Perl 6 - the future is here, just unevenly distributed

IRC log for #webwork, 2013-06-06

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

All times shown according to UTC.

Time Nick Message
00:00 goehle_ or compile it from source
00:00 mgage if it's only the // construct we can backout for now.  It's used in three files,  once in ContentGenerator at line 2016, once in ProblemUtil at line 93   and a few dozen times in Problem.pm to insure that variables are defined.
00:01 mgage I would like to start insisting on at least 5.10 soon however since it makes some of the modern perl features problem that bring perl into the 21st century
00:01 ionparticle we've been migrating to RHEL6 over the past few months for performance reasons, so it's not a problem here
00:02 mgage could you run a check on your RHEL5 machine and see if all of the errors seem to occur in one of the files I mentioned?
00:03 mgage maybe I'll create a special branch for for RHEL5 "laggards". :-)
00:03 goehle_ what about providing a third party rpm
00:03 goehle_ so they can upgrade their perl
00:03 goehle_ http://blog.tuxforge.com/tag/perl-5-10-rpm-rhel5/
00:04 ionparticle the first error was in ContentGenerator
00:05 mgage thanks. I'm reasonably sure it's limited to those three files. -- but of course it throws the parser completely off track
00:05 ionparticle Number found where operator expected at /home/john/webwork/stock_dev/webwork2/lib/WeBWorK/ContentGenerator.pm line 2039, near "# FIXME -- this gives the correct locations for release/2.7"
00:05 ionparticle (Might be a runaway multi-line // string starting on line 2036)
00:05 ionparticle (Missing operator before 2.7?)
00:05 ionparticle Bareword found where operator expected at /home/john/webwork/stock_dev/webwork2/lib/WeBWorK/ContentGenerator.pm line 2039, near "2.7 but"
00:05 ionparticle (Missing operator before but?)
00:05 ionparticle I'm guessing that it errored out near the line you mentioned but only realized it here
00:05 mgage yeah the error is actually at line 2016
00:05 ionparticle lol
00:05 goehle_ that seems familiar
00:06 mgage you could replace // by ||  or just leave it out (it will probably work and just throw a warning about XXX undefined
00:06 mgage the other place is ProblemUtil line 93
00:06 mgage after that there are several in Problem.pm
00:07 mgage the problem with the third party rpm is that many will have to convince their IT people to install it, which could be one more barrier to upgrade.
00:07 ionparticle ok, yep
00:07 mgage and it could actually cause conflicts if that same perl is being used by something else.
00:07 ionparticle replacing it with || now shows an error with ProblemUtil
00:07 ionparticle Possible unintended interpolation of @extra_answer_names in string at /home/john
00:07 ionparticle .pm line 93.
00:08 mgage yep
00:08 ionparticle Possible unintended interpolation of @extra_answer_names in string at /home/john
00:08 ionparticle .pm line 93.
00:08 mgage fix that and see if only errors in Problem.pm show up
00:08 ionparticle weird, it's cutting off the file names
00:09 mgage I thin the file name is ProblemUtil.pm
00:09 ionparticle sorry, meant my irc client
00:09 mgage oh I see.
00:10 ionparticle Possible unintended interpolation of @tableRows in string at /home/john/webwork/stock_dev/webwork2/lib/WeBWorK/ContentGenerator/Problem.pm line 237.
00:10 ionparticle Backslash found where operator expected at /home/john/webwork/stock_dev/webwork2/lib/WeBWorK/ContentGenerator/Problem.pm line 266, near "$answerMessage =~ s/\"
00:10 ionparticle (Might be a runaway multi-line // string starting on line 264)
00:10 ionparticle yeah, problem.pm next
00:11 mgage there are more lines with // in that file  -- you have to replace them carefully since you don't want to mess with
00:11 mgage s/ foo//;
00:12 mgage if it's just those three files I propose creating a "remedial" branch -- that might be enough to convince IT to upgrade eventually but it would work in the meantime.
