| *** dims__ has joined #openstack-oslo | 00:07 | |
| *** crc32 has quit IRC | 00:23 | |
| harlowja | jogo getting closer i think for https://github.com/harlowja/pippin (simple run @ http://paste.ubuntu.com/10165874/ ) it has the deep dependency graph, and is using some of taskflow's tree code to find paths through that tree (and hopefully just now need to find a valid path among those paths) | 00:31 |
|---|---|---|
| *** dims__ has quit IRC | 00:31 | |
| jogo | harlowja: neat, how long does that take to run? and can you run it against global-reqs? | 00:49 |
| harlowja | jogo running right now, lol | 00:49 |
| jogo | harlowja: sweet | 00:49 |
| harlowja | it is doing a deep download/open... repeat, so my guess not so quick, but it caches all the things it can | 00:49 |
| harlowja | but i'll time it | 00:49 |
| jogo | harlowja: can you make it use standard pip.conf settings like specifying the URL | 00:50 |
| harlowja | probably | 00:50 |
| jogo | so I can use devpi-server locally etc. | 00:50 |
| harlowja | ya, shouldn't be a problem i think | 00:50 |
| *** dims__ has joined #openstack-oslo | 00:53 | |
| *** dims_ has joined #openstack-oslo | 00:54 | |
| *** dims__ has quit IRC | 00:58 | |
| *** achanda_ has joined #openstack-oslo | 01:16 | |
| *** tsekiyam_ has joined #openstack-oslo | 01:17 | |
| *** achanda has quit IRC | 01:19 | |
| *** mtanino has quit IRC | 01:20 | |
| *** tsekiyama has quit IRC | 01:20 | |
| *** achanda_ has quit IRC | 01:21 | |
| *** tsekiyam_ has quit IRC | 01:21 | |
| *** noelbk has quit IRC | 01:21 | |
| *** hemna is now known as hemnafk | 01:22 | |
| *** daniel3_ has quit IRC | 01:24 | |
| *** david-lyle has quit IRC | 01:26 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 01:47 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 01:51 | |
| *** david-lyle has joined #openstack-oslo | 02:08 | |
| *** shardy has quit IRC | 02:14 | |
| *** tedross has joined #openstack-oslo | 02:29 | |
| openstackgerrit | Ian Wienand proposed openstack/oslo.policy: Turn off missing-directory log output for default case https://review.openstack.org/154742 | 02:40 |
| *** yamahata has quit IRC | 02:45 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 02:46 | |
| *** david-lyle has quit IRC | 02:46 | |
| *** harlowja is now known as harlowja_away | 02:46 | |
| *** david-lyle has joined #openstack-oslo | 02:50 | |
| *** pradk has quit IRC | 03:10 | |
| *** david-lyle has quit IRC | 03:31 | |
| *** rushiagr_away is now known as rushiagr | 03:39 | |
| *** achanda has joined #openstack-oslo | 03:40 | |
| *** _amrith_ is now known as amrith | 03:47 | |
| *** tedross has quit IRC | 03:50 | |
| *** dims_ has quit IRC | 04:01 | |
| *** rushiagr is now known as rushiagr_away | 04:04 | |
| *** ViswaV has quit IRC | 04:07 | |
| *** david-lyle has joined #openstack-oslo | 04:09 | |
| *** ViswaV has joined #openstack-oslo | 04:09 | |
| openstackgerrit | Steve Martinelli proposed openstack/oslo.policy: Use standard logging in oslo.policy https://review.openstack.org/154635 | 04:11 |
| *** achanda has quit IRC | 04:13 | |
| openstackgerrit | Steve Martinelli proposed openstack/oslo.policy: document the migration process and update the docs a bit https://review.openstack.org/154752 | 04:50 |
| *** rushiagr_away is now known as rushiagr | 04:52 | |
| *** amrith is now known as _amrith_ | 04:56 | |
| *** zzzeek has quit IRC | 04:59 | |
| *** dims__ has joined #openstack-oslo | 05:02 | |
| *** dims__ has quit IRC | 05:07 | |
| *** achanda has joined #openstack-oslo | 05:23 | |
| openstackgerrit | Min Pae proposed openstack/taskflow: Map optional arguments as well as required arguments https://review.openstack.org/151972 | 05:27 |
| *** achanda has quit IRC | 05:39 | |
| *** achanda has joined #openstack-oslo | 05:40 | |
| *** achanda has quit IRC | 05:56 | |
| *** MasterPiece has joined #openstack-oslo | 06:05 | |
| *** MasterPiece| has joined #openstack-oslo | 06:16 | |
| *** MasterPiece has quit IRC | 06:17 | |
| *** noelbk has joined #openstack-oslo | 06:21 | |
| *** inc0 has joined #openstack-oslo | 06:23 | |
| openstackgerrit | Min Pae proposed openstack/taskflow: Map optional arguments as well as required arguments https://review.openstack.org/151972 | 06:32 |
| openstackgerrit | Min Pae proposed openstack/oslo.utils: Add a "optional" argument to reflection.get_callable_args https://review.openstack.org/152357 | 06:35 |
| openstackgerrit | Merged openstack/debtcollector: Add universal wheel tag to setup.cfg https://review.openstack.org/153414 | 06:42 |
| *** achanda has joined #openstack-oslo | 06:43 | |
| *** sirushti has left #openstack-oslo | 06:52 | |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Let the multi-lock convert the provided value to a tuple https://review.openstack.org/154770 | 06:57 |
| *** vigneshvar has joined #openstack-oslo | 07:01 | |
| openstackgerrit | Joshua Harlow proposed openstack/oslo.utils: Add a stopwatch + split for duration(s) https://review.openstack.org/154311 | 07:01 |
| openstackgerrit | Vilobh Meshram proposed openstack/taskflow: Docstrings should document parameters return values https://review.openstack.org/152375 | 07:03 |
| *** achanda has quit IRC | 07:16 | |
| openstackgerrit | Min Pae proposed openstack/oslo.utils: Add a "optional" argument to reflection.get_callable_args https://review.openstack.org/152357 | 07:19 |
| *** achanda has joined #openstack-oslo | 07:29 | |
| *** shardy has joined #openstack-oslo | 07:58 | |
| *** dtantsur|afk is now known as dtantsur | 07:59 | |
| *** e0ne has joined #openstack-oslo | 08:10 | |
| *** alexpilotti has quit IRC | 08:24 | |
| *** dulek has joined #openstack-oslo | 08:27 | |
| *** achanda has quit IRC | 08:36 | |
| *** achanda has joined #openstack-oslo | 08:36 | |
| *** noelbk has quit IRC | 08:37 | |
| *** achanda has quit IRC | 08:37 | |
| *** andreykurilin_ has joined #openstack-oslo | 08:37 | |
| *** ajo has joined #openstack-oslo | 08:39 | |
| openstackgerrit | Merged openstack/taskflow: Map optional arguments as well as required arguments https://review.openstack.org/151972 | 08:39 |
| *** e0ne has quit IRC | 08:50 | |
| *** andreykurilin_ has quit IRC | 08:53 | |
| *** andreykurilin_ has joined #openstack-oslo | 08:53 | |
