Camelia, the Perl 6 bug

IRC log for #padre, 2012-01-01

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

All times shown according to UTC.

Time Nick Message
00:00 bowtie $year++ :)
00:10 bowtie left #padre
00:59 asarch joined #padre
04:16 * xiong is afraid to increment again, lest skip a whole year
04:17 xiong <8*D
07:07 Hyppolit svn: r17790 | adamk++ | http://padre.perlide.org/trac/changeset/17790
07:07 Hyppolit Completed the migration of the document statistics dialog from hand-crafting to FBP. Removed the refresh button, since it is now modal.
07:07 Hyppolit trunk/Padre/ trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Wx/Dialog/ trunk/Padre/lib/Padre/Wx/FBP/
07:11 Hyppolit svn: r17791 | adamk++ | http://padre.perlide.org/trac/changeset/17791
07:11 Hyppolit Add ->
07:11 Hyppolit trunk/Padre/lib/Padre/
08:18 Hyppolit svn: r17792 | szabgab++ | http://padre.perlide.org/trac/changeset/17792
08:18 Hyppolit Update the messages.pot file
08:18 Hyppolit trunk/Padre/share/locale/
08:53 dodathome joined #padre
09:50 kaare joined #padre
10:03 kaare joined #padre
11:30 bowtie joined #padre
11:41 mj41 joined #padre
12:20 sugar joined #padre
13:54 Hyppolit svn: r17793 | adamk++ | http://padre.perlide.org/trac/changeset/17793
13:54 Hyppolit The activation refresh should be on the entire app, not just the main window.
13:54 Hyppolit trunk/Padre/lib/Padre/Wx/
13:59 Hyppolit svn: r17794 | adamk++ | http://padre.perlide.org/trac/changeset/17794
13:59 Hyppolit Switching back to normal activate
13:59 Hyppolit trunk/Padre/lib/Padre/Wx/
14:03 kaare_ joined #padre
14:53 Hyppolit svn: r17795 | adamk++ | http://padre.perlide.org/trac/changeset/17795
14:53 Hyppolit Migrating more things to using Padre::Feature
14:53 Hyppolit trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Wx/ trunk/Padre/lib/Padre/Wx/Menu/
14:58 bowtie adamk any thoughts about #1369
14:58 Hyppolit # 1369 :  ToDo Panel Crashes Padre (new defect) [ http://padre.perlide.org/trac/ticket/1369 ]
15:02 Hyppolit # 1324 :  One of our features is missing...   Skelton Build or Module::Starter (assigned defect) by bowtie  [ http://padre.perlide.org/t​rac/ticket/1324#comment:5 ]
15:09 Hyppolit # 1368 :  Debug::Client tests fail against perl-5.15.6 (2 of n) (closed defect) by bowtie  [ http://padre.perlide.org/t​rac/ticket/1368#comment:1 ]
15:09 Hyppolit # 1367 :  Debug::Client tests fail against perl-5.15.6 (1 of n) (closed defect) by bowtie  [ http://padre.perlide.org/t​rac/ticket/1367#comment:3 ]
15:17 Hyppolit # 1370 :  Debug::Client, @list request fails against perl-5.15.6 (3 of n) (new defect) [ http://padre.perlide.org/trac/ticket/1370 ]
15:18 Hyppolit # 1370 :  Debug::Client, @list request fails against perl-5.15.6 (3 of n) (new defect) by bowtie  [ http://padre.perlide.org/t​rac/ticket/1370#comment:1 ]
15:49 rindolf joined #padre
16:18 Hyppolit svn: r17796 | szabgab++ | http://padre.perlide.org/trac/changeset/17796
16:18 Hyppolit Update the messages.pot file
16:18 Hyppolit trunk/Padre/share/locale/
16:34 pece joined #padre
16:35 kaare joined #padre
16:47 szabgab bowtie: maybe you could cooperate with rindolf on the Debugger?
16:47 szabgab http://blogs.perl.org/users/shlomi_fish/2011/12/b​reaking-the-perl-debugger-for-fun-and-profit.html
16:47 rindolf bowtie: hi.
