Camelia, the Perl 6 bug

IRC log for #padre, 2009-12-01

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

All times shown according to UTC.

Time Nick Message
00:17 teejay joined #padre
00:23 Hyppolit svn: r9348 | adamk++ | http://padre.perlide.org/trac/changeset/9348
00:23 Hyppolit Syncing my laptop
00:23 Hyppolit trunk/Padre-Plugin-wxGlade/lib/Padre/Plugin/
01:15 cognominal joined #padre
02:49 Hyppolit svn: r9349 | adamk++ | http://padre.perlide.org/trac/changeset/9349
02:49 Hyppolit Added missing File::Slurp dependency reported at OSDC::AU
02:49 Hyppolit trunk/Padre/
03:32 Hyppolit svn: r9350 | adamk++ | http://padre.perlide.org/trac/changeset/9350
03:32 Hyppolit Noting File::Slurp
03:32 Hyppolit trunk/Padre/
03:34 Hyppolit svn: r9351 | adamk++ | http://padre.perlide.org/trac/changeset/9351
03:34 Hyppolit Removed superfluous function import
03:34 Hyppolit trunk/Padre/lib/Padre/Wx/
03:41 Hyppolit svn: r9352 | adamk++ | http://padre.perlide.org/trac/changeset/9352
03:41 Hyppolit Adding sample Glade files for use in testing
03:41 Hyppolit trunk/Padre-Plugin-wxGlade/t/ trunk/Padre-Plugin-wxGlade/t/sample/
04:21 Alias joined #padre
04:21 Alias Righto
04:21 Alias Time to get back into Padre properly
04:28 Hyppolit svn: r9353 | adamk++ | http://padre.perlide.org/trac/changeset/9353
04:28 Hyppolit Adding a skeleton for a WXG file class
04:28 Hyppolit trunk/Padre-Plugin-wxGlade/ trunk/Padre-Plugin-wxGlade/lib/Padre/Plugin/ trunk/Padre-Plugin-wxGlade​/lib/Padre/Plugin/wxGlade/ trunk/Padre-Plugin-wxGlade/t/
04:29 Hyppolit svn: r9354 | adamk++ | http://padre.perlide.org/trac/changeset/9354
04:29 Hyppolit Test rename
04:29 Hyppolit trunk/Padre-Plugin-wxGlade/t/
06:19 holli_ joined #padre
06:19 Hyppolit svn: r9355 | adamk++ | http://padre.perlide.org/trac/changeset/9355
06:19 Hyppolit Completed the WXG file object, but not entirely sure if I'm going to need it yet. Better to just implement it while everything is fresh in my head.
06:19 Hyppolit trunk/Padre-Plugin-wxGlade/ trunk/Padre-Plugin-wxGlade​/lib/Padre/Plugin/wxGlade/ trunk/Padre-Plugin-wxGlade/t/ trunk/Padre-Plugin-wxGlade/t/sample/
06:21 Hyppolit svn: r9356 | adamk++ | http://padre.perlide.org/trac/changeset/9356
06:21 Hyppolit Remove the crazy file magic, just do it literally
06:21 Hyppolit trunk/Padre-Plugin-wxGlade​/lib/Padre/Plugin/wxGlade/
06:22 waxhead joined #padre
06:22 Sewi joined #padre
06:22 Sewi gm
06:23 Alias heya
06:23 * Alias is solving Big Problem today
06:23 Alias Integration between wxGlade and Padre
06:24 Sewi Have fun
06:24 Alias I've got most of the heavy lifting done so you can point Padre at wxGlade-generated code, and it will suck out the dialog structure and convert it into Padre-compatible dialog class
06:24 Alias So you can do visual design of GUI dialogs and such, and then just import them
06:25 Alias It's not as good as having a NATIVE Padre visual designer, but it's one step closer
06:27 szabgab gm
06:27 Alias Heya
06:27 Sewi Hi szabgab
06:28 szabgab I was thinking about turning padre into that platform for GUI applications
06:28 Alias yup
06:28 Alias PTK :)
06:28 szabgab I don't know if it is reasonable
06:28 Alias Padre Toolkit
06:28 Alias I think it's reasonable
06:28 Alias Not right now, but eventually
06:28 szabgab so you would be able to remove all the editor related menu items (sorry, actions)
06:28 szabgab and replace them with your own in some super easy way
06:28 Alias Well, I'm not sure to what extent we could rip stuff out
06:29 Alias But certainly the general application layout, the ::Locale stuff, maybe the config layer
06:29 Alias ::DB, and so on
06:29 Alias I don't think we're at the point that we should actually do it yet though
06:29 Alias When we start to see plugins that are basically applications in their own right, then we're probably closer to the time we need to do it
06:30 szabgab I am just wondering if we are not replicating stuff that is already available in wxWidgets?
