Camelia, the Perl 6 bug

IRC log for #padre, 2009-12-21

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

All times shown according to UTC.

Time Nick Message
01:19 patspam are plugins able to register new mime types?
01:49 patspam without modifying Padre::MimeTypes in the core?
01:49 patspam e.g. I want to tell Padre to open some new file extension with my plugin's document class
03:52 holli joined #padre
04:15 cognominal joined #padre
04:18 ppant joined #padre
04:19 ppant Good Morning Padre Team
04:27 patspam joined #padre
04:46 [1]awnstudio joined #padre
05:32 szabgab gm
06:02 Hyppolit svn: r9830 | szabgab++ | http://padre.perlide.org/trac/changeset/9830
06:02 Hyppolit alternative coding with full Mini::CPAN mirror
06:02 Hyppolit trunk/Perl-Dist-XL/ trunk/Perl-Dist-XL/lib/Perl/Dist/ trunk/Perl-Dist-XL/script/
06:10 szabgab Is anyone a DMOZ editor here? If so both Strawberry Perl and Padre could be added to this http://www.dmoz.org/Computers/Pr​ogramming/Languages/Perl/Tools/
07:06 marcela joined #padre
07:12 szabgab hi marcela
07:12 marcela morning
07:15 azawawi joined #padre
07:16 azawawi good morning
07:16 azawawi szabgab: http://www.dagolden.com/index.php/59​0/modulebuild-version-0-36-released/
07:17 szabgab azawawi, hi
07:17 szabgab what about it?
07:17 azawawi several of our Padre plugins could now use it...
07:17 azawawi sharedir ...
07:18 szabgab ah
07:18 szabgab let's wait a bit with that, before we jump ahead...
07:18 azawawi dont worry... i just had a note about it in my old todo list :)
07:19 kaare joined #padre
07:20 Hyppolit svn: r9831 | szabgab++ | http://padre.perlide.org/trac/changeset/9831
07:20 Hyppolit separate the debugger actions in their own class
07:20 Hyppolit trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Action/
08:13 Hyppolit svn: r9832 | szabgab++ | http://padre.perlide.org/trac/changeset/9832
08:13 Hyppolit improve the unit test running from the Run menu
08:13 Hyppolit trunk/Padre/lib/Padre/Wx/
08:20 cognominal joined #padre
08:51 szabgab oups I broke the tests, badcode
08:58 pece joined #padre
08:59 Hyppolit svn: r9833 | szabgab++ | http://padre.perlide.org/trac/changeset/9833
08:59 Hyppolit fix badcode test
08:59 Hyppolit trunk/Padre/lib/Padre/Wx/
09:02 szabgab print "getting handlers..\n" unless ( $config{silent} );  is found in /home/gabor/perl5/lib/perl5/Autodia.pm
09:17 daxim joined #padre
10:43 waxhead joined #padre
10:44 waxhead evening all
10:44 waxhead how goes locker breakage?
10:54 waxhead the last used tabbing doesn't work too well
11:20 Hyppolit #804: CRTL-Tab Last seen order not quite right (new defect) [ http://padre.perlide.org/trac/ticket/804 ]
11:21 Hyppolit #804: CRTL-Tab Last seen order not quite right (new defect) [ http://padre.perlide.org/trac/ticket/804#comment:1 ]
11:23 Hyppolit #795: Ctrl-Tab is using tab display order rather than last-used order (new enhancement) [ http://padre.perlide.org/trac/ticket/795#comment:6 ]
11:52 marcela joined #padre
12:29 Hyppolit svn: r9834 | adamk++ | http://padre.perlide.org/trac/changeset/9834
12:29 Hyppolit Added the missing ->locked shortcut method
12:29 Hyppolit trunk/Padre/lib/Padre/Wx/
12:44 szabgab well, that one blew up the tests as well
12:51 szabgab jq, ping
13:03 Hyppolit svn: r9835 | adamk++ | http://padre.perlide.org/trac/changeset/9835
13:03 Hyppolit trunk/Padre/lib/Padre/Wx/
13:13 Hyppolit svn: r9836 | adamk++ | http://padre.perlide.org/trac/changeset/9836
13:13 Hyppolit This doesn't solve bug #804 properly, but it does squelch the warning
13:13 Hyppolit trunk/Padre/lib/Padre/Wx/
