These are chat archives for ipython/ipython

30th
Jan 2015
Jonathan Frederic
@jdfreder
Jan 30 2015 00:12
Output doesn't show
the YouTube video is processing
Min RK
@minrk
Jan 30 2015 00:13
I see a 144p video where I cannot tell what's going on
Jonathan Frederic
@jdfreder
Jan 30 2015 00:13
dang
Min RK
@minrk
Jan 30 2015 00:13
aha
Jonathan Frederic
@jdfreder
Jan 30 2015 00:13
really :(
?
Min RK
@minrk
Jan 30 2015 00:13
I guess 144p was the first to transcode
it works after refresh
Thomas Kluyver
@takluyver
Jan 30 2015 00:13
could #7631 fix it?
Min RK
@minrk
Jan 30 2015 00:13
Can you check #7631?
Thomas Kluyver
@takluyver
Jan 30 2015 00:13
It's always fun when Min has fixed a bug before you notice it ;-)
Min RK
@minrk
Jan 30 2015 00:13
I think that's fixed
Jonathan Frederic
@jdfreder
Jan 30 2015 00:14
Yeah, I'll check
oh actually
I think it probably does
because I haven't been able to reproduce
it
but that particular cell
Min RK
@minrk
Jan 30 2015 00:14
I think it happens for any cell that has no output on load, but then produces output on run.
Jonathan Frederic
@jdfreder
Jan 30 2015 00:14
did overflow at one point
Min RK
@minrk
Jan 30 2015 00:15
@takluyver should terminado be included in ipython[notebook]?
(on non-Windows)
Thomas Kluyver
@takluyver
Jan 30 2015 00:15
I think we'd said no, on the grounds that it's superfluous in most installations
I'm open to reconsidering that
the UI saying 'terminals not available' instead of just keeping mum about them pushes me slightly more towards making it a dependency
Jessica B. Hamrick
@jhamrick
Jan 30 2015 00:23
@jdfreder You know, I’m pretty sure I ran into this same bug when I was giving a lecture on tuesday.
I thought it was a fluke, but it was so bizarre, because I had one cell that would run and produce output, and then other cells that wouldn’t.
Thomas Kluyver
@takluyver
Jan 30 2015 00:26
ah, I think I saw @minrk running into that same issue earlier today when testing something
Min RK
@minrk
Jan 30 2015 00:33
@takluyver we could add it only to the all extra
so that it's included in 'all the things'
Thomas Kluyver
@takluyver
Jan 30 2015 00:35
true
or even have a terminals group, so you could install ipython[notebook,terminals]
but I'm inclined towards a simple in/out for the notebook group
it's a fairly lightweight dependency if we do want to include it
Doug Blank
@dsblank
Jan 30 2015 01:35
@minrk Yes that bug has had a couple of us scratching our heads all week! I was sure I had a new kernel issue, and colleague that he had a cache issue... thanks!
Jason Grout
@jasongrout
Jan 30 2015 16:18
@takluyver - so that's how to take care of that "terminals not available" message. I was really puzzled.
Jason Grout
@jasongrout
Jan 30 2015 17:03
@rgbkrk - I just tried installing R from CRAN using their ubuntu package (http://cran.r-project.org/bin/linux/ubuntu/README), and then rzmq installed just fine.
Kyle Kelley
@rgbkrk
Jan 30 2015 17:04
ah hmmm
I could do that too then
Though it looks like we're switching to Debian soon
Min RK
@minrk
Jan 30 2015 17:50
@rgbkrk what do you think are going to be the implications of switching to Debian for things like the compmodels course?
Kyle Kelley
@rgbkrk
Jan 30 2015 17:51
scipystack isn't going to build
even before it hits the compmodels stage
I kicked it off last night before bed, noticed there are changes to make
As for what happens, I'm not sure how that effects user containers
Like, does that effect their home directory being swapped out
I'm ok with sticking with Ubuntu, it's pressure from others that makes me want to switch
Some problem will end up coming up with the Debian image too, no doubt
Jason Grout
@jasongrout
Jan 30 2015 17:56
centos rules! :)
Min RK
@minrk
Jan 30 2015 17:57
@rgbkrk should we hold the debian switch until after 3.0?
Kyle Kelley
@rgbkrk
Jan 30 2015 17:58
Yeah, let's shelve it. That would make me more comfortable.
Min RK
@minrk
Jan 30 2015 17:58
Easy enough.