06:30 Alias To some extent we are
06:30 Alias I've tried to push stuff we've done down into pure wxWidgets where possible
06:30 Sewi szabgab: Removing the menu items is something else than removing the actions. You could already remove the menu items by putting a line into your config
06:31 Alias Some stuff, we're just providing the glue between raw wxWidgets, and what a CPAN-installable Perl application needs to be layout out like
06:31 Alias Other things, we're just providing some more layers of DWIM over the top of the pure wx
06:32 Alias I have tried to keep stuff pure wx where possible though
06:32 Alias Mostly because if you can achieve that, it's AWY faster
06:32 Alias WAY
06:33 szabgab have you seen the smoke reports?
06:33 Alias me?
06:33 Alias No
06:33 szabgab http://smolder.plusthree.com/app/​public_projects/smoke_reports/11
06:33 Alias neato
06:34 szabgab I'll generate e-mails from the failing reports and send them to a mailing list
06:34 szabgab and maybe to the last committer as well
06:38 Alias Or all committers since the last smoke run
06:38 Alias or something
06:45 szabgab the process could track down the commit that broke the tests but I think more important for me to set it up on windows as well
06:46 Alias yup
06:47 Sewi There should be a delay in mail reports. Sometimes commits break tests, this is a fact, but we shouldn't send out mass-mails if it has been fixed by the author some commits later (maybe by just finishing work on this function).
06:48 Sewi A failing test should fail at least two smoke runs (times of run, not two runs at the same time on the same OS) before it is being reported.
06:49 szabgab we should see how much can we eliminate the breaking of tests
06:51 Sewi Maybe Hyppolit should report failing tests immed.
06:51 Alias It probably should
06:51 Alias Or maybe CHANGES in failing tests
06:52 Alias <Hyppolit> FAIL! Failed 8 tests, 3 more than last time. (Random encouraging quote here)
06:52 Alias <Hyppolit> PASS! Woohoo!
06:52 szabgab sounds like a good idea
06:52 Alias etc
06:52 Alias Who owns the bot?
06:53 szabgab it is in our repo
06:53 Alias i.e. Knows the structure of the code in our repo
06:53 szabgab and I run it on our server
06:53 szabgab and the smoke runner is also in our repo
06:54 Sewi Getty did most of the recent Hyppolit changes
06:55 Alias seen Getty
06:55 Alias (just on the off chance we have a bot with seen support)
06:55 Alias We should get purl in here or something
06:55 Alias Indoctrinate her into the hype team
07:00 szabgab Alias, don't wake Getty up as I am not yet running his version of Padre
07:00 szabgab and he will be angry - rightfully
07:02 Alias :)
07:10 szabgab his version of Hyppolit I meant, but yes
07:10 Alias ah
07:23 Sewi Why does PerlCritic warn about Padre/File/FTP.pm line 59? It's a "return undef, $Result;" is this forbidden, too?
07:25 kaare joined #padre
07:54 Hyppolit svn: r9357 | Sewi++ | http://padre.perlide.org/trac/changeset/9357
07:54 Hyppolit Another Critic fix
07:54 Hyppolit trunk/Padre/lib/Padre/Task/
07:54 marcela joined #padre
07:59 Alias Sewi: It's probably a bug in critic
08:00 Alias I'd say that it probably wants to be return( undef, $Result );
08:00 Alias ... or a tweak to critic
08:00 Alias Personally, I don't think we should follow the "no return undef" rule
08:00 Alias It's dubious
08:01 Sewi Usually I return undef where exactly one return argument is expected and return; otherweise.