16:47 szabgab the Debug::client (or part of it) could serve as the test cases needed for the debugger
16:47 rindolf szabgab: hi.
16:47 szabgab hi
16:53 bowtie good afternoon
16:56 bowtie rindolf, ask away
16:56 bowtie szabgab, more info please
16:56 rindolf bowtie: what should I ask?
16:57 rindolf bowtie: szabgab wants me to help you with the debugger.
16:57 bowtie rindolf, i don't know
16:58 bowtie have you seen Debug-Client 0.17_04 and #1367 and #1368
16:58 Hyppolit # 1367 :  Debug::Client tests fail against perl-5.15.6 (1 of n) (closed defect) by bowtie  [ http://padre.perlide.org/t​rac/ticket/1367#comment:3 ]
16:58 szabgab rindolf wrote more tests are needed for the debugger (see my link above)
16:58 szabgab and the Debug::Client is testing at least some aspect of the debugger API
16:59 rindolf bowtie: yes, "perl -d" was temporarily broken.
16:59 szabgab so maybe the Debug::Client should be the test suit of perl -d
16:59 rindolf It should be OK now, but there may still be some bugs.
16:59 szabgab and should maybe even becomre core...
16:59 szabgab maybe after 5.16
16:59 szabgab core in perl I mean!
16:59 rindolf I would prefer against people interfacing directly against "perl -d" and instead using a custom "Devel::" front-end.
17:00 szabgab we need to be able to talk to it via a socket
17:00 bowtie yes I have bean reading the perl deltas but they don't seam to be complete as far as perl5db.pl goes
17:00 szabgab so Padre will be able to debug remote scripts as well
17:02 bowtie szabgab, how do you suggest that the correct script will be displayed in Padre given that absolute filenames are used?
17:03 szabgab don't know
17:03 szabgab but I am off to dinner
17:03 bowtie ok
17:04 mj41 joined #padre
17:05 bowtie rindolf, ok, so what would you like to see changed with current debug-client tests?
17:09 asarch joined #padre
17:13 rindolf bowtie: I'm not familiar with Debug-Client.
17:16 bowtie rindolf, is szabgab suggesting you write a test sweet for perl5db.pl?
17:17 rindolf bowtie: s/sweet/suite/
17:17 rindolf bowtie: I've started writing one.
17:17 bowtie thanks
17:17 rindolf bowtie: well, there were some rudimentary ones, but I've written more.
17:17 rindolf bowtie: Gabor wants to utilise the Debug-Client tests to test "perl -d" too.
17:18 bowtie rindolf, where do you find the perl5db.pl changes and version number changes?
17:18 rindolf bowtie: I'd like to define a small API for writing tests for "perl -d".
17:18 rindolf bowtie: right now there's quite a lot of duplicate code.
17:18 rindolf bowtie: what do you mean?
17:19 rindolf bowtie: it's in the perl 5 git.
17:19 bowtie rindolf, ok I see where you are going, now
17:21 bowtie rindolf, I have been removing tests from Debug-Client as it's tests are to test it. not perl debugger, but I was wondering how to test perl -d
17:21 bowtie I did not know it was testless though :)
17:21 rindolf bowtie: look at lib/perl5db.t.
17:22 bowtie rindolf, is there a wiki or blog with all changes and version info for all version of perl5db.pl
17:23 rindolf bowtie: it's not completely testless.
17:24 rindolf bowtie: anyway, some of the complaints against Regexp::Common was that its test suite (> 100K assertions strong) sometimes tested the perl regex engine instead of the generated regular expressions.
17:24 rindolf bowtie: it's in the git repository.
17:24 rindolf bowtie: or should be.
17:24 rindolf bowtie: I've blogged about the recent changes, but there were earlier ones.
17:25 bowtie url please, do I have to use git?
17:25 rindolf And there are some changes that are scheduled for after 5.16.0 is out.
17:26 bowtie which are?
17:26 rindolf bowtie: http://blogs.perl.org/users/shlomi_fish/2011/12/b​reaking-the-perl-debugger-for-fun-and-profit.html - knock yourself out.
17:27 bowtie rereading now
17:27 rindolf bowtie: well, I've written a series of patches that add "use strict;" to the perl debugger.
