| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2017-11-20 | |||
| 15:30:13 | openstackgerrit | Jose Castro Leon proposed openstack/python-openstackclient master: Add support for endpoing filter commands https://review.openstack.org/520614 | |
| 15:39:30 | openstackgerrit | Merged openstack/python-openstacksdk master: Add subnet_id property for FloatingIP https://review.openstack.org/521421 | |
| 20:07:43 | mordred | dtroyer: I'm seeing an error in the osc-functional-devstack-tips job on an occ patch: | |
| 20:07:45 | mordred | dtroyer: http://logs.openstack.org/72/520472/1/check/osc-functional-devstack-tips/6fb2835/job-output.txt.gz#_2017-11-16_13_23_16_719034 | |
| 20:08:05 | mordred | dtroyer: I can't get the same error to happen locally with master of both occ and osc - wondering if you've seen that happen | |
| 20:48:53 | dtroyer | mordred: I do see that in the OSC functional -tips jobs, haven't sorted it out yet | |
| 20:55:00 | mordred | dtroyer: ok. I've mostly just eyeballed that changes that have happened since we added that test job and I don't see any that seem like they'd have any effect | |
| 20:56:49 | openstackgerrit | Merged openstack-infra/shade master: Complete move of baremetal machine tests https://review.openstack.org/502333 | |
| 21:10:32 | openstackgerrit | Merged openstack/os-client-config master: Remove setting of version/release from releasenotes https://review.openstack.org/520790 | |
| 22:50:54 | openstackgerrit | Merged openstack/python-openstacksdk master: Connection doc add arguments https://review.openstack.org/498705 | |
| #openstack-sdks - 2017-11-21 | |||
| 00:10:47 | openstackgerrit | Merged openstack/python-openstacksdk master: Remove setting of version/release from releasenotes https://review.openstack.org/521050 | |
| 00:10:48 | openstackgerrit | Merged openstack/python-openstacksdk master: Remove ansible functional tests for now https://review.openstack.org/520602 | |
| 05:46:02 | openstackgerrit | Reedip proposed openstack/python-openstackclient master: Add extra dhcp option to 'port create/set/unset' https://review.openstack.org/356263 | |
| 07:05:30 | openstackgerrit | Jose Castro Leon proposed openstack/python-openstackclient master: Add support for endpoing filter commands https://review.openstack.org/520614 | |
| 09:46:14 | openstackgerrit | Reedip proposed openstack/python-openstacksdk master: Add support to clear DHCP options in port https://review.openstack.org/475714 | |
| 12:02:30 | openstackgerrit | Vitalii Kulanov proposed openstack/cliff master: Add ability to sort data by columns for list commands https://review.openstack.org/521351 | |
| 12:52:19 | Dinesh_Bhor | cdent: Hi, do you have 10 min? | |
| 12:52:32 | cdent | Dinesh_Bhor: I can try, what’s up? | |
| 12:53:52 | Dinesh_Bhor | cdent: I am working on this blueprint: https://blueprints.launchpad.net/cinder/+spec/json-schema-validation where we decided to add json schema validation for cinder V3 API's. | |
| 12:54:13 | Dinesh_Bhor | cdent: we decided not to make any changes to V2. | |
| 12:55:28 | Dinesh_Bhor | cdent: The question is- is it allowed according to API-WG to change the exception message. | |
| 12:56:50 | cdent | Dinesh_Bhor: we have had that discussion before and the consensus has been that it is okay. | |
| 12:57:13 | cdent | Changing the response status code is considered a big deal, but the error message string, not so much. | |
| 12:57:53 | cdent | There are rare cases where it has been necessary for a client to parse a response code to make a choice on what to do (for example in placement there are multiple reasons a 409 can happen), but that’s not normals | |
| 12:59:01 | Dinesh_Bhor | cdent: okay. got it. Thanks. | |
| 14:09:12 | d0ugal | I have an osc plugin, I would like to run something before every command I register is executed. Is there a pattern for this? | |
| 14:09:24 | d0ugal | cliffs prepare_to_run_command looks like what I want, but I don't think I can access the app. | |
| 14:12:10 | d0ugal | I guess I could make my own base Command and Lister classes. | |
| 14:12:42 | cdent | d0ugal: I have no idea, but I didn’t want you to feel like you were talking to yourself. Hi. | |
| 14:12:51 | d0ugal | cdent: hello! | |