08:01 Alias yup
08:01 Alias Me too
08:01 Sewi It's clearer to the reader <-- my opinion
08:01 Alias I do a lot of stuff where calling it in list context, you WANT undef
08:02 Alias foo( bar(), 1 );
08:02 Alias That turns into foo( undef, 1 )
08:02 Alias So at least the positions of each argument stays consistent
08:02 Sewi Yeah, it's really bad if you need to hunt bugs due to list-position changes :-(
08:02 Alias return turns that into foo( 1 )
08:03 Alias So basically, this is an argument of two bugs
08:03 Alias If you do "return undef" then you get listwise true for a false return
08:03 Alias If you do "return" you get argument position-shifting
08:04 Alias Personally, I prefer return undef, so you can safely next function calls inside each other
08:04 Alias If you use return, you can never do nesting
08:05 Alias You also cant do this
08:05 Sewi Does method autocomplete work for you?
08:05 Alias foo( key => 'value', key => value() )
08:05 Alias I don't use method auto-complete
08:06 Alias Since it's not clear, and it's just dogma from PBP, I think we should remove the 'no return undef' rule
08:06 Sewi Would you try it, please? r9228/Win7 works for me but r9356/Ubuntu doesn't.
08:06 Hyppolit Changeset #9228 http://padre.perlide.org/trac/changeset/9228
08:07 Alias How does it work?
08:07 Sewi Enable it in the preferences, open a .pm - file (atm it doesn't like new files, don't know why) and type "sub foo " <-- including this last space.
08:08 Sewi It will create a complete method ({, }, my $self = shift) for you and set the cursor at the right position
08:08 Alias hrm
08:08 Alias xt/20-actions.t .................... 'cperl' is not recognized as an internal or external command, operable program or batch file. Argument "" isn't numeric in numeric ne (!=) at xt/20-actions.t line 38. The system cannot find the path specified. Argument "" isn't numeric in numeric ne (!=) at xt/20-actions.t line 38. The system cannot find the path specified
08:08 Sewi reduces typing, but it ignores files without packages.
08:09 Sewi Thanks, I didn't "update" my Win-Padre before trying the method thing with the "old" revision...
08:09 Sewi Now I know that actions.t doesn't work on Win :-)
08:10 Alias What's cperl?
08:10 Alias ctags?
08:11 Sewi Isn't it "console perl" like wperl for "window-only perl"?
08:11 Alias That would just be "perl"
08:11 Alias Did you use the Perl prober?
08:11 Sewi r9357 doesn't work on W7 for method thing
08:11 Hyppolit Changeset #9357 http://padre.perlide.org/trac/changeset/9357
08:11 Sewi No, a selfmade detection.
08:12 Alias FAIL
08:12 Alias You know we made an entire module just for this problem
08:12 Alias Padre::Perl::cperl()
08:12 Alias Actually, it's param agnostic
08:12 Sewi Is it Padre-independent (could it be used from tests)?
08:12 Alias So Padre::Perl->cperl
08:13 Alias Padre::Perl should work just fine standalone
08:13 Sewi Yes, no use's inside
08:13 Alias It's just a private improved version of Probe::Perl
08:14 * waxhead is upgrading modules
08:14 waxhead shuld have a running padre shortly
08:17 Hyppolit svn: r9358 | Sewi++ | http://padre.perlide.org/trac/changeset/9358
08:17 Hyppolit Fixed actions test
08:17 Hyppolit trunk/Padre/xt/
08:18 pece joined #padre
08:18 Sewi actions.t is more a sample for action-queue-based tests than a complete test.
08:25 Hyppolit svn: r9359 | adamk++ | http://padre.perlide.org/trac/changeset/9359
08:25 Hyppolit Adding wxGlade logo
08:25 Hyppolit trunk/Padre-Plugin-wxGlade/ trunk/Padre-Plugin-wxGlade/share/ trunk/Padre-Plugin-wxGlade/share/icons/ trunk/Padre-Plugin-wxGlade/share/icons/16x16/
08:25 Alias Sewi: Yup
08:25 Alias hrm
08:25 Alias Maybe we should write an actual Perl::Probe test
08:26 Alias Yay for automatic logo feature!
