[08:17] ₘₒᵢₙ ₘₒᵢₙ [08:18] o/ [09:03] * Chipaca doing dispatch like https://boingboing.net/2020/04/16/watch-a-plane-gracefully-land.html [09:19] grah, fake_script_* depends on $0 being the script and not a symlink to it [09:19] Symlinks don't work very well on windows [09:19] With Python, specifically [09:20] There's code in the interpreter that hides away the symlink name [09:22] i think windows doesn't have symlinks of non-directories, but i might be wrong [10:38] need to improve the test suite so it supports non-symlink things [10:38] but that can be a followup [10:59] * Chipaca ⇝ runch [10:59] Good idea [11:17] Muy buenos días a todos! [11:59] Chipaca, I'm not sure if #221 is ready to review or wip (does github have a way to signal this?) [11:59] PR #221: ops/main.py: support dispatch [12:03] oh, I've found a "Still in progress? Convert to draft" thing [12:50] * facubatista -> errands, will be back for the standup [13:08] niemeyer: to start using spread, should i reuse the snapd credentials? [13:09] Chipaca: To play with it you can set it up with your own credentials as you did in snapd.. once this is ready for merging we'll need to set up the repository as well [13:09] ok [13:32] * facubatista is back [14:58] facubatista, Chipaca: I'll need a few more minutes.. [15:36] ooh, #232 is nice [15:36] PR #232: ops.testing.Harness actions metadata support. (#231) [15:36] might need changes but as a followup [15:37] * facubatista -> lunch [15:47] Chipaca: Indeed [15:47] niemeyer: ah, missed your comment, i could fix that as well :) [15:47] * Chipaca does so [15:48] Thanks! [15:49] bah [15:49] niemeyer: all the assertEqual in that file seem to be (expected, got) [15:49] boo [15:49] no, i lie [15:49] half of them? [15:49] grmbl [15:50] the ones in the odd-numbered tests [15:50] Chipaca: :P [15:50] definitely not something to burden this pr with [15:50] Chipaca: Agreed, as mentioned there [15:50] niemeyer: yeah, i thought it was just one test or two to change, related to the pr... but no [15:50] Chipaca: I know for sure that the original tests had expected on right hand side, consistently [15:51] Chipaca: (because I wrote them.. ;) [15:51] Debating about the right way would be harder, though :) [15:51] :-D [15:51] the docs are very unhelpful [15:52] assertEqual(first, second) [15:52] psh [15:52] anyhoo, i'm going to go buy bread [15:52] Yeah.. let's just keep expected on right hand side.. [15:52] I think this is the most common convention for tests in general [15:52] Including if statements [15:52] There are certainly variations, but I think it's fair to say the constant tends to be after the variable more often than not [15:53] * Chipaca nods [15:54] i might make a pass to do that change later, if i need the distraction [15:54] (ie if dispatch gets too gnarly) [16:02] ok now yes i go out for a bit [18:08] Chipaca, we need to create the project in GH... `charm`? `charm-tool`? [18:08] `charm-cli`? [19:58] charm-tool sounds like the better of these.. not sure if there's a better option [20:03] niemeyer, I wonder how this collides with *the current* charm tool [20:03] facubatista: This is the next version of the current tool [20:04] niemeyer, yeap, but where is the old one hosted? shall we have a big sign in the README saying "don't confuse this `charm` with the other `charm`", etc [20:06] I mean, I expect some confusion about both, like I have a bug will not know in which project to open the issue, that kind of detail [20:11] Issue operator#234 opened: Debugger "at" reserved words [20:31] * facubatista eods [20:31] see you all on Monday