| *** andreykurilin_ has quit IRC | 08:54 | |
| *** andreykurilin_ has joined #openstack-oslo | 08:55 | |
| *** ihrachyshka has joined #openstack-oslo | 08:58 | |
| *** vigneshvar has quit IRC | 09:07 | |
| *** vigneshvar has joined #openstack-oslo | 09:12 | |
| *** ihrachyshka has quit IRC | 09:15 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/oslo.versionedobjects: Remove references to Nova https://review.openstack.org/154401 | 09:29 |
| *** achanda has joined #openstack-oslo | 09:38 | |
| *** achanda has quit IRC | 09:43 | |
| *** e0ne has joined #openstack-oslo | 09:45 | |
| *** stevemar has quit IRC | 09:47 | |
| *** viktors has left #openstack-oslo | 09:52 | |
| *** pblaho has joined #openstack-oslo | 09:56 | |
| *** dtantsur is now known as dtantsur|bbl | 09:56 | |
| openstackgerrit | Grzegorz Grasza proposed openstack/oslo.versionedobjects: Compatibility with Python 3 https://review.openstack.org/154407 | 10:07 |
| openstackgerrit | Merged openstack/oslo.db: Updated from global requirements https://review.openstack.org/150502 | 10:08 |
| *** alexpilotti has joined #openstack-oslo | 10:22 | |
| inc0 | guys, what should I use? from oslo import messaging or import oslo_messaging? | 10:24 |
| *** MasterPiece| has quit IRC | 10:34 | |
| openstackgerrit | Grzegorz Grasza proposed openstack/oslo.versionedobjects: Compatibility with Python 3 https://review.openstack.org/154407 | 10:34 |
| openstackgerrit | Grzegorz Grasza proposed openstack/oslo.versionedobjects: Compatibility with Python 3 https://review.openstack.org/154407 | 10:36 |
| ozamiatin | inc0, hi, if you would like to use the public api: get_rpc_server, get_notification_listener etc... you should 'from oslo import messaging'. oslo_messaging is for internal entities like drivers etc, you shouldn't import it from the outside | 10:37 |
| haypo | inc0: import oslo_messaging | 10:37 |
| *** dims__ has joined #openstack-oslo | 10:37 | |
| haypo | inc0: "oslo" namespace will slowly disappear | 10:37 |
| haypo | inc0: make sure that you have at least oslo.messaging>=1.2.0 in your requirements | 10:37 |
| inc0 | haypo, thanks for clarification | 10:38 |
| ozamiatin | haypo, thanks too, seems I didn't get it right... | 10:38 |
| inc0 | thanka for clarification guys | 10:39 |
| inc0 | thanks | 10:39 |
| haypo | ozamiatin: are you Grzegorz Grasza ? | 10:39 |
| haypo | of the versionedobjects review? | 10:40 |
| ozamiatin | haypo, no I'm Oleksii Zamiatin :) | 10:40 |
| haypo | oh ok :) sorry | 10:40 |
| ozamiatin | haypo, np ) | 10:40 |
| haypo | ozamiatin: i know, it's hard to follow recent changes in oslo | 10:40 |
| inc0 | haypo, xek is Grzegorz | 10:43 |
| inc0 | haypo, but I'm very much involved as well:) | 10:44 |
| *** ihrachyshka has joined #openstack-oslo | 10:44 | |
| haypo | xek: hey. for python3, you may take a look at https://wiki.openstack.org/wiki/Python3 thanks for helping to port openstack to python 3 ;) | 10:44 |
| *** i159 has joined #openstack-oslo | 10:45 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/oslo.versionedobjects: New field type - Json https://review.openstack.org/154820 | 10:45 |
| inc0 | speaking of which... | 10:46 |
| *** e0ne is now known as e0ne_ | 10:46 | |
| *** e0ne_ is now known as e0ne | 10:47 | |
| haypo | inc0: i don't understand what is this coerce() method :-/ | 10:48 |
| haypo | inc0: but i don't know oslo.versionedobjects at all :-p | 10:48 |
| inc0 | haypo, its part of idea that versionedobject should be serializable | 10:49 |
| inc0 | so you can transport it via RPC | 10:49 |
| inc0 | and then we have typed field, like these: https://review.openstack.org/#/c/146841/11/heat/objects/stack.py | 10:50 |
| inc0 | coerce is method which turns serialized thing to unserialized | 10:51 |
| inc0 | haypo, and I don't blame you about not knowing it, gerrit appeared in Monday;) | 10:52 |
| openstackgerrit | Grzegorz Grasza proposed openstack/oslo.versionedobjects: Rename utils -> _utils https://review.openstack.org/154409 | 10:52 |
| xek | haypo, thanks, I will look into it | 10:55 |
| *** alexpilotti has quit IRC | 11:00 | |
| *** dtantsur|bbl is now known as dtantsur | 11:01 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/oslo.versionedobjects: Remove references to Nova https://review.openstack.org/154401 | 11:03 |
| *** e0ne_ has joined #openstack-oslo | 11:09 | |
| *** i159 has quit IRC | 11:10 | |
| *** e0ne has quit IRC | 11:10 | |
| *** i159 has joined #openstack-oslo | 11:10 | |
| *** f13o has quit IRC | 11:19 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/oslo.versionedobjects: New field type - Json https://review.openstack.org/154820 | 11:29 |
| inc0 | is Victor Stinner here?:) | 11:39 |
| *** achanda has joined #openstack-oslo | 11:40 | |
| *** f13o has joined #openstack-oslo | 11:43 | |
| *** achanda has quit IRC | 11:45 | |
| *** cdent has joined #openstack-oslo | 11:48 | |
| *** boris-42 has quit IRC | 11:52 | |
| *** ankit_ag has joined #openstack-oslo | 11:52 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/oslo.versionedobjects: Fixes for heat implementation https://review.openstack.org/154835 | 11:53 |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/oslo.versionedobjects: Fixes for heat implementation https://review.openstack.org/154835 | 11:53 |
| *** _amrith_ is now known as amrith | 12:04 | |
| openstackgerrit | Alexander Makarov proposed openstack/oslo-specs: oslo logging needs a way to be reconfigured https://review.openstack.org/154838 | 12:14 |
| *** boris-42 has joined #openstack-oslo | 12:28 | |
| *** pblaho_ has joined #openstack-oslo | 12:29 | |
| *** pblaho has quit IRC | 12:29 | |
| haypo | inc0: i'm victor stinner | 12:30 |
| inc0 | ahh, good to know;) | 12:33 |
| inc0 | I've replied to your comments in review, thanks | 12:33 |
| haypo | inc0: ok, i will see that later | 12:33 |
| inc0 | basically versionedobjects is pulled straight out of nova | 12:33 |
| inc0 | hence all this nova-reference removals | 12:34 |
| inc0 | we've pulled out whole thing including all its flaws, like hardcoded stuff:) | 12:34 |
| haypo | yes, i understood that (that it comes from nova) | 12:34 |
| haypo | inc0: it is possible to implement new fields in applications using versionedobjects ? | 12:35 |
