*** jtomasek has joined #storyboard | 07:52 | |
*** jtomasek_ has joined #storyboard | 07:59 | |
*** jtomasek_ has quit IRC | 08:48 | |
SotK | I should really make the PUT endpoint for "worklist items"/cards less dumb | 10:46 |
---|---|---|
Zara | off you go then, I'll just wait here | 10:47 |
Zara | :P | 10:47 |
openstackgerrit | Merged openstack-infra/storyboard: Move permissions checks into storyboard.db.api https://review.openstack.org/284312 | 10:58 |
openstackgerrit | Merged openstack-infra/storyboard: Fix permissions being able to be set by any authenticated user https://review.openstack.org/284313 | 10:58 |
Zara | so, trying to add a task to a board in my test instance; it's not showing up. queried the api directly for the task, and this is what I got: http://paste.openstack.org/show/488337/ | 11:37 |
Zara | for some reason it has no story id | 11:37 |
Zara | also, this is the request and response, which differ: http://paste.openstack.org/show/488338/ | 11:38 |
Zara | presumably it's somehow because the task has no story id | 11:39 |
Zara | the request is asking for the right thing. the response is talking about a different task, so that's strange | 11:39 |
Zara | I don't know how long this has been an issue, if it's just a problem with my db, or what | 11:40 |
Zara | just flagging it up in case anyone else sees something similar | 11:40 |
Zara | it's possible I just deleted a story in my test instance, and boards have a bug where they show tasks from deleted stories in the search, so I'm going to try that now | 11:42 |
Zara | okay, so that's part of what's going on | 11:45 |
Zara | though I don't know why the request and response bodies differ in the above case | 11:46 |
Zara | but I tried removing a story and then adding a task from it; sb lets you click on the task and then click 'add items' in a board, but that task won't display in the lane | 11:46 |
Zara | so that's part of what's going on, anyway. :) | 11:47 |
SotK | hmm, can you paste the log of your API by any chance? | 11:47 |
Zara | sure, though I restarted it between the first weird one and that one | 11:50 |
Zara | well, I can just try to add teh weird one again, hang on | 11:51 |
Zara | http://paste.openstack.org/show/488341/ | 11:53 |
Zara | so that's got me deleting a task from the board, then deleting the story it belonged to, going back to the board, trying to readd that task (now it has no story), and then trying to add the weird task from earlier for good measure | 11:53 |
Zara | I can post logs earlier than that but I don't think it's much help | 11:54 |
* SotK decides the POST endpoint needs to be less dumb too | 11:57 | |
SotK | can you do `select * from worklist_items where list_id = 122;` on your database and see if there is an entry for the task you tried to add? | 11:58 |
Zara | looks like there are two! :D | 12:00 |
Zara | I think I tried to add it twice, so that makes sense | 12:00 |
Zara | http://paste.openstack.org/show/488343/ | 12:01 |
Zara | they are both in the 4th position on the list xD | 12:01 |
Zara | also, for the weird one: | 12:03 |
Zara | http://paste.openstack.org/show/488345/ | 12:03 |
SotK | oh dear | 12:04 |
SotK | and none of them show up on the web ui? | 12:04 |
Zara | tasks 50 and 92 don't, no | 12:04 |
Zara | hm, I should find out what task 18 is | 12:04 |
Zara | hahaha I was looking at the db in the tab which was running the server, hang on... | 12:05 |
Zara | ahaha, that was the other one I tried to add which didn't appear | 12:06 |
Zara | but I couldn't find it with search and then forgot about it | 12:06 |
Zara | this one also says it has no story ID | 12:06 |
SotK | hm, can you paste the response body for the board GET request? | 12:07 |
Zara | http://paste.openstack.org/show/488348/ | 12:09 |
Zara | hang on | 12:09 |
Zara | that formatted weird | 12:09 |
Zara | oh... it always puts it on one line :/ | 12:11 |
Zara | that might be a bit old, too | 12:12 |
*** krotscheck_dcm is now known as krotscheck | 12:13 | |
Zara | http://pastebin.com/4cYcjytV | 12:13 |
Zara | that's a newer one and is more readable | 12:13 |
SotK | this is really weird | 12:20 |
Zara | I have to continue my tradition of finding the best bugs | 12:20 |
Zara | oh! something I noticed at home ysterday and forgot until now | 13:45 |
Zara | https://storyboard.openstack.org/#!/project/839 takes much longer than other projects to load | 13:45 |
Zara | I'm wondering if it's a bit like the strange marked thing before, and the length of the project title is affecting the loading speed | 13:46 |
Zara | still just speculation at this point. | 13:46 |
persia | Seems to be about the same time for my browser, when I'm not logged in, but my browser has issues as previously discussed. | 13:47 |