17:27 rindolf I'm contemplating adding "use warnings;" too.
17:27 bowtie cool
17:27 rindolf And I also want to refactor t/perl5db.t by defining an OOP API.
17:28 bowtie rindolf, if perl 5.14 forces strict, is this not implemented in debugger?
17:29 rindolf bowtie: it only forces strict if you do "use 5.014;"
17:30 rindolf Which lib/perl5db.pl does not do.
17:32 bowtie rindolf, thanks for pointing out perl5db.t, is that all there is by perl for testing perl5db.pl?
17:33 rindolf bowtie: yes, but there's some more tests in bleadperl.
17:34 rindolf bowtie: in the same file.
17:39 bowtie rindolf, is there a wiki or blog page just for perl debugger with relevant info collocated in a single location, with out the normal perl terse wording so it can just be read
17:40 rindolf bowtie: normal perl "terse wording"?
17:40 rindolf bowtie: what are you looking for exactly?
17:41 rindolf bowtie: are you looking for a manual?
17:41 rindolf bowtie: for a book?
17:41 rindolf http://perl-begin.org/topics/debugging/ - there are the resources here.
17:49 kaare joined #padre
17:50 bowtie rindolf, my point exactly, only 5.14.2 where is the info for 5.15.x?
17:50 rindolf bowtie: there's also http://debugger.perl.org/
17:53 rindolf bowtie: what?
17:53 rindolf bowtie: the info for perl-5.15.x is in the perl 5 git repository.
17:53 rindolf bowtie: you may be able to interface with it using a different version control system.
17:53 rindolf Or download tarballs or use the github interface.
17:53 rindolf TIMTOWTDY
17:53 rindolf TIMTOWTDI
17:53 bowtie rindolf, but as I said earlier where is the detailed info for 5.15.x which is eluded to in perldeltas?
17:54 rindolf bowtie: in the perl -d documentation.
17:54 rindolf bowtie: what are you looking for, exactly?
17:56 bowtie rindolf, changes to perl5db.pl with respect to perl version number and perl5db.pl version number to start with
17:56 rindolf a changelog?
17:57 rindolf bowtie: use git for htat.
17:57 rindolf bowtie: use git for that*.
17:57 bowtie rindolf, and some, the POD dose not cover that
17:58 bowtie git at what url
17:58 rindolf bowtie: first clone into a local repository.
17:58 rindolf bowtie: then filter the log based on lib/perl5db.pl .
17:58 bowtie rindolf, first I don't use git
18:00 rindolf I'm not sure about the exact syntax, but there's #git on Freenode, #p5p here, and Google search is also useful.
18:00 rindolf And if those fails - there's Stack Overflow.
18:00 rindolf bowtie: well, you can start using git.
18:01 rindolf bowtie: I don't like git very much, but I still use it when I need to.
18:04 bowtie rindolf, you have missed the point, the current documentation for perl5db.pl v5.15.x is naff, as to the changes, logic etc it's all missing, there should be a perldoc for 5.15.x perl5db with changes and implications
18:09 rindolf bowtie: well, I'm not sure who wrote the perldelta.
18:09 rindolf bowtie: I think I contributed some entries for it too.
18:09 asarch joined #padre
18:09 rindolf bowtie: why are you unhappy with the perldelta?
18:11 kaare joined #padre
18:11 bowtie rindolf, the only way to decide what a perl -d can or will do is to use it's version, but there is no feature table to reference, hence we mere mortals are blindfolded :)
18:14 rindolf bowtie: like I said, IMNSHO, you should not parse and process the "perl -d" output because it's not easily parsable, and is primarily meant for end-users. Instead, define or reuse a Devel:: API module and use that - there's Devel::ebug and Devel::Trepan and more could be written.
18:16 rindolf Maybe given enough refactoring, "perl -d" will have a stable API, but it's not going to happen soon.
18:17 bowtie rindolf, ok this is new to me, can you explain the Devel:: way, please
18:19 rindolf bowtie: if you do perl -d:Trace then Perl will look for Devel::Trace and debug Perl using it.
18:19 rindolf Instead of lib/perl5db.pl.