| inc0 | yes | 12:35 |
| inc0 | all you need to do is to inherit from base classes | 12:35 |
| inc0 | so nova, neutron and so on will be able to implement their own field types, I think thats reasonable | 12:36 |
| inc0 | I'd hate to see patchset for review made by Cinder guys with something like "ISCSITargetField" ;) I wouldn't even know how to review it;) | 12:37 |
| *** pblaho_ is now known as pblaho | 12:37 | |
| dims__ | go flaper87 go! | 12:50 |
| *** pblaho_ has joined #openstack-oslo | 12:51 | |
| *** pblaho has quit IRC | 12:51 | |
| *** rushiagr is now known as rushiagr_away | 12:59 | |
| *** akhivin has quit IRC | 13:00 | |
| *** mkoderer_cloud has joined #openstack-oslo | 13:03 | |
| *** khivin has joined #openstack-oslo | 13:06 | |
| *** vipul has quit IRC | 13:08 | |
| openstackgerrit | Mehdi Abaakouk proposed openstack/oslo.messaging: Skip functional tests that fail due to a qpidd bug https://review.openstack.org/154638 | 13:10 |
| openstackgerrit | Mehdi Abaakouk proposed openstack/oslo.messaging: amqp1: fix functional tests deps https://review.openstack.org/153754 | 13:10 |
| *** vipul has joined #openstack-oslo | 13:11 | |
| *** pblaho_ is now known as pblaho | 13:14 | |
| *** tedross has joined #openstack-oslo | 13:15 | |
| openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/oslo.versionedobjects: Fixes for heat implementation https://review.openstack.org/154835 | 13:18 |
| *** inc0 has quit IRC | 13:18 | |
| *** dims__ has quit IRC | 13:19 | |
| *** tedross has left #openstack-oslo | 13:19 | |
| *** dims__ has joined #openstack-oslo | 13:20 | |
| openstackgerrit | Mehdi Abaakouk proposed openstack/oslo.messaging: amqp1: fix functional tests deps https://review.openstack.org/153754 | 13:25 |
| *** e0ne_ has quit IRC | 13:40 | |
| *** amrith is now known as _amrith_ | 13:41 | |
| *** kgiusti has joined #openstack-oslo | 13:44 | |
| *** e0ne has joined #openstack-oslo | 13:45 | |
| openstackgerrit | Mehdi Abaakouk proposed openstack/oslo.messaging: amqp1: fix functional tests deps https://review.openstack.org/153754 | 13:46 |
| *** vigneshvar has quit IRC | 13:48 | |
| *** bknudson has joined #openstack-oslo | 13:49 | |
| *** jaosorior has joined #openstack-oslo | 13:50 | |
| openstackgerrit | Grzegorz Grasza proposed openstack/oslo.versionedobjects: Compatibility with Python 3 https://review.openstack.org/154407 | 13:56 |
| openstackgerrit | Grzegorz Grasza proposed openstack/oslo.versionedobjects: Compatibility with Python 3 https://review.openstack.org/154407 | 13:56 |
| openstackgerrit | Grzegorz Grasza proposed openstack/oslo.versionedobjects: Rename utils -> _utils https://review.openstack.org/154409 | 13:58 |
| *** inc0 has joined #openstack-oslo | 13:59 | |
| *** mriedem has joined #openstack-oslo | 14:29 | |
| *** zzzeek has joined #openstack-oslo | 14:34 | |
| *** gordc has joined #openstack-oslo | 14:36 | |
| *** mtanino has joined #openstack-oslo | 14:49 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 14:55 | |
| openstackgerrit | Merged openstack/oslo.messaging: Skip functional tests that fail due to a qpidd bug https://review.openstack.org/154638 | 14:57 |
| *** jecarey has joined #openstack-oslo | 15:11 | |
| *** e0ne is now known as e0ne_ | 15:12 | |
| inc0 | dansmith, hi, I'm thinking about this ip address fields in oslo | 15:15 |
| dansmith | inc0: hi | 15:15 |
| inc0 | I'm not convinced I did good removing it, but it still seems a bit too domain-specific | 15:15 |
| inc0 | most of projects will not use this kind of fields | 15:16 |
| dansmith | which projects don't deal with ip addresses? | 15:17 |
| inc0 | cinder, heat | 15:17 |
| inc0 | not on object level | 15:17 |
| dansmith | you mean they don't currently persist any plain IP addresses, I assume, but they do deal with network addresses in many forms and could at some point decide to persist an IP, right? | 15:18 |
| inc0 | sure, but if we want to add normal ipv4 address, we may as well add subnet, then ipv6 and so on and so on | 15:19 |
| inc0 | we'll end up having ton of fields around network | 15:19 |
| inc0 | it might be good idea to just separate it to another file...like network-related fields | 15:19 |
| dansmith | the reason to have them there, IMHO, is to offer an easier alternative than the other projects just using a string for those things because "it's harder to build a proper field" | 15:19 |
| dansmith | inc0: if you think separating them is important, then sure :) | 15:20 |
| *** prad has joined #openstack-oslo | 15:20 | |
| inc0 | yeah, but as I understand idea of a field, its closer to raw data structure than complex one | 15:20 |
| dansmith | it's really just to codify a specific type of data | 15:20 |
| dansmith | and there are more complicated ones than IPAddress | 15:20 |
| inc0 | yeah, another topic I wanted to move;) | 15:21 |
| dansmith | the address one does a lot of validation and canonicalization, especially in the case of ipv6 | 15:21 |
| inc0 | whats point of "list of dicts field" and so on? | 15:21 |
| inc0 | that doesn't seem right... | 15:21 |
| dansmith | one of the reasons we added the address one is because mysql stores it as a string and does no checking, postgres does checking, and our API was depending on the database to validate (!) | 15:21 |
| dansmith | the list-of-dicts (et al) are mostly because we haven't been able to force people to properly model all their data all in one go | 15:22 |
| dansmith | so I'm happy to have those stay in nova if you want | 15:22 |
| inc0 | wouldn't be better to do something like string_field(validator=ipaddress_calidator) | 15:22 |
| dansmith | we have things like lists of cpu maps and such that fit into those | 15:22 |
| *** e0ne_ has quit IRC | 15:22 | |
| inc0 | I was thinking of jsonfield... | 15:23 |
| inc0 | I had to implement it for heat (we store templates as deeply nested json..impossible to cast to "list of whatever") | 15:23 |
| dansmith | it just seems cleaner to me to define the field in terms of what it is, and not in terms of the primitive type plus a validator | 15:23 |
| inc0 | de gustibus non est disputandum | 15:25 |
| inc0 | :) | 15:25 |
| dansmith | translation please :) | 15:25 |