persia | Also seems to be the highest-numbered story, if that might be relevant | 13:48 |
Zara | I wondered about that, but I don't think it was a problem with the previous highest-numbered project | 13:48 |
Zara | I also notice s.o.o projects contain slashes and underscores, which storyboard doesn't permit a user to put in a title manually | 13:49 |
Zara | (I think projects are currently created in s.o.o automatically when a new project is created, somehow) | 13:49 |
Zara | oooh, interesting, after trying to load the page a few times it froze for a while and came up with the 'want to stop a script?' dialog | 13:51 |
Zara | (screenshot for the intersted: http://imgur.com/DnzFduS ) | 13:52 |
Zara | (I check new stories/projects every so often to catch this sort of thing, because I'm boring) | 13:54 |
persia | It probably makes sense to fix the project naming model to allow anything that can be a git repo, really. | 13:56 |
Zara | I don't yet know if that is what's causing the problem; other projects also seem to have characters that sb doesn't allow, and they may not have the same problems loading (I don't disagree, just talking relevant to this bug) | 13:58 |
Zara | the function starts `function(window,document,undefined` so it could be to do with the offsetWidth directive (iirc that's the one that causes some undefined errors) or the marked stuff. | 14:01 |
* Zara makes story https://storyboard.openstack.org/#!/story/2000496 | 14:03 | |
Zara | it may even be another 'this thing is too long' bug | 14:06 |
Zara | (project titles also have a length limit by default | 14:06 |
Zara | I think it's the title length | 14:07 |
Zara | I found something with a 27 character title and it behaved similarly | 14:07 |
Zara | actually, I can't count at all, because it's 26 characters *after* `openstack-infra/` | 14:12 |
Zara | things that are around 20ish characters after `openstack-infra` seem to behave sensibly; it seems to freak out around the 25character mark (so actually 25 +16 so 41 characters) | 14:13 |
Zara | not the most scientific investigation, but there we go | 14:14 |
Zara | (was the other longish title: https://storyboard.openstack.org/#!/project/805) | 14:14 |
Zara | I'm not sure how the clauses of that sentence ended up that way round, oh well | 14:14 |
* Zara updates story and goes back to reviewing | 14:16 | |
persia | The colon makes up for the delayed subject | 14:17 |
* SotK wonders why we have both https://storyboard.openstack.org/#!/project/836 and https://storyboard.openstack.org/#!/project/830 | 14:19 | |
Zara | my guess is it was renamed and it doesn't automatically delete old projects? I don't know. | 14:19 |
SotK | incidentally, the one with the underscore seems to take much longer to render | 14:20 |
Zara | ooh | 14:21 |
Zara | maybe it's a combination of length and underscores... or maybe longer titles are likely to contain more underscores | 14:22 |
Zara | though it's best checking a few times in case one is cached and the other isn't or something | 14:22 |
Zara | but I'm finding hte smae thing | 14:23 |
Zara | er, just rearrange all the letters in that sentence | 14:23 |
Zara | until it makes sense | 14:23 |
SotK | GET request takes ~150ms for each of them, it looks like its the rendering of the result that is slow to me | 14:24 |
Zara | quick, make a title that's nothing but underscores :P | 14:26 |
*** faybrocklebank has quit IRC | 14:36 | |
Zara | hm, card movement seems a little less smooth with https://review.openstack.org/#/c/284316/ , but I'm not sure if I'm imagining it | 14:36 |
* Zara goes back to master to compare | 14:36 | |
Zara | I think I'm imagining things. | 14:38 |
* SotK would expect so | 14:39 | |
Zara | :) | 14:41 |
Zara | I *think* what happened was that I made a new board for testing with slightly smaller cards, so they were more fiddly to move | 14:41 |
Zara | but we may never know for sure | 14:41 |
Zara | all seems fine, anyway | 14:41 |
*** faybrocklebank has joined #storyboard | 14:50 | |
Zara | okay, I think I've reached the end of my API reviewing, at least. all seems fine; any bugs I've found have been problems that already existed. https://review.openstack.org/#/c/284314/ and https://review.openstack.org/#/c/284315/4 are in need of +1s. I've tested the behaviour, but my python is a bit rubbish, so the code-checking will have been less thorough. | 14:54 |
Zara | (though that said, my migration was a bit of an odd custom job, so it'd be good if someone could check the behaviour there.) | 14:54 |
SotK | thanks Zara :) | 14:58 |
Zara | yw :) | 15:04 |
Zara | just thinking, last time a detail page was slow to display, it had a knockon-effect on the list page | 15:05 |