08:26 Alias I really should move the logos for everyone else
08:27 Alias It's silly at this point that people are coding their own logo registration
08:31 marcela joined #padre
08:37 Hyppolit svn: r9360 | adamk++ | http://padre.perlide.org/trac/changeset/9360
08:37 Hyppolit I've managed to reimplement "Open File" :)
08:37 Hyppolit trunk/Padre-Plugin-wxGlade/lib/Padre/Plugin/
09:27 daxim joined #padre
09:52 Potatohead joined #padre
10:41 pece joined #padre
10:43 Getty installing padre
10:43 Getty on my workstation at work, and then padre php plugin and then sewi gets his comments :-P
10:43 Getty and 20< bug reports hrhr
10:43 Getty Sewi: i test with padre-0.50 and padre-php-plugin which is on cpan, not svn
10:46 Sewi ok
10:58 Hyppolit svn: r9361 | adamk++ | http://padre.perlide.org/trac/changeset/9361
10:58 Hyppolit Checkpoint commit.
10:58 Hyppolit Bulk constructor code works.
10:58 Hyppolit Lexification of unimportant variables works.
10:58 Hyppolit Wx::wxCONSTANT normalisation works.
10:58 Hyppolit trunk/Padre-Plugin-wxGlade/lib/Padre/Plugin/ trunk/Padre-Plugin-wxGlade/t/
11:19 siegfri3d joined #padre
11:20 pece2 joined #padre
11:33 pece joined #padre
12:18 waxhead nite all
12:43 Hyppolit svn: r9362 | adamk++ | http://padre.perlide.org/trac/changeset/9362
12:43 Hyppolit Dialog import now automatically applies lexify_unimportant and normalise_constants
12:43 Hyppolit trunk/Padre-Plugin-wxGlade/lib/Padre/Plugin/ trunk/Padre-Plugin-wxGlade/t/
13:00 patspam joined #padre
13:08 Alias Sewi: What's up with the whole Action subclass thing?
13:08 Sewi Anything broken?
13:08 Alias It makes no sense to me, and it seems to have broken some stuff
13:08 Alias What is an instance of an Action::Plugins meant to be
13:09 Alias All the module tools are broken
13:09 Sewi I did the split because actions and menus were mixed up.
13:09 Sewi Which module tools?
13:10 Alias So what did the split do?
13:11 Alias It seems to mostly have just moved half the code into two different files
13:11 Sewi Not much more than splitting the action handler and the menu structure.
13:12 Sewi There was a fixed relationship in the code between menu item and action but this wasn't true. The split was the first step to get configurable menus, toolbars and key bindings.
13:13 Sewi The main menu bar is now configurable (it is build using a configuration default value atm), same for toolbar.
13:13 Alias Do we really want the menu to be configurable?
13:14 Alias Anyways
13:14 Alias It can't hurt I guess, and we can track how many people end up doing it
13:16 Alias Why are the event registration things in an object we make but don't use?
13:18 Sewi Maybe a class method or function would be better... I tried to keep the grouping and not move everything to one big ugly file where you get lost looking for something
13:27 Alias It is a good idea in principle, the problem is that everything is still menu-oriented
13:29 Sewi I kept the groups because I thought they make sense.
13:38 Getty oh
13:38 Getty the install failed
13:39 Getty stupid question, how i can tell CPAN to redo the tests?
13:39 Getty he just tells me that test not OK and thats it
13:39 Sewi test Package
13:39 Sewi force test Package
13:40 Getty then he gives me still the saved result
13:40 Getty still
13:40 Getty was also my first try
13:40 Getty wait i show you what i see
13:40 Getty http://paste.debian.net/52860/
13:41 Getty or is it really that ExtUtils error which make it fail?
13:41 Sewi I'ld guess yes
13:42 Getty i think its all old version stuff
13:43 Getty some minimum version needs to be added to Padre i think, or other modules it depends on
13:43 Getty i now let him do a full upgrade of anything
13:43 Getty i love local::lib
13:47 Hyppolit svn: r9363 | adamk++ | http://padre.perlide.org/trac/changeset/9363
13:47 Hyppolit Completed what I hope is the first working version
13:47 Hyppolit trunk/Padre-Plugin-wxGlade/lib/Padre/Plugin/
13:48 Getty what is wxGlade plugin? making all stuff nicer?