18:19 rindolf You can put lib/perl5db.pl in Devel::Foo and just do perl -d:Foo
18:26 bowtie rindolf, to vague more info required
18:34 rindolf bowtie: too vague?
18:35 rindolf bowtie: perl defines a debugger interface and you can write alternative debuggers.
18:35 rindolf bowtie: which is specified using perl -d
18:36 rindolf So there's Devel::Trace
18:36 rindolf And Devel::NYTProf
18:36 rindolf And Devel::Cover
18:36 rindolf And Devel::LineTrace
18:36 rindolf They all do different things.
18:36 rindolf And there are also "perl -d" alternatives written this way.
18:38 bowtie rindolf, ok I have used Devel::Cover
18:39 bowtie but what have they got to do with perl -d ?
18:40 rindolf bowtie: they use the same interface.
18:40 rindolf bowtie: you can do perl -d:Cover to run it using Devel::Cover.
18:41 bowtie instead of cover -test
18:42 rindolf The only thing special about perl -d without any :$SOMETHING following it is that it runs stuff using lib/perl5db.pl which is just a normal perl code.
18:42 rindolf Nothing particularly special about it.
18:43 bowtie so why do you want to do Devel::
19:00 rindolf bowtie: because then you can provide a structure API instead of having to process the output of "perl -d".
19:01 rindolf Which is useful for writing a GUI debugger.
19:02 bowtie rindolf, ok, got that, but how?
19:02 rindolf bowtie: look at Devel::ebug and Devel::Trepan.
19:04 bowtie rindolf, this is what you are intending to write?
19:04 bowtie will look at examples later
19:04 rindolf bowtie: no, I don't intend to write that.
19:04 rindolf bowtie: I'm not really a Padre contributor.
19:05 rindolf bowtie: well, I am to an extent, but I'm still using Vim for Perl development.
19:06 bowtie rindolf, ok
19:07 bowtie rindolf, any thoughts about http://padre.perlide.org/trac/wiki/F​eatures/Perl5Debugger#Debug::Client
19:09 rindolf bowtie: can you translate it to a t/perl5db.t test?
19:10 bowtie why?
19:10 rindolf bowtie: so I can fix the breakage.
19:11 rindolf Also, did you try with bleadperl? There are one or two fixes there.
19:11 bowtie rindolf, breakage?
19:12 rindolf bowtie: the regression.
19:12 rindolf bowtie: what was once fine, and now is broken.
19:12 bowtie do not know about bleadperl what is it?
19:12 rindolf bowtie: http://dev.perl.org/perl5/docs/p5p-faq.html
19:13 rindolf bowtie: it's like perl's "svn trunk".
19:15 bowtie rindolf, I am quite happy, just need to fix debug::client due to the naff give me @list calls, rather than do action then ask where am I
19:15 rindolf bowtie: what? I can't parse your sentence.
19:15 bowtie rindolf, I am quite happy,
19:15 rindolf bowtie: what is "naff"?
19:15 bowtie just need to fix debug::client
19:15 bowtie due to the naff give me @list calls,
19:16 bowtie rather than do action then ask where am I
19:17 bowtie see #1367
19:17 Hyppolit # 1367 :  Debug::Client tests fail against perl-5.15.6 (1 of n) (closed defect) by bowtie  [ http://padre.perlide.org/t​rac/ticket/1367#comment:3 ]
19:18 bowtie also #1368
19:18 Hyppolit # 1368 :  Debug::Client tests fail against perl-5.15.6 (2 of n) (closed defect) by bowtie  [ http://padre.perlide.org/t​rac/ticket/1368#comment:1 ]
19:43 rindolf joined #padre
20:11 mj41 joined #padre
21:53 Hyppolit # 1371 :  [0.92 osx] checkmarks for case sensitivity, etc. cannot be toggled in Replace window (new defect) [ http://padre.perlide.org/trac/ticket/1371 ]
23:58 Hyppolit svn: r17797 | adamk++ | http://padre.perlide.org/trac/changeset/17797
23:58 Hyppolit Removed $DB::single
23:58 Hyppolit trunk/Padre/lib/Padre/Wx/Dialog/

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