00:12 mgage ionparticle: thanks for catching this for us.
00:12 ionparticle np
00:13 ionparticle "check_user_id" is not exported by the WeBWorK::DB module\nCan't continue after import errors at /home/john/webwork/stock_dev/webwork2/lib/WeBWorK/ContentGenerator/Instructor/UserList2.pm line 74
00:13 ionparticle this is on the develop branch, btw
00:14 mgage ok -- the develop branch might be a bit rougher than the release/2.7 branch right at the moment.
00:15 mgage seems a little weird however -- that is an old function
00:18 ionparticle yeah, not sure why it's doing that
00:18 ionparticle http://pastebin.com/gKNbdAVF
00:18 ionparticle diff so far
00:21 mgage thanks -- ur fast -- I was just starting
00:22 mgage that looks ok except for the $answerResult and $studentAnswer lines -- you might end up replacing 0 by '' which will often cause warnings -- usually not worse than that.
00:23 mgage but I'm trying to get something that doesn't clutter up the logs with spurious warnings.
00:24 mgage ||this.focus() really should be left as //this.focus() -- that's almost certainly inside a literal javaScript
00:35 ionparticle ouch
00:35 ionparticle I hope that's not common
00:36 mgage well it is in some other files -- that's why I'm not 100% sure I found every relevant instance of //
00:36 mgage goehle: still around?
00:38 Paul_Pearson @mgage My bug fix to Stats.pm did not make it into 2.7 yet.
00:38 Paul_Pearson I've found the two lines of code that need to be changed in Stats.pm
00:39 Paul_Pearson Git and github are vexing me.  Could you help?
00:39 mgage sure
00:39 Paul_Pearson What do I want to clone if I want to get the 2.7 version of webwork2?
00:40 aubreyja joined #webwork
00:40 Paul_Pearson that is, how should I modify this
00:40 Paul_Pearson git remote add webwork2-dev git@github.com:<your user name>/webwork2-dev.git
00:40 Paul_Pearson to work for me to get 2.7 (not dev)
00:40 Paul_Pearson sorry, I wanted the clone:
00:40 Paul_Pearson git clone git@github.com:<your user name>/webwork2-dev.git
00:41 mgage instructions at  http://webwork.maa.org/wiki/Github#Quick_instructions_for_using_other_branches_of_webwork
00:41 Paul_Pearson OK, give me a minute to read that.
00:42 mgage also are you moving something from openwebwork to your desktop machine or are you moving it to your github repo?
00:42 mgage note that webwork2-dev has gone away and been replaced by webwork2
00:45 Paul_Pearson I know that I do not want to use webwork2-dev.  The only reason I mentioned webwork2-dev is that it was mentioned in the README.md file at the location https://github.com/openwebwork/webwork2/tree/release/2.7
00:46 Paul_Pearson Should the README.md file have instructions that point to http://webwork.maa.org/wiki/Github#Quick_instructions_for_using_other_branches_of_webwork ?
00:46 mgage yes -- that's actually on my list of things to change. -- that was the original read me from webwork2-dev
00:48 mgage you don't need to change it yet -- it should probably point to parts of Github that haven't been rewritten yet -- that are for developers rather than those just trying to download various versions of webwork.
00:49 Paul_Pearson I'm not sure where I want to move things.  I think the workflow could be either of these:
00:50 Paul_Pearson 1. openwebwork 2.7 -- fork --> my github
00:50 Paul_Pearson 2. openwebwork 2.7 -- clone --> my hard drive
00:51 mgage the return has to be    my github --> github/openwebwork
00:51 mgage so do 1.
00:51 mgage then clone  github/paultpearson  or whatever it is
00:52 mgage when you are done you can push back to your github and issue a pull request from there.
01:09 mgage I see 4 lines that you have changed in Stats.pm
01:09 mgage Paul_Pearson:
01:10 Paul_Pearson line 564 should be
01:10 Paul_Pearson $problabelxpixel = $leftmargin + ($probID-1) * $totalbarwidth + $barsep + sprintf("%d",$totalbarwidth/2);
01:11 Paul_Pearson I thought I knew what line 565 should be, but now I'm not sure.  That's one of the hang-ups.