13:48 Alias So there's this visual design tool for Wx
13:48 Alias Called wxGlade
13:48 Alias And it can even generate Perl code
13:49 Alias But, it is (of course) horrible standalone useless Perl
13:50 Alias The wxGlade plugin will examine this useless generated Perl, extract the useful bits, then rewrite the code to be Padre compatible, and turn it into a working dialog class
13:50 Getty ah ok
13:51 Alias So basically, this is a tool for spitting out good GUI code
13:51 Getty "to be Padre compatible" so its more likely a Padre-Plugin-WxGlade-For-Padre ?
13:51 Getty ah so not really necessary to run the code in Padre?
13:51 Getty that sounds like something i need then
13:51 Alias Well, at the moment you have to run it in Padre
13:51 Alias Because there's no PTK (Padre Toolkit) analogue yet
13:52 Getty no i mean in general, that the resulting code is independent, sounded like it just runs as Padre stuff
13:52 Alias It just runs as Padre stuff
13:52 Alias Since this is version 0.01, we have no idea how to produce good GUI code in the general case
13:52 Getty the resulting code of the stuff i do with Padre-Plugin-WxGlade?
13:52 Getty ah ok
13:52 Getty get it now :)
13:52 Alias And the horrible generated code does actually run
13:52 Getty ghehehe
13:52 Alias So anyone doing regular stuff can use wxGlade directly :)
13:54 Alias Much of the nicer style of our GUI code relies on things like Padre::Wx and Padre::Locale to set up
13:54 Alias I am, however, trying to keep the functionality separate
13:54 Getty yeah and as long you dont put up a heavy concept for making it sugar-style, its good to have a clean generation process
13:54 Alias Right
13:55 Alias At the moment, by far the biggest userbase for Wx is Padre developers :)
13:55 Getty yeah
13:55 Alias So it makes sense for now to be mostly-about-Padre
13:55 Getty it just feels a bit
13:55 Getty taht you should rename it
13:55 Getty Padre-Plugin-wxGlade-Padre
13:55 Alias It now says sub plugin_name { 'Padre wxGlade Tools' }
13:55 Getty so that longtime if someone makes an independent version it doesnt collide
13:55 Alias You can't use sub-names like that fwiw
13:56 Alias It makes the plugin loading sequence faster
13:56 Getty then: Padre-Plugin-wxGladePadre
13:56 Getty ;)
13:56 Alias For the moment, I'm not wirred
13:56 Alias worried
13:56 Getty hehe
13:56 Getty if the case comes... ;)
13:56 Alias If the case comes, I'll change it then
13:56 Alias Move the code into the developer plugin or something
13:57 Alias Anyways, this works just fine for now
13:57 Getty hehe :)
13:57 Alias The version I just comitted works pretty damned well
13:57 Alias Try it
13:58 Getty i still wait for padre itself
13:58 Getty have to upgrade all CPAN
13:58 Alias ah :)
13:58 Getty so that i can test again
13:58 Getty on standard debian, if i just do local::lib and then install Padre, it gives out a big chaos
13:58 Getty ExtUtils::Install + Module::Install version chaos
13:58 Getty so in result: it doesnt install
13:58 Getty so now i upgrade all CPAN packages in my local::lib and then see if that helps
14:01 Alias Longer term I really want a visual designer thingy
14:01 Alias But that will be an enormous amount of work
14:01 Getty i'm with yeah
14:02 Getty ;)
14:02 Getty that would kick ass perl interface development
14:02 Alias I'm sure there's a good reason that wxGlade is partly abandoned
14:02 Alias For now though, wxGlade does a pretty decent job
14:02 Getty could be the area where perl makes it "big jump" (so that everybody wants to make perl)
14:02 Alias I've already made about 4 dialogs with it
14:03 Alias Hard to say on the big jump thing
14:03 Getty for making interfaces so far every language suxx beside Microsoft supported stuff
14:03 Alias oh, that
14:03 Getty i mean with C# and Visual Studio and stuff, its so damn easy to make an interface
14:03 Alias Yeah, for windows there's not really anything simple
14:03 Getty what you do for non-Windows?
