| opendevreview | Yaguang Tang proposed openstack/keystoneauth master: fix: correct oidc request error handling logic https://review.opendev.org/c/openstack/keystoneauth/+/941203 | 04:14 |
|---|---|---|
| opendevreview | Yaguang Tang proposed openstack/keystoneauth master: fix: correct oidc request error handling logic https://review.opendev.org/c/openstack/keystoneauth/+/941203 | 05:18 |
| opendevreview | Yaguang Tang proposed openstack/keystoneauth master: fix: correct oidc request error handling logic https://review.opendev.org/c/openstack/keystoneauth/+/941203 | 05:30 |
| opendevreview | Yaguang Tang proposed openstack/keystoneauth master: fix: correct oidc request error handling logic https://review.opendev.org/c/openstack/keystoneauth/+/941203 | 05:54 |
| opendevreview | Yaguang Tang proposed openstack/keystoneauth master: fix: correct oidc request error handling logic https://review.opendev.org/c/openstack/keystoneauth/+/941203 | 06:14 |
| opendevreview | Yaguang Tang proposed openstack/keystoneauth master: fix: correct oidc request error handling logic https://review.opendev.org/c/openstack/keystoneauth/+/941203 | 06:33 |
| joelm21 | Long shot, but is there a document that covers the openstacksdk's functions, in relation to their relevant policy? i.e. `identity.get_user` has a corresponding policy, but `identity.find_user` doesn't, and I'm just trying to work out what privileges I need to grant my user to get this working. | 06:49 |
| opendevreview | Yanks proposed openstack/openstacksdk master: pass stream value https://review.opendev.org/c/openstack/openstacksdk/+/927778 | 07:08 |
| opendevreview | Yaguang Tang proposed openstack/keystoneauth master: fix: correct oidc request error handling logic https://review.opendev.org/c/openstack/keystoneauth/+/941203 | 07:26 |
| opendevreview | Yaguang Tang proposed openstack/keystoneauth master: fix: correct oidc request error handling logic https://review.opendev.org/c/openstack/keystoneauth/+/941203 | 07:41 |
| Uggla | stephenfin, gtema, could you please have a look at 880056: Add shares API to compute | https://review.opendev.org/c/openstack/openstacksdk/+/880056 and 881540: Add share commands to compute | https://review.opendev.org/c/openstack/python-openstackclient/+/881540 | 09:35 |
| stephenfin | gtema: Thoughts on merging OSC into SDK early in E? | 11:09 |
| stephenfin | ...and making OSC a meta package | 11:09 |
| stephenfin | or at least the compute, image, network and object storage parts of it since those are pure SDK now | 11:10 |
| gtema | generally - ok. What I am concerned with is that at some stage everything is being splitted, later again re-merged. It is not about sdk/cli in particular, just an observation. | 11:10 |
| gtema | are we now talking about merging into single repo while still having different pypi packages? | 11:11 |
| stephenfin | At least in the short term, to avoid breaking workflows | 11:12 |
| stephenfin | I think we can do one of two things | 11:12 |
| stephenfin | Either (a) we do a wholesale import of everything from OSC into SDK but put it behind an extra in setup.cfg, and then have OSC literally have a requirements.txt e.g. 'openstacksdk[cli]' | 11:13 |
| stephenfin | (a.k.a. similar to what the openstackclient meta-package is currently doing https://github.com/openstack/openstackclient/blob/master/requirements.txt) | 11:14 |
| stephenfin | Or (b) we only import the actual service commands (i.e. 'openstackclient.compute', 'openstackclient.network') and add a 'openstack.cli.base' entry to our 'setup.cfg' | 11:15 |
| stephenfin | I think (b) is conceptually nicer since that is using the same machinery that e.g. python-manilaclient is using to provide their OSC commands | 11:15 |
| stephenfin | On your meta point of things being split out and remerged, I think that's a consequence of (a) tooling and (b) latterly the decreasing size of the community | 11:16 |
| stephenfin | There was all this focus on various "lib" packages to share code between projects, mainly because Python doesn't historically make it easy to export multiple packages from a single repo | 11:17 |
| gtema | (b) makes sense but then we still have 2 different repos and perhaps a mess questioning yourself: where the heck is x.y.z now coming from | 11:17 |
| stephenfin | Plus Gerrit doesn't have anything like the k8s OWNER file thing: someone either has commit rights to the entire repo or they don't. You can't grant it for part of the repo. At least not natively. | 11:18 |
