| [01:21] <wgrant> cprov, cjohnston: people.canonical.com/~wgrant/launchpad/almost-final‽.png | |
| [01:22] <cprov> wgrant: hey hey | |
| [01:24] <cprov> it is coming up very nicely and since the new layout is more flexible (div) it will be easier to increment. | |
| [01:24] <wgrant> I hope so. | |
| [01:25] <cprov> wgrant: I've just opened an asana task about finding a way to materialise the IC threads, specially in emails | |
| [01:25] <wgrant> I've also dropped the "draft" and "publish" terminology. It nows says "Unsaved comment", and then "Include X inline comments" next to the "Save Comment" button. | |
| [01:25] <wgrant> Yeah, I've been wondering about how to do that. | |
| [01:25] <wgrant> It's a bit awkward, as they're all at different depths potentially. | |
| [01:25] <cprov> wgrant: it's annoying to to read the reviews right now | |
| [01:27] <cjohnston> looks good wgrant | |
| [01:27] <wgrant> Someone want to review? https://code.launchpad.net/~wgrant/launchpad/ic-js-cleanup/+merge/219531 | |
| [01:28] <cprov> wgrant: I will | |
| [01:29] <cprov> wgrant: can you review https://code.launchpad.net/~cprov/launchpad/delete-branches-with-ics ? | |
| [01:30] <wgrant> Hm | |
| [01:30] <wgrant> What do you guys think about renaming "inline" to "diff"? | |
| [01:30] <wgrant> That pretty much just affects the "Include X inline comments" checkbox | |
| [01:31] <wgrant> Makes it clearer what it refers to. | |
| [01:31] <cprov> wgrant: I wonder if I can just add few other 'ouch' in your MP and let you figure them out ;-) | |
| [01:31] <wgrant> Heh | |
| [01:31] <cjohnston> +1, more ouch | |
| [01:31] <cprov> wgrant: +1 on "diff comments" | |
| [01:32] <cjohnston> if you do that you should probably do the same in the email | |
| [01:32] <cjohnston> what about 'View inline comments' | |
| [01:32] <wgrant> Oh yeah, email and link too, forgot those. | |
| [01:33] <cjohnston> I would argue that 'regular' comments are diff comments too.. | |
| [01:33] <cjohnston> IC is actually inline | |
| [01:34] <cprov> wgrant: table-layout: fixed, not only predictable but also faster/easier for browsers. Nice one! | |
| [01:35] <cprov> had to look it up ... | |
| [01:36] <wgrant> It's pretty useful here, since divs expand to fill their container, and tds expand to fit their contents. | |
| [01:38] <cprov> wgrant: what about the TextWidget for review_type ? what does it bring ? | |
| [01:38] <wgrant> cprov: I narrowed it from 20 characters to 15, to ensure that the widgets all fit on one line. | |
| [01:39] <cprov> ah, cool | |
| [01:45] <wgrant> cprov: The rename is pushed. | |
| [01:54] <cprov> wgrant: what does the CSS added in the BMP template do ? | |
| [01:55] <wgrant> cprov: That's a variation on cjohnston's. It adds a bit of space before the publish checkbox. | |
| [01:55] <wgrant> Also, I've just pushed a regression fix for something that seems to have broken a few days ago: the publish checkbox was being added unconditionally, so inline comments crashed and failed to load for anonymous users. | |
| [01:57] <cprov> I think I had fixed that in 17011, on top of cjohnston fixes ... | |
| [01:57] <wgrant> Oh, so you did | |
| [01:58] <wgrant> I'll uncommit and merge | |
| [02:02] <wgrant> cprov: Have you seen any more issues on (qa)staging since those firewall fixes a few hours ago? | |
| [02:02] <cprov> wgrant: wait, there were 2 problems, you have to manually merge the conflicting hunks. I've only fixed the checkbox being rendered multiple times (in tests, essentially) and you have fixed the not-logged-in issue. | |
| [02:03] <wgrant> Also, you added a tab! | |
| [02:03] <wgrant> Die | |
| [02:08] <cprov> I must, I don't know why the emacs in my vm is allowing this shit ... I will review my config | |
| [02:09] <cprov> jslint should have exploded in my face when I try to commit, that's how we should setup project for dummies (like me) | |