| inc0 | There’s no arguing about tastes and colors. | 15:25 |
| dansmith | hah | 15:25 |
| inc0 | its from latin | 15:26 |
| inc0 | I just want to agree on one standard now, before anyone actually implement it;) | 15:26 |
| inc0 | because then we might not have that chance | 15:26 |
| inc0 | ;) | 15:26 |
| *** andreykurilin_ has quit IRC | 15:26 | |
| inc0 | and what do you think about jsonfield replacing list-of-whatever-field? | 15:27 |
| openstackgerrit | Merged openstack/oslo.messaging: Add missing unit test for a recent commit https://review.openstack.org/153122 | 15:28 |
| dansmith | inc0: we specifically don't want a jsonfield in nova | 15:29 |
| inc0 | ow? and why is that? | 15:30 |
| dansmith | inc0: it is the definition of what we were trying to replace with this whole infrastructure | 15:30 |
| dansmith | inc0: it's unversioned, and that's the problem | 15:30 |
| *** e0ne has joined #openstack-oslo | 15:30 | |
| dansmith | inc0: before this, we were passing unversioned dicts (jsonified) all over RPC. The point of this is to define a schema attached to a version, and thus it should be a set of well-defined data types | 15:30 |
| dansmith | so I'm definitely -2 on putting that in here, but heat can do what they want of course :) | 15:31 |
| inc0 | well, we can't really dodge that bullet I think, I'll try to think of way around that | 15:32 |
| inc0 | dansmith, nova doesn't use sqlalchemy's jsonfield? | 15:35 |
| dansmith | no | 15:35 |
| dansmith | we store json in the database, but we're moving to using the serialized (and thus versioned) form of these objects everywhere we do | 15:35 |
| inc0 | but what about objects without strictly defined schema? | 15:36 |
| dansmith | so that when we pull json out of the database, we know what version it was stored in, and we can apply the same semantics to reading it as we do to receiving an older object over rpc | 15:36 |
| inc0 | like heat template - it can be nested indefinetly and so on? | 15:36 |
| dansmith | inc0: we don't have anything like that, AFAIK | 15:36 |
| dansmith | inc0: the closes thing we have is metadata, and it's strictly key=value, 255 chars each | 15:37 |
| *** e0ne is now known as e0ne_ | 15:37 | |
| *** e0ne_ is now known as e0ne | 15:37 | |
| inc0 | well, then I'd ask you to remove -2 on this idea, because there are cases like that | 15:38 |
| openstackgerrit | Mehdi Abaakouk proposed openstack/tooz: file: Fix multiple call of get_lock https://review.openstack.org/154913 | 15:38 |
| inc0 | heat is an example | 15:39 |
| *** amotoki has joined #openstack-oslo | 15:40 | |
| *** yamahata has joined #openstack-oslo | 15:40 | |
| dansmith | inc0: right, so heat can define their own such field, right? | 15:40 |
| dansmith | inc0: my point is, it goes contrary to the whole goal of this, so it seems extremely problematic to build in such functionality to the base library object | 15:41 |
| inc0 | well, sure we can do that in heat, but I bet there are other cases like that | 15:42 |
| dansmith | well, I'd argue we should start without it, and see how many cases pop up | 15:43 |
| dansmith | heat seems to be the only one I can think of, | 15:43 |
| dansmith | where you do have arbitrary complexity in a structure.. can't see what in cinder would be similar | 15:43 |
| inc0 | taskflow? | 15:44 |
| dansmith | I don't know enough about it to say | 15:45 |
| dansmith | so let me ask you this: | 15:45 |
| *** _gryf has joined #openstack-oslo | 15:46 | |
| dansmith | if you have half your deployment running old code and half running new code, how do you handle passing this unversioned structure from one to the other and handling potential changes in what those components expect? | 15:46 |
| dansmith | I don't know if taskflow has such a potential.. heat either for that matter, although I expect it does | 15:46 |
| dims__ | inc0: does not look like harlowja_away needs it in taskflow (http://docs.openstack.org/developer/taskflow/persistence.html) | 15:46 |
| dansmith | persistence is the other side, where you've stuffed something in the db, to be read by a later version of the code at some undetermined point in the future | 15:47 |
| inc0 | well, what I'm saying is, people will use stringfield to do it | 15:48 |
| inc0 | so exactly what you wanted to avoid by putting ipv4 field out there | 15:48 |
| dansmith | but ipv4 is a well-defined thing | 15:49 |
| dansmith | "random complexity json" is not | 15:49 |
| dansmith | unless you have some weird five-byte IPv4 addresses that we don't :P | 15:49 |
| inc0 | I know, but still do we really want close this road because people might use it wrong? | 15:49 |
| *** hemnafk is now known as hemna | 15:50 | |
| inc0 | I know, we can make our own jsonfield, we can even use stringfield...its just seem wrong | 15:50 |
| dansmith | we're not closing the road if we don't add it now, we are opening the road forever if we do, right? | 15:50 |
| inc0 | what about jsonschema or something like that? | 15:51 |
| dansmith | ...more words | 15:51 |
| *** _amrith_ is now known as amrith | 15:52 | |
| *** denis_makogon has joined #openstack-oslo | 15:52 | |
| dansmith | inc0: sorry, I meant: can you describe more of what you mean by "jsonschema or something like that" ? | 15:55 |
| inc0 | well, its something that allow us to put some arbitrary schema for json | 15:55 |
| inc0 | its more flexible than what we have now, and might just be common ground | 15:56 |
| inc0 | because we can use it for some version compatibility logic | 15:56 |
| dansmith | if you define a field in terms of jsonschema, | 15:56 |
| dansmith | and then you have version 1.0 which pertains to some specific schema of that json field, | 15:57 |
| inc0 | well, then you can schema per field, which is reasonable | 15:57 |
| dansmith | and then you want to change the schema and bump to 1.1, | 15:57 |
| dansmith | what happens if you receive an object of version 1.0? and what happens if you need to send a 1.1 object to something that only knows 1.0? | 15:57 |
| inc0 | then you'll have to add make_obj_compatible logic to turn one into another | 15:57 |
| inc0 | this is the case of anything really | 15:58 |
| inc0 | unless its just "string or none" | 15:58 |
| *** harlowja_at_home has joined #openstack-oslo | 15:58 | |