| 15:03:45 | openstackgerrit | Brian Haley proposed openstack/python-openstackclient master: Fix file mode on network-topology.rst https://review.openstack.org/521878 | |
| 15:47:02 | openstackgerrit | OpenStack Proposal Bot proposed openstack/openstackclient master: Updated from global requirements https://review.openstack.org/521902 | |
| 15:52:37 | openstackgerrit | OpenStack Proposal Bot proposed openstack/python-openstackclient master: Updated from global requirements https://review.openstack.org/521903 | |
| 15:58:08 | dtroyer | d0ugal: to be clear, you want to hook every command from your plugin or just the plugin commands? | |
| 16:00:57 | dtroyer | d0ugal: because for the latter you may be able to use cliff's CommandHook to do this. it has some limitations but works | |
| 16:05:03 | d0ugal | dtroyer: just the plugin commands. | |
| 16:05:35 | d0ugal | I spotted the CommandHook - looks like I need to hook into each command again. I guess duplicating that in the setup.cfg is fairly easy. | |
| 16:05:53 | d0ugal | (and perhaps more impportantly, easy to write a linter for) | |
| 16:07:10 | dtroyer | ok, for commands you already control the class for, another option is to override Comamnd.run(), that is where the hooks get called from anyway | |
| 16:07:30 | d0ugal | That is what I have done, in a base class for my plugin | |
| 16:07:49 | d0ugal | it was kinda annoying changing every command to use my new base class - but otherwise okay | |
| 16:08:34 | dtroyer | We did a similar thing in osc-lib to add deprecation and beta command support | |
| 16:08:59 | dtroyer | osc0lib/command/command.py | |
| 16:09:30 | dtroyer | that's how we ocvered the show/list command variants | |
| 16:09:51 | d0ugal | That is neater. | |
| 16:10:26 | d0ugal | I just copied and pasted a little bit :) | |
| 16:10:45 | d0ugal | Thanks! | |
| 17:29:54 | openstackgerrit | Merged openstack/python-openstackclient master: Remove hard-coded policy choices for creating a server group https://review.openstack.org/521397 | |
| 18:21:07 | tonythomas | (copied from #openstack-dev - looks like this is the place to ask) tonythomas> is there a common documentation of python-openstack (nova, glance, neutron) etc to do common CRUD on an instance ? starting from creating it to boot, delete ? | |
| 18:50:05 | openstackgerrit | Eric Fried proposed openstack/keystoneauth master: WIP: Return the endpoint_override from EndpointData https://review.openstack.org/491947 | |
| 19:54:22 | openstackgerrit | Merged openstack/python-openstackclient master: Updated from global requirements https://review.openstack.org/521903 | |
| 20:30:44 | openstackgerrit | Trygve Vea proposed openstack-infra/shade master: Implement availability_zone_hints for networks and routers. https://review.openstack.org/522004 | |
| 22:02:40 | openstackgerrit | Trygve Vea proposed openstack-infra/shade master: Implement availability_zone_hints for networks and routers. https://review.openstack.org/522004 | |
| 23:16:36 | openstackgerrit | Dean Troyer proposed openstack/osc-lib master: Use converged SDK insead of os-client-config https://review.openstack.org/518240 | |
| 23:35:12 | openstackgerrit | Dean Troyer proposed openstack/osc-lib master: Use converged SDK insead of os-client-config https://review.openstack.org/518240 | |
| #openstack-sdks - 2017-11-22 | |||
| 06:02:00 | openstackgerrit | Reedip proposed openstack/python-openstacksdk master: Add support to clear DHCP options in port https://review.openstack.org/475714 | |
| 07:11:44 | openstackgerrit | Vitalii Kulanov proposed openstack/cliff master: Add ability to sort data by columns for list commands https://review.openstack.org/521351 | |
| 10:07:10 | openstackgerrit | Reedip proposed openstack/python-openstacksdk master: Add support to clear DHCP options in port https://review.openstack.org/475714 | |
| 15:44:26 | dhellmann | dtroyer : does osc load any settings other than clouds.yaml on startup? | |
| 15:47:25 | dtroyer | dhellmann: no | |
| 15:47:40 | dhellmann | ok. I'm thinking about how to implement command aliases | |
| 15:48:10 | dhellmann | current idea is an "alias" command and then to just read a bunch of commands from a ~/.apprc file on startup in interactive mode | |