Kyle Kelley
@rgbkrk
Jan 30 2015 17:58
I'll just use @jasongrout's suggestion for installing R from CRAN
Min RK
@minrk
Jan 30 2015 18:02
@takluyver we may want to summarize/link to PRs, Issues rather than including them all in sphinx. There are over 1200 issues and PRs in 3.0, and GitHub milestones make it an easy link to the same lists.
Thomas Kluyver
@takluyver
Jan 30 2015 18:06
Yeah, I think including the full list in the release notes is superfluous
Min RK
@minrk
Jan 30 2015 18:07
It wasn't when the list wasn't super long and the search to find the right ones was complicated, and couldn't be done via GitHub UI
but now it's a single URL to the milestone.
Thomas Kluyver
@takluyver
Jan 30 2015 18:11
yep
makes sense to me
though perhaps we should take a copy somewhere, in case those Github URLs change, or we one day move away from Github
Min RK
@minrk
Jan 30 2015 18:15
We can probably maintain a dump of our GitHub Issues and PRs somewhere
Jason Grout
@jasongrout
Jan 30 2015 18:15
I think gitlab imports from github :)
Jonathan Frederic
@jdfreder
Jan 30 2015 18:55
@jasongrout really? Or just jokingly?
Jason Grout
@jasongrout
Jan 30 2015 19:09
I'm pretty sure the newest release does.
so maybe it's still unsupported, but possible.
Min RK
@minrk
Jan 30 2015 21:29
@/all I'm about ready to cut 3.0b1
If anyone has a reason to hold it, speak up. I don't see anything on GitHub that suggests it should wait. We can do b2 next week.
Jonathan Frederic
@jdfreder
Jan 30 2015 21:31
yay!
Damian Avila
@damianavila
Jan 30 2015 21:32
yeah!!
Doug Blank
@dsblank
Jan 30 2015 22:06
Woot... wow...
Min RK
@minrk
Jan 30 2015 23:32
@/all 3.0b1 is out the door
Scott Sanderson
@ssanderson
Jan 30 2015 23:32
@minrk congrats!
Sylvain Corlay
@SylvainCorlay
Jan 30 2015 23:33
Bravo!
jdfreder @jdfreder frowns
Jonathan Frederic
@jdfreder
Jan 30 2015 23:38
another weird bug
Min RK
@minrk
Jan 30 2015 23:40
@jdfreder I think that's been reported - lost click events because of an invisible hscrollbar
Jonathan Frederic
@jdfreder
Jan 30 2015 23:41
Oh
That's probably what it is
I'll mark the issue I opened as a duplicate
@minrk all I could find is this:
ipython/ipython#4404
Sylvain Corlay
@SylvainCorlay
Jan 30 2015 23:46
Are you guys setting the current version to 4.0dev and backporting to certain things to 3 ?
Jonathan Frederic
@jdfreder
Jan 30 2015 23:47
Thanks @minrk
Min RK
@minrk
Jan 30 2015 23:47
@SylvainCorlay not yet
Jonathan Frederic
@jdfreder
Jan 30 2015 23:47
@SylvainCorlay 3.0 hasn't been released yet
just the beta
Min RK
@minrk
Jan 30 2015 23:47
not before we ship 3.0
Sylvain Corlay
@SylvainCorlay
Jan 30 2015 23:48
ok, thanks.
Min RK
@minrk
Jan 30 2015 23:48
and generally, we'll leave it at 3.1-dev until the first change that we don't plan to backport
that doesn't usually take long, though
Jason Grout
@jasongrout
Jan 30 2015 23:49
especially if splitting repos happens right after 3.0
I thought the plan was to go 3.0->4.0 very quickly
Min RK
@minrk
Jan 30 2015 23:49
it is
Sylvain Corlay
@SylvainCorlay
Jan 30 2015 23:49
Oh, do you want to split right away after?
Min RK
@minrk
Jan 30 2015 23:49
it's the first thing we plan to do for 4.0
there's a lot of planning to do before we can really get started, though.
it's not going to be immediate.
Sylvain Corlay
@SylvainCorlay
Jan 30 2015 23:50
just the time to merge a few PR s:)
Min RK
@minrk
Jan 30 2015 23:50
We have to change how some subpackages interact before we can start pulling them out
config and subcommands, in particular
how to deal with shared utils is going to be tricky, as well
Sylvain Corlay
@SylvainCorlay
Jan 30 2015 23:54
I bet that the mechanisms that you will use will be very useful for people like us who use a lot of different extension points.
Min RK
@minrk
Jan 30 2015 23:58
Either that, or we'll make a huge mess.
Let's hope for the first option!