01:11 mgage go here https://github.com/openwebwork/webwork2/blame/release/2.7/lib/WeBWorK/ContentGenerator/Instructor/Stats.pm
01:12 Paul_Pearson Anyway lines 564 and 565 are the ones that need to be fixed, and I need to determine what line 565 needs to be.
01:12 ionparticle mgage, sorry, can't seem to find why it doesn't like check_user_id
01:13 ionparticle it honestly makes no sense
01:14 mgage which file was that called from again?
01:15 ionparticle lib/WeBWorK/ContentGenerator/Instructor/UserList2.pm
01:15 mgage kk
01:16 mgage Paul_Pearson:  I think you'll have to figure those two lines out -- as far as I can tell those are the original lines that were written last august.  -- maybe compare them against the version in Stats_old.pm?
01:18 mgage and the check_user_id is in the develop branch? right?
01:18 mgage ionparticle:
01:20 ionparticle ?
01:21 ionparticle oh
01:21 ionparticle yes
01:21 ionparticle it's in DB.pm
01:21 ionparticle I checked the history for both UserList2.pm and DB.pm
01:21 ionparticle they apparently haven't been touched for 2 months
01:22 ionparticle but it looks like they might have been recently merged from release/2.7
01:23 ionparticle err, I mean, the last commit for them was 2 months ago, but I'm wondering if those commits were only brought in recently
01:27 mgage so peter staab has been working on the develop branch and Homework Manager --- he might have kicked something that caused trouble with UserList2.
01:27 mgage I can get the develop version on openwebwork to compile for me.
01:27 mgage It is a little weird because DB.pm doesn't seem to have the Exporter module or the @EXPORT_OK array of
01:28 ionparticle there's another // to || replacement at line 320 of WeBWorK.pm
01:28 ionparticle I removed the import
01:28 ionparticle and now it works
01:28 mgage things that it is alright to export from the module.  (see Utils.pm to see what I'm talking about) -- why it would work for me and not on your machine is a mystery.
01:28 ionparticle the server boots up at least
01:29 mgage ok
01:29 whytheplatypus joined #webwork
01:29 ionparticle the theme looks really broken on my end
01:29 ionparticle lol
01:30 mgage nice catch for the // -- I hope that is the last one.
01:31 ionparticle there might be another 5.8.8 incompat error being hidden by the export problem
01:31 mgage I won't guarantee the themes in the develop branch.   they should be ok in the release/2.7 branch -- we've been tweaking them all day :-)
01:31 ionparticle but yeah, that should be it
01:31 ionparticle lol
01:32 mgage but we haven't moved those changes into the develop branch yet.
01:32 ionparticle I'm going home for the day, will check them out tomorrow
01:33 mgage ok -- I'll wait for you to finish and then check over your pull request -- no point in the two of us duplicating work -- and you can check that the changes are enough.
01:33 mgage thanks for the help
01:35 mgage one thing I'm worried about -- it would be best if we could put these changes into release/2.7   -- to do that you'll need to pull a copy of the current version of release/2.7 in openwebwork,  make changes there to get rid of the //   and then issue the pull request to branch release/2.7.
01:36 mgage if you try to send the pull request built on development to release/2.7  it won't work because of all of the  other differences between release/2.7 and evelop
01:47 Paul_Pearson mgage I submitted a pull request (I think) to release/2.7
01:54 mgage joined #webwork
01:54 mgage kk -- just a sec.
01:55 mgage yep.  that's perfect.
01:55 mgage it's all merged now
01:56 Paul_Pearson :) :) :) :) :)
02:40 mgage joined #webwork
02:40 mgage afk
02:45 goehle_ left #webwork
03:46 mgage joined #webwork
11:59 aubreyja joined #webwork
13:46 goehle joined #webwork
14:04 mgage joined #webwork
15:17 Paul_Pearson joined #webwork
16:52 mgage joined #webwork
17:45 mgage joined #webwork
20:26 Paul_Pearson joined #webwork

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