| opendevreview | Takashi Kajinami proposed openstack/codegenerator master: Replace obsolete PCRE packages https://review.opendev.org/c/openstack/codegenerator/+/973945 | 05:56 |
|---|---|---|
| frickler | cardoe: well it currently isn't mergeable due to the wrong depends-on. once that is fixed, you could specifically ping the PTL and maybe some other cores | 07:14 |
| opendevreview | Merged openstack/openstacksdk master: Add microversion to tag operation API request https://review.opendev.org/c/openstack/openstacksdk/+/973956 | 08:33 |
| PavloKostianov[m] | Hey there! I am Pavlo - new here. I was reviewing release patch (https://review.opendev.org/c/openstack/releases/+/974030) which suggest releasing "replacement of deprecated formatter from osc_utils" in python-ironicclient (https://review.opendev.org/c/openstack/python-ironicclient/+/949704). My question is: | 11:41 |
| PavloKostianov[m] | Are osc_lib.utils (as I understand this is what it is about) really deprecated? I see osc_lib.tests.utils are deprecated, but can't find anything about osc_lib_utils deprecation. | 11:41 |
| PavloKostianov[m] | * Hey there! I am Pavlo - new here. I was reviewing release patch (https://review.opendev.org/c/openstack/releases/+/974030) which suggests releasing "replacement of deprecated formatter from osc\_utils" in python-ironicclient (https://review.opendev.org/c/openstack/python-ironicclient/+/949704). My question is: | 11:42 |
| PavloKostianov[m] | Are osc\_lib.utils (as I understand this is what it is about) really deprecated? I see osc\_lib.tests.utils are deprecated, but can't find anything about osc\_lib\_utils deprecation. | 11:42 |
| stephenfin | PavloKostianov[m]: https://github.com/openstack/osc-lib/commit/0e7dada02b825b2f97ac3eb68544bdd438a3a099 | 14:59 |
| stephenfin | PavloKostianov[m]: osc_lib.utils *isn't* deprecated. Using the functions from there (or anything other than a formatter class *was* deprecated and is not removed | 15:00 |
| stephenfin | correct: *(or anything other than a formatter class) with the get_dict_properties or get_item_properties functions | 15:01 |
| PavloKostianov[m] | I see, thanks stephenfin! When you say "formatter class" you mean all the functions related to formatting? So let's say format_dict() is not deprecated, is it? No need to replace it in another dependency then? | 15:46 |
| stephenfin | PavloKostianov[m]: correct | 15:51 |
| opendevreview | cid proposed openstack/openstacksdk master: Add support for filtering by `conductor_groups` https://review.opendev.org/c/openstack/openstacksdk/+/973287 | 16:40 |
| opendevreview | Jacob Anders proposed openstack/openstacksdk master: Add node health field for Redfish health monitoring https://review.opendev.org/c/openstack/openstacksdk/+/973264 | 16:53 |
| opendevreview | Jacob Anders proposed openstack/openstacksdk master: Add node health field for Redfish health monitoring https://review.opendev.org/c/openstack/openstacksdk/+/973264 | 16:55 |
| cardoe | frickler: I've got some others as well https://review.opendev.org/c/openstack/openstacksdk/+/955103 and https://review.opendev.org/c/openstack/openstacksdk/+/973321 both of those have 2+ ironic's +2s. | 17:06 |
| cardoe | And for the permissions... https://review.opendev.org/c/openstack/project-config/+/974261 I'm not sure of the ask of me is? | 17:07 |
| frickler | cardoe: you should consider stacking those changes, else they'll have merge conflicts. also gtema and stephenfin are the sdk cores most likely to approve these | 17:10 |
| stephenfin | cardoe: approve those sdk changes | 17:13 |
| frickler | cardoe: I have updated https://review.opendev.org/admin/groups/33c03b5e000a75ffc476c51d484c5111648d399c,members now, in theory any openstack-sdk-core could have done this. with that IMO 974261 should no longer be necessary | 17:14 |
| stephenfin | frickler: infra related, but what role do the files in that gerrit/acls directory given you can modify things directly in the Gerrit UI? | 17:15 |
| frickler | stephenfin: well the acl file needs to define which groups are allowed to do something. but if e.g. ironic-reviewers is already a member of openstacksdk-service-core, it is no longer necessary for the former to be explicitly added in the sdk acl | 17:17 |
| frickler | same goes for all the other $project-core groups | 17:17 |
| cardoe | stephenfin: thank you. | 17:18 |
| cardoe | frickler: I'll stack the others and/or resolve merge conflicts. | 17:19 |
| stephenfin | frickler: ack, so does https://review.opendev.org/c/openstack/project-config/+/974347 make sense? | 17:23 |
| frickler | yes, that's what I suggested, will wait for gtema to +1 in case there was some reasoning behind this that I'm not aware of | 18:47 |
| opendevreview | Merged openstack/openstacksdk master: Add Node.instance_name support https://review.opendev.org/c/openstack/openstacksdk/+/955103 | 18:47 |
| opendevreview | Merged openstack/openstacksdk master: feat: add missing baremetal port fields https://review.opendev.org/c/openstack/openstacksdk/+/973321 | 18:47 |
| gtema | Stephenfin, frickler: I think the main reason for that setup was to have a clear separation of concerns. When all service cores are included in the sdk-core group we cannot prevent them from having also workflow permissions, which we wanted to protect. I think it is better to have Gerrit acls and not to play with group nesting | 18:52 |
| opendevreview | Doug Goldstein proposed openstack/openstacksdk master: Add support for filtering by `conductor_groups` https://review.opendev.org/c/openstack/openstacksdk/+/973287 | 18:55 |
| opendevreview | cid proposed openstack/openstacksdk master: Add support for filtering by `conductor_groups` https://review.opendev.org/c/openstack/openstacksdk/+/973287 | 18:58 |
| cardoe | cid: <spiderman meme> | 18:59 |
| cid | :D | 18:59 |
| opendevreview | Piotr Sipika proposed openstack/python-openstackclient master: Make --all-stores behave the same as in glanceclient. https://review.opendev.org/c/openstack/python-openstackclient/+/974374 | 21:34 |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!