These are chat archives for astropy/astropy

13th
May 2016
Michael Seifert
@MSeifert04
May 13 2016 12:37
Maybe a stupid question but is it possible to order inline-comments in a PR by their date? So that the most recently updated inline-comment is at the bottom? I don't like the ordering that the date of the first inline-comment determines the position I want it to be the last comment date.
Thomas Robitaille
@astrofrog
May 13 2016 14:03
@MSeifert04 - hmm, I don't think so :-/
Thomas Robitaille
@astrofrog
May 13 2016 16:14
@mwcraig, are you around?
Stuart Mumford
@Cadair
May 13 2016 16:15
WHAT IS THIS MADNESS
Thomas Robitaille
@astrofrog
May 13 2016 16:15
lol yeah we've been active
Stuart Mumford
@Cadair
May 13 2016 16:21
but you hid it away on this proprietary chat platform :p
Michael Seifert
@MSeifert04
May 13 2016 16:39
does anyone know or could check if sunpy, specutils,spectral-cube have problems because of the nddata PRs?
Thomas Robitaille
@astrofrog
May 13 2016 17:40
@MSeifert04 - you mean because of the already merged ones?
@Cadair, do you want to try running the test suite for sunpy with the latest astropy dev
?
Michael Seifert
@MSeifert04
May 13 2016 17:41
yes. I've checked sunpy (there was no failure against dev-astropy but I'm not sure if that was really against the development branch) and ccdproc (there I've submitted a patch)
Thomas Robitaille
@astrofrog
May 13 2016 17:44
spectral-cube doesn't use NDData yet
so no need to test that
Stuart Mumford
@Cadair
May 13 2016 17:48
We have a Travis build that does that.
Has something landed? I can restart it.
Thomas Robitaille
@astrofrog
May 13 2016 17:50
No I think @MSeifert04 already tested sunpy so all good
Michael Seifert
@MSeifert04
May 13 2016 17:52
@Cadair If you notice any failures on that build please let me know. I think I've tested it but wasn't sure.
Steve Crawford
@crawfordsm
May 13 2016 18:26
@astrofrog quick question -- I just merged Jake's pull request but I forgot to ask him to squash the commits. Big deal? Or should I revert and squash all of the commits?
Thomas Robitaille
@astrofrog
May 13 2016 18:26
No need to squash the commits
and there's no way to revert really, all it would do is add another commit that removes the code and then you'd have to add a squashed commit, so the history would contain the implementation twice
@crawfordsm - please don't revert it
Steve Crawford
@crawfordsm
May 13 2016 18:27
okay I won't
Thomas Robitaille
@astrofrog
May 13 2016 18:27
In the past, we've only asked people to squash commits if the history has things like merge conflicts or merge commits
but if it's a 'clean' history, even with 10-20 commits, as long as it's not for a trivial thing, it's fine
Steve Crawford
@crawfordsm
May 13 2016 18:28
Ah, thanks for the clarification
Thomas Robitaille
@astrofrog
May 13 2016 18:31
@crawfordsm - thanks for merging that PR! :)
Steve Crawford
@crawfordsm
May 13 2016 18:32
Thanks for yours (and everyone else's) help with it!
So question, I did have a someone who reviewed the performance of it and compared it to other packages (basically peer review), do we have a method for people to be appropriately acknowledge for that type of review? (it's something I'd like to do more with the stats package in the future)
Thomas Robitaille
@astrofrog
May 13 2016 18:38
@crawfordsm - we don't at the moment but there's no reason we can't
Maybe this is something you could bring up on the astropy-dev list?
Steve Crawford
@crawfordsm
May 13 2016 18:47
sure will do post-freeze.
Brigitta Sipocz
@bsipocz
May 13 2016 18:55
@olebole - What I do is to switch to the top of the bugfix branch, and cherry-pick the commits from the given PR I want to test. (Recently I started using gitx, a git gui, previously always did this on the command line). This way I can do local testing, or by creating a copy branch, it can be pushed up and tested by travis if you enable travis on your fork.
Brigitta Sipocz
@bsipocz
May 13 2016 19:01
(@crawfordsm @astrofrog thanks for not squashing the big PR, I always find it easier to figure out things with smaller commits and their appropriate commit message, in case of bug hunting ).
Michael Seifert
@MSeifert04
May 13 2016 22:53
@mwcraig are you around?