| *** yamahata has quit IRC | 15:58 | |
| dansmith | it is, but my argument here is: we've been getting it wrong for a long time, despite best intentions | 15:59 |
| *** yamahata has joined #openstack-oslo | 15:59 | |
| dansmith | it's much easier to define "version 1.3 gained a string field that is just a name of a widget. In order to send to 1.2, remove that field" | 15:59 |
| inc0 | dansmith, I'm not saying that we should replace whole thing with jsonschema, I'm just saying that if you consider ListOfSetsOfIntegersField you still can end up with function that has to add None to this at some point | 16:00 |
| inc0 | or floatr | 16:01 |
| inc0 | and you have exactly the same problem | 16:01 |
| dansmith | I understand what you're saying, I just don't agree :) | 16:01 |
| *** arek has joined #openstack-oslo | 16:02 | |
| inc0 | what you really did by adding class like that, is specyfing schema of something | 16:02 |
| *** arek has quit IRC | 16:02 | |
| *** rushiagr_away is now known as rushiagr | 16:03 | |
| inc0 | but still it doesn't have any migrations and anything | 16:03 |
| *** arek_ch has joined #openstack-oslo | 16:03 | |
| *** prad has quit IRC | 16:03 | |
| inc0 | and you can still have exactly the same problem you've described earlier | 16:03 |
| dansmith | you really don't. It's not a list of sets of strings, it's a list of sets of integers | 16:03 |
| dansmith | it can't change, it's a known format | 16:04 |
| dansmith | it can't include a key, inside of which could be arbitrary complexity of other stuff | 16:04 |
| inc0 | yeah, but what if between versions it will have to change? | 16:04 |
| inc0 | you'll end up with problem where you'll have to translate v1.1 to v1.0 | 16:05 |
| inc0 | and mutually incompatible formats | 16:05 |
| dansmith | perhaps this is unspoken and needs to be written, but we don't change formats of fields for this specific reason | 16:05 |
| dansmith | you can't turn Foo.cpus from a string to a list-of-sets-of-integers and then to a dict | 16:06 |
| dansmith | because that breaks everything | 16:06 |
| inc0 | but what if we could at some point add 0.5 of cpu to vm? it might not break anything as this value is just passed by to libvirt, but it will break on versioning | 16:07 |
| inc0 | thing is, predicting is hard, especially the future... | 16:08 |
| dansmith | that spans two domains, the modeling and the "business logic" I think | 16:08 |
| inc0 | and its harder to make translation rules between several field than to one better defined and more flexible | 16:08 |
| inc0 | it does, but still it might require hacks, or migrations | 16:09 |
| dansmith | I dunno what to say | 16:09 |
| inc0 | well, I'll do patchset, we'll have prototype to discuss | 16:10 |
| inc0 | how about that? | 16:10 |
| *** andreykurilin_ has joined #openstack-oslo | 16:10 | |
| dansmith | I guess | 16:10 |
| inc0 | ok, by this positive accent I conclude this day:) cya later dansmith | 16:12 |
| inc0 | and rest | 16:12 |
| dansmith | inc0: see ya :) | 16:13 |
| *** andreykurilin_ has quit IRC | 16:19 | |
| *** andreykurilin_ has joined #openstack-oslo | 16:20 | |
| *** MasterPiece has joined #openstack-oslo | 16:23 | |
| openstackgerrit | Mehdi Abaakouk proposed openstack/oslo.messaging: amqp1: fix functional tests deps https://review.openstack.org/153754 | 16:25 |
| *** stevemar has joined #openstack-oslo | 16:26 | |
| *** tsekiyama has joined #openstack-oslo | 16:27 | |
| *** harlowja_at_home has quit IRC | 16:27 | |
| *** ihrachyshka has quit IRC | 16:35 | |
| *** saikrishna has joined #openstack-oslo | 16:35 | |
| *** denis_makogon has quit IRC | 16:36 | |
| *** MasterPiece has quit IRC | 16:40 | |
| *** MasterPiece has joined #openstack-oslo | 16:40 | |
| *** ankit_ag has quit IRC | 16:49 | |
| *** MasterPiece has quit IRC | 16:55 | |
| *** daniel3_ has joined #openstack-oslo | 16:56 | |
| *** MasterPiece has joined #openstack-oslo | 17:03 | |
| *** MasterPiece has quit IRC | 17:07 | |
| *** MasterPiece has joined #openstack-oslo | 17:08 | |
| *** e0ne is now known as e0ne_ | 17:09 | |
| *** dulek has quit IRC | 17:09 | |
| *** e0ne_ is now known as e0ne | 17:11 | |
| *** noelbk has joined #openstack-oslo | 17:21 | |
| *** andreykurilin_ has quit IRC | 17:21 | |
| openstackgerrit | Steve Martinelli proposed openstack/oslo.policy: document the migration process and update the docs a bit https://review.openstack.org/154752 | 17:23 |
| *** alexpilotti has joined #openstack-oslo | 17:25 | |
| *** dtantsur is now known as dtantsur|afk | 17:28 | |
| *** i159 has quit IRC | 17:33 | |
| *** arek_ch has quit IRC | 17:33 | |
| openstackgerrit | Dan Smith proposed openstack/oslo.versionedobjects: Rename tests/fixtures to test/obj_fixtures to avoid import collisions https://review.openstack.org/154971 | 17:39 |
| openstackgerrit | Dan Smith proposed openstack/oslo.versionedobjects: Replace metaclass-based registry with a decorator-based one https://review.openstack.org/154972 | 17:39 |
| *** MasterPiece has quit IRC | 17:45 | |
| dhellmann | bnemec: what's your take on the subprocess monkeypatching change? https://review.openstack.org/#/c/153216/3 | 17:45 |
| openstackgerrit | Steve Martinelli proposed openstack/oslo.utils: Add fileutils to oslo_utils https://review.openstack.org/154975 | 17:45 |
| bnemec | dhellmann: I still think it's enabling bad behavior in consuming applications. | 17:47 |
| *** harlowja_away is now known as harlowja | 17:47 | |
| bnemec | And I think neutron is working on a patch to monkey patch earlier so it isn't necessary. | 17:47 |
| dhellmann | bnemec: ok, I'll let you -2 it then | 17:48 |
| harlowja | dims__ what u say about taskflow, lol | 17:48 |
| dhellmann | bnemec: I agree that the apps should be setting up the monkeypatching themselves, although I wonder if we can make it easier to get right? is that just a documentation issue, or can we code something? | 17:48 |
| harlowja | * dims__ inc0 (me just trying to figure out what was mentioned, ha) | 17:50 |
| bnemec | dhellmann: I dunno, I did consider an oslo.eventlet library or something, but I think it's mostly a documentation issue. | 17:50 |
| bnemec | dhellmann: Basically the eventlet best practices spec is my attempt to resolve this once and for all. | 17:50 |