| gtema | i am now often confused where is the actual implementation of the identity stuff that is somehow split between multiple different places | 11:18 |
| stephenfin | The source of the CLI implementations is always shown if you do '--help' | 11:19 |
| gtema | fyi: OWNER is not a k8s specific stuff, it is suported by git hosters | 11:19 |
| gtema | I mean the code itself which is chaining pieces from different places and different repos | 11:19 |
| gtema | not finding the entry point | 11:19 |
| stephenfin | Gotcha. If we do (b) then I'm only proposing moving the services that are 100% SDK right now | 11:20 |
| stephenfin | so there should be no novaclient, no swiftclient etc. | 11:20 |
| gtema | ok, lets discuss more details during ptg | 11:20 |
| stephenfin | OSC will take the shell utility stuff from osc-lib and cliff, and it will take the actual commands from either OSC itself (identity, block storage), SDK (compute, image, network and object storage), or one of the python-*client packages (shared filesystems, clustering etc.) | 11:21 |
| gtema | we can meaybe also discuss merging osc-lib into osc itself. I do not understand why it is now a separate repo | 11:22 |
| stephenfin | neither do I, so I'm onboard with that | 11:23 |
| stephenfin | lemme draft up a PoC to demo what I'm thinking | 11:23 |
| stephenfin | only a PoC since I'd like to find a way to preserve history if possible | 11:24 |
| gtema | sure | 11:25 |
| opendevreview | Stephen Finucane proposed openstack/python-openstackclient master: Permit use of tuple API_VERSIONS https://review.opendev.org/c/openstack/python-openstackclient/+/941636 | 12:43 |
| opendevreview | Stephen Finucane proposed openstack/openstacksdk master: WIP: Add compute, image, network and object storage CLIs https://review.opendev.org/c/openstack/openstacksdk/+/941637 | 12:50 |
| opendevreview | Stephen Finucane proposed openstack/python-openstackclient master: WIP: Remove compute, image, network and object storage CLIs https://review.opendev.org/c/openstack/python-openstackclient/+/941638 | 12:52 |
| opendevreview | Stephen Finucane proposed openstack/python-openstackclient master: WIP: Remove compute, image, network and object storage CLIs https://review.opendev.org/c/openstack/python-openstackclient/+/941638 | 12:57 |
| stephenfin | gtema: ^ | 12:57 |
| stephenfin | That's going to fail so hard but that's the general idea | 12:57 |
| gtema | cool, will look later today. Thanks | 12:58 |
| opendevreview | Merged openstack/python-openstackclient master: Temporary ignore "is_vlan_qinq" column in the output https://review.opendev.org/c/openstack/python-openstackclient/+/940111 | 13:23 |
| opendevreview | Merged openstack/openstacksdk master: Add "vlan_qinq" attribute to the "network" resource https://review.opendev.org/c/openstack/openstacksdk/+/939703 | 13:23 |
| opendevreview | Merged openstack/python-openstackclient master: Add "qinq-vlan" and "no-qinq-vlan" params to the "network create" cmd https://review.opendev.org/c/openstack/python-openstackclient/+/939706 | 13:23 |
| *** tkajinam is now known as Guest9129 | 13:39 | |
| opendevreview | Rodolfo Alonso proposed openstack/python-openstackclient master: [Neutron] "uplink-status-propagation" enabled by default https://review.opendev.org/c/openstack/python-openstackclient/+/941552 | 14:13 |
| opendevreview | Rodolfo Alonso proposed openstack/python-openstackclient master: [Neutron] Support ``uplink-status-propagation-updatable`` extension https://review.opendev.org/c/openstack/python-openstackclient/+/941555 | 14:25 |
| opendevreview | Merged openstack/keystoneauth master: typing: Deprecate use of AuthConstructor https://review.opendev.org/c/openstack/keystoneauth/+/935770 | 16:29 |
| opendevreview | Merged openstack/keystoneauth master: typing: Removing additional unused kwargs (1/2) https://review.opendev.org/c/openstack/keystoneauth/+/935771 | 20:48 |
| opendevreview | Merged openstack/keystoneauth master: typing: Removing additional unused kwargs (2/2) https://review.opendev.org/c/openstack/keystoneauth/+/941242 | 21:18 |
| opendevreview | Merged openstack/keystoneauth master: pre-commit: Bump versions https://review.opendev.org/c/openstack/keystoneauth/+/941253 | 21:18 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!