| [02:10] <wgrant> Heh | |
| [02:15] <wgrant> cprov: The line-no col is wide enough to contain five digits plus the add sprite. | |
| [02:17] <cprov> up to 99999 lines diffs then, not at problem | |
| [02:17] <wgrant> Well, if you go back to your old days... :P | |
| [02:17] <cprov> wgrant: so, you really want new methods on CRICSet ... | |
| [02:18] <wgrant> cprov: All the other manipulation code (apart from person merge) is in codereviewinlinecomment.py | |
| [02:18] <wgrant> I really don't want it to get like various other LP model classes, where there are ten different modules that alter them. | |
| [02:19] <wgrant> Makes it very difficult to change anything. | |
| [02:20] <cprov> right, you think I should implement also the removal on CRICSet methods, or should I return ResultSet to be removed on the callsite ? | |
| [02:21] <wgrant> I think you should just implement a remove method. We don't expose any ResultSets from CRICS at the moment, and I can't really see any valid reason for someone to want one. | |
| [02:23] <cprov> wgrant: agreed, also the removal() should really consider both CRIC and CRICD as it is implemented in BMP, right ? they are likely to be removed together every time. | |
| [02:25] <wgrant> cprov: Right, that might well be a good thing to do. | |
| [02:25] <wgrant> I can't see much of a reason to split them. | |
| [02:32] <cprov> wgrant: code moved and pushed | |
| [02:39] <wgrant> cprov: I'd rename the method to removeFromDiffs or similar, but that looks great otherwise. Thanks. | |
| [02:41] <cprov> wgrant: no problemo, renamed version pushed | |
| [02:42] <wgrant> cprov: How much longer are you likely to be around? I think I might implement a quick fix to only include commented files (not just hunks, at least not yet) in emails. | |
| [02:42] <cprov> longer enough. | |
| [04:36] <cprov> wgrant: http://lpbuildbot.canonical.com/builders/lucid_lp_lxc/builds/1172/steps/shell_9/logs/summary is that old spurious failure, right ? | |
| [04:36] <wgrant> cprov: Yup, already forced. | |
| [04:36] <cprov> wgrant: thanks | |
| [04:37] <cprov> wgrant: how is the ic-email improvement ? | |
| [08:20] <cjwatson> wgrant: Any luck with the livefs reviews? I know there's been the whole pg9.3 thing, and lots of work on IC ... | |
| [08:20] <cjwatson> I've put the remaining things people are asking for on the shelf for now, since I reached the point where I could do a touch image build with an added PPA | |
| [08:22] <wgrant> cjwatson: Sorry, busy with staging and postgres and inline comments and PPA microservice and ephemeral PPAs and Malta... Three of those are coming to a close tomorrow, so I might actually have some time. | |
| [08:24] <cjwatson> OK, thanks | |
| [08:26] <wgrant> cjwatson: Is Laney's cron.germinate change sane? I saw some debate over what the name should be. | |
| [08:57] <cjwatson> Oh, that bit of the naming is fine. I was meaning to review and land that for him today. | |
| [08:58] <cjwatson> It will cause tasks to appear for sdk-libs or something like that, but they were actually wondering why they weren't there anyway | |
| [08:58] <cjwatson> And it will cause cron.germinate to take that bit longer, of course, but I've been looking at parallelising it ... | |
| [09:14] <cjwatson> Laney: Could you please set a commit message for https://code.launchpad.net/~laney/launchpad/germinate-ubuntu-touch/+merge/220233 ? | |
| [09:14] <Laney> cjwatson: Oh right, yes | |
| [09:15] <Laney> try that | |
| [09:17] <cjwatson> Laney: Thanks, landing | |
| [09:17] <Laney> cheers | |
| [12:05] <cjohnston> wgrant: have the new IC changes been released? | |
| [12:05] <cjohnston> I guess maybe it's DC changes now? ;-) | |
| [12:06] <wgrant> cjohnston: In progress. | |
| [12:06] <cjohnston> cool | |
