Perl 6 - the future is here, just unevenly distributed

IRC log for #webwork, 2013-11-25

| Channels | #webwork index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:10 rbeezer joined #webwork
14:01 goehle joined #webwork
15:29 goehle hey mgage
15:56 aubreyja joined #webwork
15:56 aubreyja joined #webwork
16:06 aubreyja mgage - around?
18:12 mgage hi
18:19 mgage goehle: aubreyja: hi
18:19 goehle hey mgage
18:20 goehle I figured out my issue
18:20 goehle I fixed the ** pull, btw
18:20 mgage that was easy :-)
18:21 goehle anything that needs working on now?
18:25 mgage would you be interested in taking a look at bug 2922? submitted by alex jordan   -- this is a really long time behavior which is just not correct.  -- when you create a new version you are still pointed at the old file to edit.  -- everyone expects that you would be pointed at the new file.  It's been like this forever and I don't know if it's worth fixing now with a new interface coming -- but ....
18:26 mgage the place to look for the behavior is in the preheater initialize subroutine of the code for the PGProblem editor.  It's fairly hairy because -- particularly at the time -- we weren't' sure what kind of file you were editing.  Is it a header file? a problem file in a se? a problem file in a library , a problem file not in a set, or a tex header file?    …..
18:27 mgage never mind -- I see you just fixed it
18:27 mgage fast work :-)
18:28 goehle :D
18:29 goehle I pushed my branch to the wrong git though
18:29 goehle and am having trouble deleting it
18:29 goehle um
18:29 mgage I'll need some help with peter's stuff once he gets the last stage submitted
18:29 goehle I didn't do anything super intelligent
18:29 goehle if it was using that option then it was saving the file as the last option in the set
18:29 goehle *problem
18:30 goehle so I had the redirect point to the last problem
18:31 mgage that sounds right.  However check the behavior for the other two options -- replacing the current problem in the set, and creating a new independent file -- I _think_ that it still points to the original file after those saves.  This might be ok in the case of creating a new independent file -- not sure about that -- but good visual cues as to where you are would be welcome.
18:32 goehle ok
18:32 goehle so new version and replace seems tow ork
18:34 goehle append now switched to the newly created file
18:34 goehle and so does the unnattached option
18:34 goehle the append *tab* still has wierd behavior
18:34 mgage ok -- sounds good then
18:35 rbeezer joined #webwork
18:35 goehle both of the otehr two options pointed to the new file
18:35 goehle or rather, the redirect points them to the appropriate problem, which has the new file stored in the db
18:36 goehle I cant seem to delete pgediting from the openwebwork git
18:39 mgage hmm -- did you click on the "branches" tab in webwork2?
18:39 goehle yeah
18:39 goehle its not there
18:39 goehle but it is listed in the dropdown on the "main" page
18:40 mgage ok -- I see the same thing
18:42 mgage it also says there are 5 branches -- I wonder if this is an update thing and github will catch up in bit?
18:42 goehle maybe
18:42 goehle usually its pretty fast
18:42 goehle and its been over 10 min
18:43 mgage yeah -- I don't know
18:43 mgage if it doesn't go away we can ask github gods
18:44 goehle well, it will go away if you pull 246
18:44 goehle or at least it will become a "merged" branch
18:44 goehle and will be more or less invisible
18:45 mgage ok -- I'll check it out and then pull 246 -- might take an hour or so -- I'm helping peter get things ready.
18:46 goehle no worries
18:46 goehle take your time
18:46 goehle i'll test the peter pull if you want
18:47 goehle my hope is that it will run on a "fresh" pull of develop
18:47 goehle with no further modification
18:47 goehle ah
18:47 goehle figured out how to delete it via command line
19:11 mgage goehle:  I'm wondering whether the timezone validation that JJ worked on can be easily submitted to 2.8 (separately from the other stuff) -- it strikes me as something that would reduce calls for help
19:11 goehle probably
19:11 goehle what calls for help would it reduce though?
19:12 mgage it doesn't look that complicated
19:12 goehle as far as I can do it just validates timezones that go in course config
19:12 goehle but for the vast majority of users they won't ever change that field
19:12 goehle it will be set by the site admin as a default
19:13 goehle speaking of that though
19:13 mgage yeah but we get a call every month or so from someone who has locked themselves out by entering an invalid timezone
19:13 goehle in that case it should be implementable
19:14 goehle my other question is: why do we even show timezones on things like homework sets editor and set detail
19:14 mgage the problem is that if it's wrong it makes the entire process die and you can't log in to fix it from the web
19:15 mgage that's a reasonable question -- actually I think they should be shown, but I don't see why they should be editable
19:15 goehle well it just causes confusion with the daylight savings stuff
19:15 mgage I'm sure this was done on the first pass just for convenience.
19:15 goehle I looked into making that check automatically, but its too complicated
19:16 goehle would you want jj to backport the timezone validation stuff
19:16 goehle that would be best because he wrote it.  But he also says he is busy
19:16 goehle actually wait a sec
19:16 mgage it didn't look that bad and you said you had some time -- I don't think it's at all urgent but it would be useful
19:16 goehle the pull request says that timezone validation wasl aready backported
19:17 goehle hmm
19:17 goehle doesnt' seem like it is to me though
19:20 goehle well that was easy
19:20 goehle yet another pull for you to look at
19:20 goehle I feel like I'm creating more work than I'm doing !
19:22 mgage np -- I've been getting caught up -- at least it's not too hard to pull these small ones -- I spent quite a bit of time over the weekend trying to pull Peter's -- and I really should pull Wheeler's as well -- that's not too massive but it's all intertwined
19:22 mgage I'll get to yours first.
19:23 mgage sometimes when I change course configuration from the web I get a bunch of error messages of variables being undefined -- does that happen to you?  I have always been busy and haven't documented it yet as a bug.