| bnemec | We'll see how that goes. :-) | 17:50 |
| dhellmann | bnemec: does an app developer need to know which modules oslo libs need patched? or are we expecting them to patch everything? | 17:50 |
| bnemec | dhellmann: So that's an interesting question. harlowja has found that if, for example, you don't patch the thread module you can get random hangs no matter what we do. | 17:52 |
| *** yamahata has quit IRC | 17:52 | |
| bnemec | My preference would be patch everything or nothing. Accept eventlet into your heart, as harlowja put it. ;-) | 17:52 |
| harlowja | *when u pull in libraries from pypi u don't know the patterns they are doing inside that library (and really won't ever know) | 17:52 |
| bnemec | But I know that's not what we're doing today, so I think discussion is going to be needed. | 17:52 |
| harlowja | and if they are doing certain patterns they can potentially hang :-/ | 17:52 |
| * dhellmann is so tired of talking about eventlet | 17:53 | |
| harlowja | bnemec amen, lol | 17:53 |
| bnemec | Yeah, the whole monkey patching thing is one of those ideas that was too clever by half. | 17:53 |
| dhellmann | ok, I think you're both right that this isn't a patch we want, and maybe the spec bnemec is writing will help | 17:53 |
| *** MasterPiece has joined #openstack-oslo | 17:53 | |
| *** MasterPiece| has joined #openstack-oslo | 17:54 | |
| harlowja | accept it into your heart! | 17:55 |
| harlowja | lol | 17:55 |
| *** achanda has joined #openstack-oslo | 17:55 | |
| *** MasterPiece has quit IRC | 17:58 | |
| *** vigneshvar has joined #openstack-oslo | 18:02 | |
| dims__ | harlowja: that the persistence was not complicated | 18:03 |
| harlowja | dims__ depends on who u ask, everything is complicated to someone :-P | 18:03 |
| harlowja | but ya, i tried to not make it that complicated, ha | 18:04 |
| harlowja | if u have the ability to integrate running things with persistence (then i think its less complicated) but if u try to layer it in later (then i think it gets more complicated) | 18:04 |
| harlowja | taskflow does the first path (not the second) | 18:04 |
| harlowja | ^ if that made any sense :-P | 18:05 |
| dims__ | harlowja: y trying to understand. the other thing they were talking about was nested versionedobjects and if taskflow needed it (heat templates are nested) | 18:12 |
| dansmith | dims__: not really nested versionedobjects, but versioned objects with fields that contain unversionable arbitrary complexity blobs of stuff :) | 18:13 |
| dansmith | I don't expect taskflow needs to do that, but I don't know | 18:13 |
| harlowja | i could see it being useful for stuff that taskflow persists (aka when people return things from tasks, they get persisted, so if people returned versionedobjects then peoples upgrade path when they resume/upgrade stuff could be nicer) | 18:13 |
| dansmith | nested objects is a thing we already do | 18:14 |
| dims__ | right | 18:15 |
| harlowja | i'll use it if i can (pretty please make versionedobjects pretty generic, ha) | 18:16 |
| harlowja | *and pretty | 18:17 |
| harlowja | lol | 18:17 |
| *** saikrishna has quit IRC | 18:26 | |
| openstackgerrit | Merged openstack/debtcollector: Add doctested examples into the documentation https://review.openstack.org/153752 | 18:28 |
| *** yamahata has joined #openstack-oslo | 18:28 | |
| *** rushiagr is now known as rushiagr_away | 18:29 | |
| *** david-lyle is now known as david-lyle_afk | 18:31 | |
| harlowja | ok, mtreinish lets get out a release for debtcollector | 18:37 |
| mtreinish | harlowja: cool | 18:38 |
| *** jaosorior has quit IRC | 18:41 | |
| harlowja | mtreinish https://pypi.python.org/pypi/debtcollector/0.2.0 ok dokie | 18:44 |
| harlowja | guess requirements repo will need adjustment if u want to use it | 18:44 |
| gordc | is there suppose to be an oslo.cache lib somewhere or is google lying to me? | 18:44 |
| harlowja | only yahoo doesn't lie to u | 18:45 |
| harlowja | lol | 18:45 |
| gordc | :) /me goes to yahoo | 18:45 |
| harlowja | haha | 18:46 |
| gordc | hmm.. didn't realise bing still powered yahoo. | 18:47 |
| *** e0ne has quit IRC | 18:47 | |
| harlowja | for a little while longer i thinkk | 18:47 |
| gordc | ah i see. time for me to go to bing and look at pretty moving pictures. | 18:48 |
| *** mkoderer_cloud has quit IRC | 18:48 | |
| harlowja | jogo fyi, that debtcollector release is new this cycle (so shouldn't block your work/cause problems) | 18:51 |
| mtreinish | harlowja: well you can use it now (well once the mirror syncs) you only need to bump in g-r if a project starts having a hard req on 0.2.0 | 18:55 |
| harlowja | mtreinish isn't that what u wanted to do (use it but needed 0.20) :-P | 18:55 |
| harlowja | so seems like a hard req no? | 18:55 |
| mtreinish | harlowja: heh, yeah. I was just saying | 18:56 |
| harlowja | ah, k | 18:56 |
| harlowja | ya, agreed | 18:56 |
| mtreinish | I'll let jogo handle that because it's his patch... | 18:56 |
| *** crc32 has joined #openstack-oslo | 18:59 | |
| harlowja | k | 18:59 |
| * harlowja nearly forgot who's patch it was, ha | 18:59 | |
| *** MasterPiece| has quit IRC | 19:00 | |
| *** ViswaV has quit IRC | 19:01 | |
| *** ViswaV has joined #openstack-oslo | 19:02 | |
| * jogo is happy to let mtreinish handle it if he wants | 19:03 | |
| jogo | harlowja: just -1 my patch with a link to what I should do | 19:03 |
| *** noelbk has quit IRC | 19:04 | |
| harlowja | k | 19:04 |
| *** MasterPiece| has joined #openstack-oslo | 19:04 | |
| *** amotoki has quit IRC | 19:04 | |
| *** alexpilotti has quit IRC | 19:10 | |
| *** amotoki has joined #openstack-oslo | 19:12 | |
| *** noelbk has joined #openstack-oslo | 19:17 | |
| dhellmann | gordc: you might be thinking of http://specs.openstack.org/openstack/oslo-specs/specs/kilo/oslo-cache-using-dogpile.html, which we haven't started on yet | 19:19 |
| *** inc0 has quit IRC | 19:21 | |
| *** ViswaV has quit IRC | 19:21 | |
| *** ViswaV has joined #openstack-oslo | 19:26 | |
| *** achanda has quit IRC | 19:27 | |
| *** boris-42 has quit IRC | 19:32 | |
| *** noelbk has quit IRC | 19:32 | |
| *** achanda has joined #openstack-oslo | 19:42 | |