| 15:51:11 | dtroyer | just interactive? | |
| 15:54:41 | dhellmann | yeah, I was undecided on that | |
| 15:54:46 | dhellmann | I guess it could load them all the time | |
| 15:54:51 | dhellmann | that would be simpler to implement anyway | |
| 16:03:35 | dhellmann | dtroyer : do you think we need to handle multi-word alias names? like 'alias "foo bar" "blah --bletch"'? or is it enough to support 1-word aliases? | |
| 16:03:46 | dhellmann | 1-word items are going to be easier to expand | |
| 16:09:33 | dtroyer | one of the requested reasons for the aliases was to emulate old commands, ie "backup create" is now "volume backup create" and the alias would be used for compat | |
| 16:09:41 | dhellmann | ah | |
| 16:10:01 | dhellmann | is that something we would expect the user to set, then? | |
| 16:10:20 | dtroyer | that kind, probably not. so maybe we have two classes of aliases then | |
| 16:10:33 | dtroyer | one for compat, and one more shell/git-style? | |
| 16:11:17 | dtroyer | yeah, they are two different use cases | |
| 16:11:33 | dhellmann | I wonder how hard it would be to build on top of http://xon.sh | |
| 16:11:34 | dtroyer | ok, I've talked myself out of calling the multi-word thing 'alias' | |
| 16:12:00 | dhellmann | I don't want to re-build xonsh for sure | |
| 16:12:53 | dhellmann | if what we really need is a way to have those old commands optionally registered, that's different from an alias | |
| 16:13:01 | dhellmann | at least what I think of as a user-defined alias | |
| 16:13:38 | dtroyer | right, that is what I had to thing through (again) | |
| 16:14:40 | dhellmann | is there any reason not to have those old names available all the time? I guess we don't want them to show up in help, so we need to know they are dupes | |
| 16:14:57 | cdent | dtantsur: did you get/see my email about api-wg invitation? Often my mail falls into spam folders. | |
| 16:15:17 | dtantsur | cdent: hey, was it sent to me personally? with what subject? | |
| 16:15:20 | dtroyer | We basically do it by hand now, using help.SUPPRESS in the parser declarations | |
| 16:16:02 | dhellmann | what if the command manager had a method to register a legacy command mapped to the new name? | |
| 16:16:04 | cdent | dtantsur: yes, you personally, your redhat address, subject “api-wg core?” :) | |
| 16:17:15 | dhellmann | dtroyer : I'll give that some more thought; I have to go pick up a turkey from the smokehouse | |
| 16:17:30 | dtroyer | we often have more than just the commands change, options change too | |
| 16:17:50 | dtroyer | I think this pat of it is probably the lesser important of the two | |
| 16:18:33 | dtroyer | dhellmann: and thanks for the pointer to xon.sh, now I'll be distracted for the rest of the day! | |
| 16:35:57 | openstackgerrit | LIU Yulong proposed openstack/python-openstackclient master: Add floating IP qos_policy actions https://review.openstack.org/517845 | |
| 19:20:56 | dhellmann | dtroyer : there's a video of a presentation about xonsh at pycon from a year or two ago. very cool project. | |
| 19:21:10 | dhellmann | having the options change seems like a different level of aliasing | |
| 19:21:42 | dhellmann | that feels like something that's going to have to be handled on a command-by-command basis | |
| 19:35:05 | openstackgerrit | Dean Troyer proposed openstack/osc-lib master: Make -tips job non-voting https://review.openstack.org/522367 | |
| 19:40:40 | openstackgerrit | Monty Taylor proposed openstack/cliff master: DNM testing docs and releasenotes jobs https://review.openstack.org/522370 | |
| 19:40:40 | openstackgerrit | Monty Taylor proposed openstack/cliff master: Move doc requirements to doc/requirements.txt https://review.openstack.org/522369 | |
| 19:50:01 | dhellmann | dtroyer : using SUPPRESS to hide deprecated options seems like the right approach. Registering deprecated commands as multi-word mappings between the old and new version would be easy enough to add. | |
| 20:11:17 | openstackgerrit | Doug Hellmann proposed openstack/cliff master: add support for legacy command name translation https://review.openstack.org/522380 | |
| 20:11:22 | dhellmann | dtroyer : ^^ | |