| [12:06] <cjohnston> How'd the DB stuff go? | |
| [12:06] <wgrant> Not quite uneventful, but eventually successful. | |
| [12:06] <cjohnston> cool | |
| [12:06] <wgrant> The master is now 9.3, and the slaves are rebuilding on 9.3. | |
| [12:07] <cjohnston> oh nice.. so the upgrade was done.. | |
| [12:10] <wgrant> Yep | |
| [12:10] <wgrant> I poked the comments stuff pretty thoroughly this evening, and fixed a few bugs | |
| [12:11] <wgrant> Including a hilarious off-by-one error in the email formatter, where comments were showing up a line later, and the first line of the diff was always missing. | |
| [12:16] <jelmer> wgrant: is launchpad really going to do inline comments on code reviews? | |
| [12:17] <wgrant> jelmer: On Monday, with a bit of luck. | |
| [12:17] <jelmer> wgrant: good stuff | |
| [12:20] <cjohnston> by "with a bit of luck" he means pencil and paper | |
| [12:21] <cjohnston> wgrant: take a look at https://code.launchpad.net/~ursinha/uci-engine/ts-add-mp-support/+merge/220275 I think we may need to rethink the design of my "comment" a little... cjohnston a moment ago: (and then nothing) | |
| [12:21] <cjohnston> the way it is it looks really odd without any sort of review or comments or anything | |
| [12:21] <wgrant> cjohnston: Yeah, I think we probably want to move the "Show diff comments" link into the body somewhere. We really need space on the bottom of each comment to have links | |
| [12:22] <wgrant> Show diff comments, Reply, and Save/Cancel in the case of the inline editor. | |
| [12:22] <wgrant> There's no way to have actions today. | |
| [12:22] <wgrant> Also the Delete button, and possibly a future Edit feature | |
| [12:22] <cjohnston> edit! | |
| [12:22] <wgrant> It's all really awful atm, no consistency. And the inconsistent bits are dreadful anyway. | |
| [12:58] <wgrant> cjohnston: prod's all up to date now, if you want to prepare that email. | |
| [12:58] <cjohnston> ack. thanks | |
| === Ursinha is now known as Ursinha-afk | |
| === Ursinha-afk is now known as Ursinha | |
| === cprov_ is now known as cprov | |
| === _mup__ is now known as _mup_ | |
| [23:02] <cjohnston> wgrant: cprov, what do you think about moving the 'Add comment' stuff to below the diff... | |
| [23:02] <cjohnston> I think it would improve workflow | |
| [23:02] <cjohnston> If you are replying to a comment you may not need to see the diff, but otherwise, your probably going to be adding a comment after reading the diff | |
| [23:02] <cjohnston> so instead of having to click a link to go back up, why not just have the box below | |
| [23:46] <wgrant> cjohnston: The problem is that the diff is enormous. | |
| [23:46] <wgrant> This is why other sites have it on a separate tab. | |
| [23:47] <cjohnston> it apparently isn't completely discoverable That you have to click save comment | |
| [23:49] <wgrant> Yeah, that's one of the big remaining issues. | |
| [23:49] <wgrant> But moving the comment form below the diff would make it very difficult to find. | |
| [23:49] <wgrant> Even for MP-wide comments. | |
| [23:50] <cjohnston> personally i would expect it to be at the bottom and not Where it is | |
| [23:50] <cjohnston> it seems odd to have to scroll back up after scrolling to the bottom | |
| [23:51] <wgrant> Sure, but then you have to scroll through 5000 lines of diff. | |
| [23:51] <wgrant> Just to find the comment form | |
| [23:51] <wgrant> + it separates the comment form from the comment thread | |
| [23:51] <wgrant> Which is weird | |
| [23:51] <cjohnston> or scroll 10k lines to read the diff and then go back up | |
| [23:53] <cjohnston> what do you think about changing the new bottom link to be either 'return to add comment' or 'return to save comments' depending on if drafts are present or not | |
| [23:54] <cjohnston> that may help discoverability slightly | |