| *** noelbk has joined #openstack-oslo | 19:48 | |
| *** david-lyle_afk is now known as david-lyle | 19:51 | |
| *** BrianShang has quit IRC | 19:57 | |
| *** cdent has quit IRC | 19:57 | |
| *** prad has joined #openstack-oslo | 20:05 | |
| *** crc32 has quit IRC | 20:06 | |
| *** amrith is now known as _amrith_ | 20:07 | |
| *** crc32 has joined #openstack-oslo | 20:10 | |
| dansmith | dims__: https://review.openstack.org/#/c/154971/1 | 20:11 |
| dansmith | dims__: that fixes a problem with import ambiguity | 20:11 |
| *** _gryf has quit IRC | 20:11 | |
| dansmith | dims__: also interested in feedback on the patch on top of that, which is one of the work items | 20:12 |
| *** e0ne has joined #openstack-oslo | 20:12 | |
| dims__ | dansmith: we seem to be following a pattern in naming the fixture - http://paste.openstack.org/show/171567/ am ok with leaving the name what you have, but just checking | 20:16 |
| gordc | dhellmann: i see... so i guess if i did want some sort of cache i should use the one sitting in oslo-incubator... or an external lib | 20:16 |
| dims__ | dansmith: looking at the other one | 20:18 |
| dansmith | dims__: I normally don't name things test_foo unless it's testing foo, but that's fine, don't much care | 20:19 |
| dims__ | good point | 20:20 |
| *** ihrachyshka has joined #openstack-oslo | 20:26 | |
| *** ViswaV has quit IRC | 20:28 | |
| *** fnordahl has quit IRC | 20:32 | |
| *** fnordahl has joined #openstack-oslo | 20:32 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 20:33 | |
| *** crc32 has quit IRC | 20:34 | |
| *** amotoki has quit IRC | 20:34 | |
| *** achanda has quit IRC | 20:36 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 20:36 | |
| dims__ | dansmith: if one of the other projects define their own class extending base.NovaObject, is there something in the test harness they can use (other than fixtures) | 20:37 |
| openstackgerrit | Merged openstack/oslo.policy: Use standard logging in oslo.policy https://review.openstack.org/154635 | 20:37 |
| dansmith | dims__: meaning, to reuse some of the test infrastructure that is here? | 20:37 |
| openstackgerrit | Joshua Harlow proposed openstack/oslo.serialization: Avoid using strtime for serializing datetimes https://review.openstack.org/154547 | 20:38 |
| dims__ | dansmith: y (like we use _VirtDriverTestCase for various drivers) | 20:38 |
| dansmith | dims__: yeah, I was thinking about that when doing the registry patch.. the versions test is really not even relevant for the library, but is something that other projects will want to have to check their objects like we do in nova | 20:39 |
| dansmith | dims__: so I'm not sure what the prior art is there, I'll have to think on it a bit | 20:39 |
| dims__ | thanks dansmith | 20:39 |
| dhellmann | gordc: use dogpile | 20:41 |
| dhellmann | gordc: the oslo lib was just going to be a bunch of configuration options for it, so we can extract whatever you come up with to start that lib | 20:41 |
| dhellmann | morganfainberg: ^^ | 20:42 |
| morganfainberg | dhellmann: ++ | 20:42 |
| morganfainberg | gordc: what dhellmann said. And it was going to be used to be able to replicate what is in keystone caching. | 20:43 |
| morganfainberg | Elsewhere (but also better) and replace the keystone caching with something a bit more standard across the projects. | 20:43 |
| gordc | cool cool. sounds like a plan. | 20:44 |
| gordc | will work with dogpile for now. | 20:45 |
| openstackgerrit | Merged openstack/oslo-incubator: Update policy.py status in MAINTAINERS https://review.openstack.org/154565 | 20:45 |
| morganfainberg | gordc: zzzeek is the maintainer of dogpile. So we have expertise in house :) | 20:47 |
| openstackgerrit | Dan Smith proposed openstack/oslo.versionedobjects: Remove the ambiguous context argument on remotable methods https://review.openstack.org/155048 | 20:48 |
| *** ajo has quit IRC | 20:54 | |
| *** ajo has joined #openstack-oslo | 20:55 | |
| openstackgerrit | Dan Smith proposed openstack/oslo.versionedobjects: Remove the ambiguous context argument on remotable methods https://review.openstack.org/155048 | 20:57 |
| *** ajo has quit IRC | 21:01 | |
| dims__ | dansmith: if the object does not have an emedded context then we'll see OrphanedObjectError? | 21:02 |
| dansmith | dims__: yep | 21:02 |
| stevemar | dhellmann, i'm calling it - we should be ready to tag a release of oslo.policy, i double checked all the items here: https://etherpad.openstack.org/p/graduate-policy | 21:03 |
| dhellmann | stevemar: ok, I'm wrapped up in this version capping thing at the moment, but I'll review everything and if it looks good I will cut 0.1.0 tomorrow | 21:05 |
| stevemar | dhellmann, \o/ | 21:06 |
| dhellmann | stevemar: also, if you haven't already, please update the work items on the blueprint | 21:06 |
| stevemar | will do sir | 21:06 |
| dhellmann | ty! | 21:06 |
| openstackgerrit | Mehdi Abaakouk proposed openstack/oslo.messaging: rabbit: Fix behavior of rabbit_use_ssl https://review.openstack.org/154446 | 21:07 |
| dims__ | good work stevemar! | 21:07 |
| stevemar | dims__, ty - was also wondering about fileutils in incubator | 21:08 |
| stevemar | i posted a patch to move it to oslo.utils, i'm wondering if i need to create a whole blueprint for that too? | 21:08 |
| dims__ | stevemar: will check. give me a few | 21:10 |
| *** ViswaV has joined #openstack-oslo | 21:12 | |
| *** kgiusti has quit IRC | 21:12 | |
| *** ajo has joined #openstack-oslo | 21:15 | |
| dims__ | dansmith: what does serialize_entity use context for? | 21:16 |
| dansmith | dims__: I think it's there for symmetry with deserialize() | 21:17 |
| dims__ | dansmith: cool, it's not really needed or used from what i can tell | 21:23 |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Move 'provides' and 'name' to instance attributes https://review.openstack.org/155060 | 21:23 |
| dansmith | dims__: we put it there because it was part of oslo.messaging api, and since the serializer is defined in each project, we didn't want to limit them to only the context on deserialize | 21:23 |
| dansmith | dims__: see that that class inherits from messaging.NoOpSerializer ? | 21:24 |
| dims__ | ah. got it. thanks i did miss that | 21:24 |