19:29 goehle it only happened to me when I was testing jwj's pull request
19:29 goehle but I have seen it before
19:31 goehle I was thinking it might be a good time to merge 2.8 into develop
19:32 goehle also people are too agressive about moving stuff to legacy
19:33 goehle a lot of that is still iln active use and not being replaced any time soon
19:34 goehle hmm
19:34 goehle are we using //
19:34 goehle or not?
19:36 goehle also also, I am getting that config uninitialized error on devel
19:37 goehle actually nm
19:37 goehle wait no... getting the error
19:37 goehle if I save changes in devleop
19:38 goehle as for the merge of 2.8 and develop, I am now thinking it would be best to wait until we get peters stuff in there
19:44 mgage yes -- I've been waiting for that as well.
19:54 mgage goehle: small question (and only because you asked whether things need doing) -- do we still need the Complex1.pm file in the pg/lib?  If so to what extent?  are there many places where perl complex numbers are being used instead of MathObject versions?
19:54 goehle I honestly have no idea
19:55 goehle my grasp of pg is a lot weaker than of core webwork
19:55 mgage it it's not being used then it will be less confusing  if it is not there
19:56 mgage -- I'm not sure either -- and it will take a while to hunt it down.  I use whole directory searches for this kind of thing ---- but I noticed the file there a few days ago and was thinking that it was time to start pruning files that really aren't being used
19:58 mgage another thing that needs work is to research the error reporting path -- particularly for errors reported from within pg.  I think it's been hacked a few times and is pretty squirrely by now.  I would like to improve the error reporting for problem authors -- particularly to those that aren't used to output from a perl compiler :-)
19:59 mgage on the easier side I have a pull request into pg   -- involving askSage -- that would probably be fun to look at -- Jason Grout made several suggestions and I've tried to incorporate them.
19:59 mgage I can send you a test homework set if you like
19:59 mgage afk
20:00 mgage brb in 10 minutes
20:02 goehle kk.  I've got some other stuff to do atm (but if you send me an email I can put it on the list)
20:03 pstaab joined #webwork
20:03 pstaab mgage: if you're here, I figure it'll be quicker to chat this way.
20:10 mgage kk -- I'm back
20:12 mgage pstaab: I'm back
20:13 pstaab I'm trying to pull down bits of js libraries right now.  Then you can pull in a smaller pull request
20:15 mgage before you do too much work -- just create a new pull request (you'll need to fork a new branch in order not to add to the old pull request) and see if it looks different from the "more refined new UI-tools" pull request.   The comparison with the existing develop branch might be different
20:16 mgage if that doesn't work then we can try making the js changes in an advance  pull request and then submitting a pull request for the more substantial changes.
20:17 mgage I think that will get this unblocked.
20:22 pstaab almost have something for you.
20:31 pstaab mgage: I just added a couple of js libraries to stage3 and sent another pull request.
20:38 mgage kk -- just a sec and I'll pull it -- this changes 118 files -- can you see those stats when you compare and prepare for a pull request? or can I only see them once you have submitted?
20:42 mgage kk -- it's merged now -- see if other commits now have fewer file changes
20:43 mgage ah -- yes they do -- you don't even need to submit a new request -- now "more refined new UI-tools" only changes 65 files -- that will be much easier to check
20:43 mgage I'll pull that onto my laptop as soon as I get home and do some testing
20:49 pstaab mpage:  I just did a stage4 pull request that is the big commit, but hopefully this will be smaller.  If so, then merging pstaabp/develop shouldn't be too bad.
20:58 mgage stage 4 has only 2 file changes -- can you see that statistic when you submit?
21:03 pstaab for stage4 I select the commit that was a big problem.  I wonder if we've sucked up the other file changes elsewhere.
21:05 mgage hopefully yes -- you can see the commits for each pull when you look on github -- I think you can see those pages -- particularly after they are submitted or committed. (look under the closed pull request tab)
21:06 pstaab Is there an easy way to tell the difference between pstaabp/develop and pstaabp/stage4?
21:07 mgage one pull request is labeled "stage4" and the other "more refined new UI-tools"
21:09 pstaab I see now.  It still looks like the libraries in stage4 aren't in sync with those in pstaabp/develop
21:18 pstaab mgage:  I tried one more stage to pull.  I'm not seeing where the big differences are.
21:19 pstaab Thanks for being patient.  I'm learning the ins and outs of git and source tree...
21:26 mgage pstaab: how many files changed to you see for this stage 5 pull request?
21:26 pstaab looks like 55 file changes.
21:27 pstaab That's from github
21:27 mgage that's what I see too -- do you see why there are 10 more file changes in "more refined new UI-tools"?
21:28 mgage yep -- it's pretty handy -- I think it will actually give you that statistic when you start to do the pull request with the initial compare -- before you even submit the request
21:29 pstaab They aren't at exactly the same point in time.  There a few commits apart.  I was trying to find where all the js libraries were going to come in, but I really can't find that.  They may be in already.
21:30 mgage if you go to the closed pull requests you'll see that a lot of them came in with "updated backbone-validation"
21:31 mgage there were even more in stage3
21:33 pstaab Now I see.  Good.  All those changes weren't in a single commit.  It just looked like it because the commit file with the merge contained all the changes in the merge, but the commits that included those js libraries updates were before that.
21:57 mgage headed home -- I'll check the pull requests on my laptop and then pull them -- hopefully later this evening if all goes smoothly
21:57 pstaab Thanks Mike.

| Channels | #webwork index | Today | | Search | Google Search | Plain-Text | summary