14:03 Alias Best thing is just to make stuff easier and more brain-dead one step at a time
14:03 Getty i mean i say its just simple with windows, with nothing else
14:03 Alias All my $work code is for Unix
14:03 Alias For years
14:03 Alias Actually, forever
14:04 Getty yeah but you wouldnt say its idiot proofed to make an interface on unix
14:04 Alias I'm pretty much a Windows Client + Linux Server guy
14:04 Getty not as idiot proofed as for windows
14:04 Getty me too
14:04 Getty or generall depending on language, making interfaces is still hard work
14:04 Getty only with Visual Studio and the designers there its "snap easy"
14:05 Getty if someone cracks down that area... could be a big rush
14:19 siegfri3d joined #padre
14:22 awnstudio joined #padre
14:29 Sewi svn?
14:29 Hyppolit svn is http://padre.perlide.org/trac/wiki/Download#​Moredetailedinstructionformanualinstallation
14:30 kthakore oh hi Alias
14:31 Alias hiya
14:31 kthakore Alias: so you are back huh
14:31 kthakore Alias: crazy month is done?
14:32 kthakore Sewi: Has anyone looked into the min/max bug on windows
14:32 Alias kthakore: Crazy month is done
14:32 kthakore Alias: nice ... mines is begining soon
14:39 kthakore Alias: oh btw http://sdl.perl.org :) our work for last month... Getting closer to Frozen Bubble hacking
14:40 patspam joined #padre
14:41 Alias kthakore: I saw the announcement of the first Generation 2 dev releases
14:53 kthakore Alias: yup :)
14:54 kthakore Alias: it is still very rough ... but I think it is a good effort.
14:58 CromeDome joined #padre
15:03 dapatrick joined #padre
15:55 Haarg joined #padre
16:07 Hyppolit svn: r9364 | adamk++ | http://padre.perlide.org/trac/changeset/9364
16:07 Hyppolit Updating
16:07 Hyppolit trunk/wxglade/
16:07 Hyppolit svn: r9365 | adamk++ | http://padre.perlide.org/trac/changeset/9365
16:07 Hyppolit Updating
16:07 Hyppolit trunk/Padre-Plugin-wxGlade/t/sample/
16:25 Hyppolit svn: r9366 | Sewi++ | http://padre.perlide.org/trac/changeset/9366
16:25 Hyppolit Fixed invalid ref
16:25 Hyppolit trunk/Padre-Plugin-SVN/lib/Padre/Plugin/
16:45 Hyppolit svn: r9367 | adamk++ | http://padre.perlide.org/trac/changeset/9367
16:45 Hyppolit The code we generate is now a bit bloatier, but thankfully not extremely long
16:45 Hyppolit trunk/Padre-Plugin-wxGlade/lib/Padre/Plugin/ trunk/Padre-Plugin-wxGlade/t/
17:33 Woody2143 joined #padre
17:38 Hyppolit svn: r9368 | adamk++ | http://padre.perlide.org/trac/changeset/9368
17:38 Hyppolit Adding Padre::Plugin::wxGlade-generated warning dialog (for use in a future "Do not show again" functionality)
17:38 Hyppolit trunk/Padre/lib/Padre/Wx/Dialog/
18:00 garu zamolxes, you have a commit bit, right?
18:01 zamolxes if it's on piet, yes
18:11 Haarg joined #padre
18:40 garu joined #padre
18:42 tsee joined #padre
19:08 isec joined #padre
19:09 ingy joined #padre
23:21 Hyppolit svn: r9369 | garu++ | http://padre.perlide.org/trac/changeset/9369
23:21 Hyppolit adding zamolxes++'s ideas to TODO file
23:21 Hyppolit trunk/Padre-Plugin-Catalyst/
23:22 garu zamolxes, how is showing the server in a special panel better than using the output panel?
23:31 zamolxes garu: dunno, it would be vertical :)
23:32 garu hmmm... I think you just gave me some ideas, thanks
23:33 zamolxes garu: i mean, i always have a server running in some screen
23:34 zamolxes garu: but I also run commands that would use the output window
23:34 garu you're right. since we should be able to run other projects and scripts independently of the server being up or not
23:34 zamolxes and the server should be vertical :)
23:37 * garu still wonders what exactly you mean by that :)
23:38 Alias hi
23:38 garu hi Alias
23:40 Hyppolit svn: r9370 | garu++ | http://padre.perlide.org/trac/changeset/9370
23:40 Hyppolit yet more ideas
23:40 Hyppolit trunk/Padre-Plugin-Catalyst/

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