Zara | so we might find the projects list page displays faster if we can get to the bottom of the weird thing. | 15:05 |
Zara | (and maybe project groups list, too) | 15:05 |
Zara | though at this point I think storyboard's perceived speed is lagging behind its actual speed. looked around at some UX research last night, which suggested it was better to avoid a spinner for things that loaded in under a second (or maybe even two). | 15:06 |
Zara | well, it didn't specify spinner, it was talking more generally about loading progress displays | 15:07 |
Zara | the other thing is that our spinner turns reeeeeally slowly, and is a bit stop-start, which makes things seem to load more slowly than they load. | 15:09 |
pedroalvarez | any 400 or 401 so far? :) | 15:49 |
* SotK hasn't seen any that he remembers today | 15:49 | |
Zara | I had a couple this morning, but they might have been leftovers from before the patches were merged | 15:50 |
Zara | I haven't had any others today | 15:50 |
Zara | So I should update that story... | 15:50 |
openstackgerrit | Adam Coldrick proposed openstack-infra/storyboard: Archive cards instead of deleting them https://review.openstack.org/285392 | 16:00 |
Zara | \o/ | 16:01 |
openstackgerrit | Merged openstack-infra/storyboard: Add Due Dates to the API https://review.openstack.org/284314 | 16:17 |
openstackgerrit | Adam Coldrick proposed openstack-infra/storyboard-webclient: Show a modal to confirm archiving cards https://review.openstack.org/285417 | 16:41 |
SotK | might be more annoying than useful, idk | 16:41 |
Zara | yeah, I think it will be, but I'll test! | 16:42 |
Zara | I think cards should be pretty lightweight and wanting to unarchive them will be pretty rare | 16:43 |
Zara | (but should still be possible for those rare cases) | 16:43 |
Zara | but I'm saying this without testing the modal; maybe I'll love it. | 16:43 |
* SotK expects you won't :P | 16:45 | |
SotK | I was only worried about people accidentally clicking the archive button, but in that situation I expect most people will just re-add the card | 16:45 |
pedroalvarez | what if you didn't read the card before clicking? | 16:46 |
SotK | heh, that is a good point | 16:47 |
pedroalvarez | :) | 16:47 |
SotK | (when we have an events timeline for boards it won't matter, but until then that would be bad) | 16:47 |
Zara | I think (un)archiving is only useful in the case where the card has a due-date and the user has accidentally clicked 'delete'. which is important enough to account for, but we shouldn't assume it's the default. | 16:48 |
Zara | if it's really crucial then I'd imagine people would be aware of it anyway. | 16:49 |
Zara | unless it's a really far off deadline and/or nobody is assigned to it yet | 16:49 |
Zara | which I think is probably rare | 16:49 |
persia | My thought was the deprioritisation - reprioritisation cycle. | 16:50 |
persia | Where something is given a date, and nobody can do it, and it gets dropped, and then later it gets to be important again: it's handy to be reminded of the old date. | 16:50 |
persia | Other than that, I don't think anyone will notice the difference between deletion/archiving except if they are looking at card ID numbers. | 16:51 |
* SotK also thinks nobody will notice the difference, until they suddenly want a card back | 16:53 | |
SotK | at which point its nice to be able to bring it back | 16:53 |
SotK | aaaaaaalso it means we don't need to worry about adding more metadata to cards if necessary (I can't think of any examples) | 16:54 |
*** pleia2_ has joined #storyboard | 17:15 | |
*** colettecello has joined #storyboard | 17:15 | |
*** anteaya has quit IRC | 17:20 | |
*** krotscheck has quit IRC | 17:20 | |
*** gothicmindfood has quit IRC | 17:20 | |
*** pleia2 has quit IRC | 17:20 | |
*** krotscheck has joined #storyboard | 17:20 | |
*** anteaya has joined #storyboard | 17:28 | |
*** mrmartin has joined #storyboard | 17:36 | |
Zara | I'm going to hold off workflow +1 because it's friday night, but testing api stuff seems fine so far | 18:00 |
Zara | (I feel a bit mean doing that, buuut nobody will be able to see it until UI stuff is merged, and I've yet to test that, so I imagine the time we'll lose on Monday won't be more time than it takes to test) | 18:13 |
SotK | heh, the ui stuff was more of a bonus :) | 18:37 |
*** yarkot_ has joined #storyboard | 20:22 | |
*** mrmartin has quit IRC | 20:23 | |
*** yarkot_ has quit IRC | 20:26 | |
*** yarkot_ has joined #storyboard | 21:13 | |
*** yarkot_ has quit IRC | 21:29 | |
*** colettecello has quit IRC | 22:50 | |
*** gothicmindfood has joined #storyboard | 22:50 | |
*** jtomasek has quit IRC | 22:54 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!