/c/589675/ should reduce some node churn | |
| 13:22:23 | Shrews | mordred: well, the tests I asked about are for that one | |
| 13:24:32 | mordred | ah - yeah | |
| 14:33:06 | amotoki | dtroyer: could you look at https://review.openstack.org/#/c/586617/ ? it is a revert in osc-lib which breaks neutronclient OSC plugin | |
| 14:38:37 | dmsimard | mordred: if it's any consolation, after doing even a bit of AWS I find the OpenStack APIs and Ansible modules for OpenStack MUCH better (and especially more consistent) | |
| 14:39:30 | dmsimard | The AWS APIs and modules are very inconsistent and unpredictable, it's almost like if they had been developed by entirely different companies | |
| 14:39:34 | dmsimard | so... good job ? :D | |
| 14:39:48 | dtroyer | amotoki: +2+W | |
| 14:40:10 | mordred | dmsimard: yay! thanks - that's great to hear | |
| 14:40:34 | amotoki | dtroyer: thanks. | |
| 14:43:13 | openstackgerrit | Merged openstack/osc-lib master: Remove PyPI downloads https://review.openstack.org/589369 | |
| 14:46:42 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Rename Resource get and update to not clash with dict https://review.openstack.org/589193 | |
| 14:46:42 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Add computed attribute type and location to base resource https://review.openstack.org/589194 | |
| 14:46:43 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Remove special handling of stacks https://review.openstack.org/589195 | |
| 14:46:43 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Make resource a dict subclass usable by shade layer https://review.openstack.org/530835 | |
| 17:43:42 | openstackgerrit | Merged openstack/osc-lib master: Revert "Fixes a bug with issubclass() being called on an instance" https://review.openstack.org/586617 | |
| 18:31:25 | frickler | Shrews: mordred: I'd guess those two failures are just racy tests that have unstable results, the job did pass before the recheck | |
| 18:32:39 | mordred | frickler: you mean https://review.openstack.org/#/c/589675/ ? - oh,you're right | |
| 18:32:54 | frickler | mordred: yes | |
| 18:33:00 | mordred | ++ | |
| 18:33:16 | mordred | I am pretty sure they are unrelated, since that patch doesn't touch that job at all | |
| 18:33:38 | frickler | ack | |
| 18:34:20 | frickler | dtroyer: with the dependency merged, I hope https://review.openstack.org/558824 at last is now ready for some final reviews | |
| 18:34:27 | Shrews | are we comfortable approving it then? | |
| 18:35:29 | mordred | Shrews: yeah - I think it should be fine | |
| 20:03:59 | openstackgerrit | Merged openstack/openstacksdk master: Stop using the -consumer devstack jobs https://review.openstack.org/589675 | |
| 20:15:08 | openstackgerrit | Merged openstack/python-openstackclient master: Add DNS support to floating IP commands https://review.openstack.org/558824 | |
| #openstack-sdks - 2018-08-09 | |||
| 00:09:42 | openstackgerrit | Monty Taylor proposed openstack/openstacksdk master: Make resource a dict subclass usable by shade layer https://review.openstack.org/530835 | |
| 00:09:53 | mordred | Shrews: I found new terrible things I can't wait to tell you about | |
| 00:10:21 | mordred | adriant: no worries re: time - you've been busy! I mostly want to make sure that once this stack goes green, we make sure it doesn't break you before we land it | |
| 00:11:12 | adriant | mordred: it mostly looks like stuff that shouldn't affect me, but I won't get a chance to play with it until next week at the earliest | |
| 00:11:25 | mordred | adriant: kk. no rush | |
| 00:11:30 | adriant | it's just the changing of the Resource internals to make it dict based right? | |
| 00:11:56 | mordred | adriant: yah- and renaming two of the methods - but most things shoudl be touching those Resource bits via proxy methods | |
| 00:12:18 | adriant | Yeah, none of my code goes that deep into the Resource class | |
| 00:12:27 | mordred | woot. then you should be fine | |
| 00:12:44 | adriant | Most of the stuff that I always ran into was the session stuff you fixed :) | |
| 00:14:13 | adriant | mordred: oh and https://review.openstack.org/#/c/544544/ totally forgot about that patch, but the person who proposed it never responded | |
| 00:45:56 | samueldmq | mordred: Shrews: I liked having openstack cloud and operator cloud as separate things. having that boundary made sense from user's perspective | |
| 00:48:43 | samueldmq | for instance, say I want to validate the abstraction layer against all public clouds. that means I would only consider calls for users, excluding operator's calls (which were in a separate thing) | |