| *** achanda has joined #openstack-oslo | 21:36 | |
| *** achanda has quit IRC | 21:39 | |
| *** achanda has joined #openstack-oslo | 21:39 | |
| *** kgiusti has joined #openstack-oslo | 21:40 | |
| dims__ | stevemar: left a note and added dhellmann to the review | 21:42 |
| stevemar | dims__, cool! | 21:42 |
| dims__ | stevemar: i had exactly the same thought before paris, so am with you | 21:43 |
| *** tsekiyama has quit IRC | 21:45 | |
| *** tsekiyama has joined #openstack-oslo | 21:45 | |
| stevemar | dims__, looks like concurrency depends on utils now anyway =\ | 21:47 |
| dims__ | stevemar: cool, so we should be all set :) | 21:48 |
| *** daniel3_ has quit IRC | 21:48 | |
| dims__ | may be i was mis-remembering some other dependency chain. will +2 it and let's wait for dhellmann | 21:48 |
| *** daniel3_ has joined #openstack-oslo | 21:50 | |
| *** cdent has joined #openstack-oslo | 21:50 | |
| *** daniel3_ is now known as Guest34776 | 21:50 | |
| stevemar | sounds good to me | 21:51 |
| dims__ | stevemar: dug up the etherpad - https://etherpad.openstack.org/p/kilo-oslo-library-proposals | 21:51 |
| dims__ | "We may want to split some of the things like fileutils, imageutils, etc. out to separate libraries just to avoid dependency bloat " | 21:52 |
| dims__ | quote from dhellmann | 21:52 |
| stevemar | theres also, re: oslo.context: This library is needed to release oslo.log, and oslo.log is blocking oslo.concurrency because currently oslo.concurrency includes a version of the log settings in the vendored copy of log needed by fileutils | 21:52 |
| stevemar | oh i guess... i mean fileutils just depends on std libs, no 3rd party stuff | 21:54 |
| dims__ | right | 21:54 |
| stevemar | if it's packaged solo that would be nice | 21:54 |
| stevemar | but its not like oslo.utils has tons of bloat | 21:55 |
| stevemar | https://github.com/openstack/oslo.utils/blob/master/requirements.txt | 21:55 |
| dims__ | either way we have to revisit stuff to see where we are since that etherpad is old | 21:55 |
| dims__ | stevemar: right | 21:55 |
| stevemar | i'm legitimately indifferent, on one hand it's nice to make something like oslo.io, but OTOH the reqs for utils are small | 21:56 |
| stevemar | what are the odds that a project isn't going to use oslo.utils | 21:56 |
| * stevemar shrugs | 21:56 | |
| dims__ | stevemar: next oslo meeting, we can take it to vote | 21:57 |
| stevemar | fair enough, when are those? | 21:57 |
| dims__ | monday noon-ish i think | 21:58 |
| *** Guest34776 has quit IRC | 21:58 | |
| dims__ | 1600 UTC | 21:58 |
| dims__ | https://wiki.openstack.org/wiki/Meetings/Oslo | 21:59 |
| *** daniel3__ has joined #openstack-oslo | 22:01 | |
| *** noelbk has quit IRC | 22:02 | |
| dhellmann | stevemar, dims__ : the client libs use oslo.utils, so we don't want to drag too much in as secondary requirements | 22:02 |
| stevemar | dhellmann, but fileutils doesn't add to any secondary requirements, AFAICT it's just std lib | 22:04 |
| dhellmann | stevemar: ok, that's fine -- I just wanted to make sure that was kept in mind | 22:04 |
| *** ihrachyshka has quit IRC | 22:04 | |
| dhellmann | I also want to avoid throwing everything into oslo.utils, just as a random bag of whatever we can't think of where else to put it | 22:05 |
| dhellmann | some of this stuff is really stable, so making a new library is a one time cost | 22:05 |
| stevemar | thats fair | 22:06 |
| dhellmann | the benefit is we get the whole history of the file, which we don't if we copy the file between existing repositories -- though I'm told there may be a way to merge those histories, too | 22:06 |
| dhellmann | I probably didn't do a good enough job of writing down why I came up with each of those libraries, so as dims__ said we should go through it again | 22:07 |
| dhellmann | I would lean towards saving history, myself, but I can be persuaded it's not worth the trouble :-) | 22:07 |
| dhellmann | stevemar: for some reason I thought fileutils had code that only glance was using, but it doesn't look like it, so this may be a good candidate to get into oslo.utils | 22:08 |
| stevemar | dhellmann, afaik, keystone and oslo.policy would use it too | 22:09 |
| dhellmann | stevemar: yeah, I just mis-remembered what it was | 22:12 |
| *** kgiusti has left #openstack-oslo | 22:13 | |
| *** noelbk has joined #openstack-oslo | 22:13 | |
| dims__ | dhellmann: we are growing old dhellmann :) | 22:15 |
| dhellmann | dims__: my "friend" said to me recently, "there are people younger than us who are old" | 22:16 |
| dims__ | dhellmann: classic | 22:23 |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Move 'provides' and 'name' to instance attributes https://review.openstack.org/155060 | 22:25 |
| *** cdent has quit IRC | 22:47 | |
| *** cdent has joined #openstack-oslo | 22:48 | |
| *** mriedem has quit IRC | 22:52 | |
| openstackgerrit | Manish Godara proposed openstack/taskflow: Clarify documentation related to inputs https://review.openstack.org/155093 | 22:54 |
| *** ajo has quit IRC | 22:58 | |
| *** achanda has quit IRC | 23:00 | |
| *** achanda has joined #openstack-oslo | 23:05 | |
| *** e0ne has quit IRC | 23:20 | |
| *** cdent has quit IRC | 23:30 | |
| *** achanda has quit IRC | 23:32 | |
| *** achanda has joined #openstack-oslo | 23:33 | |
| *** MasterPiece| is now known as MasterPiece | 23:40 | |
| *** prad has quit IRC | 23:41 | |
| openstackgerrit | Merged openstack/oslo.policy: document the migration process and update the docs a bit https://review.openstack.org/154752 | 23:47 |
| *** dims__ has quit IRC | 23:49 | |
| *** dims__ has joined #openstack-oslo | 23:50 | |
| *** achanda has quit IRC | 23:52 | |
| openstackgerrit | Joshua Harlow proposed openstack/oslo.serialization: Avoid using strtime for serializing datetimes https://review.openstack.org/154547 | 23:52 |
| *** achanda has joined #openstack-oslo | 23:53 | |
| openstackgerrit | Joshua Harlow proposed openstack/taskflow: Improve upon/adjust/move around new optional example https://review.openstack.org/155111 | 23:53 |
| *** achanda has quit IRC | 23:53 | |
| *** dims__ has quit IRC | 23:54 | |
| *** achanda has joined #openstack-oslo | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!