13:16 Alias joined #padre
13:33 szabgab Alias, have you tried the debugger on window?
13:59 patspam joined #padre
14:02 Hyppolit svn: r9837 | garu++ | http://padre.perlide.org/trac/changeset/9837
14:02 Hyppolit updated brazilian portuguese translation
14:02 Hyppolit trunk/Padre/share/locale/
14:06 Hyppolit svn: r9838 | garu++ | http://padre.perlide.org/trac/changeset/9838
14:06 Hyppolit updated brazilian portuguese translation
14:06 Hyppolit trunk/Padre-Plugin-Catalyst/share/locale/
14:17 submersible1 joined #padre
14:33 jq szabgab: pong
14:48 Alias szabgab: Nope
14:48 Hyppolit svn: r9839 | jquelin++ | http://padre.perlide.org/trac/changeset/9839
14:48 Hyppolit updated french translation
14:48 Hyppolit trunk/Padre-Plugin-Catalyst/share/locale/
14:48 * Alias is trying to avoid distractions while rewriting Hook::LexWrap and Aspect.pm
14:51 szabgab jq, I've just downloaded mandriva 2010.0, will try to install it on a virtual box, is there an easy way to build an arbitrary live CD based on Mandriva ?
14:51 jq what do you want to do?
14:51 jq run a live cd?
14:52 jq or build a new one with what you want on it?
14:53 jq run an existing live cd: download mandriva one which is a live cd that can also act as an installer
14:53 jq create your own live cd: http://wiki.mandriva.com/en/Draklive
14:54 szabgab I'd like to create a live CD
14:54 jq create your own live cd: http://wiki.mandriva.com/en/Draklive
14:54 szabgab and install on it tons of interesting stuff
14:55 szabgab and whatever is missing, nag you to add it
14:55 jq then you're looking for draklive
14:55 jq which is used by mandriva people themselves to create their live cd
14:56 szabgab and I guess I can build a live CD using Cooker, right ?
14:56 jq sure
14:56 jq i've never done it, but cooker is just a 'living' dist
14:57 jq other than that, it's the same than a regular dist
14:58 szabgab Will you have time to add tons of Perl modules to Cooker?
14:58 jq sure, if you tell me which one you want
14:58 szabgab e.g. many of the Padre plugins
14:58 szabgab Perl6 and Catalys plugins
14:58 szabgab Smolder
14:58 szabgab for now
14:58 jq note that lots of them are *not* installable from cpan
14:58 jq which means that i cannot ship them as rpm for mdv
14:59 szabgab what do you mean?
14:59 szabgab that are not installabel from cpan?
14:59 jq when i want to install some of them, i get 'not in author tree' or sthg like that
14:59 El_Che_ szabgab: If all go well, I would have some time to assist you on fosdem. Time and no money: taking time off for phd (still fighting with the boss :) )
14:59 szabgab El_Che_, I found the source of the warning on the consolo
15:00 szabgab print "getting handlers..\n" unless ( $config{silent} );  is found in /home/gabor/perl5/lib/perl5/Autodia.pm
15:00 El_Che_ szabgab: /nick El_Che
15:00 El_Che_ oops
15:01 szabgab El_Che, it would be great your help on fosdem
15:01 szabgab do you see any chance organizing a course as well?
15:02 szabgab jq, so basically this looks like a set of perl scripts to build the liveCD
15:02 Hyppolit svn: r9840 | jquelin++ | http://padre.perlide.org/trac/changeset/9840
15:02 Hyppolit updated french translation
15:02 Hyppolit trunk/Padre/share/locale/
15:02 jq szabgab: seems so.
15:03 jq ask on cooker@ for more information
15:03 jq i'm no expert of this, since i never used it
15:03 jq or on freenode#mandriva-cooker
15:03 jq it should work, since one of the community member is using them to create a live cd based on xfce
15:06 El_Che szabgab: I am preparing a request for a 1 year fellowship
15:06 El_Che szabgab: zero time left
15:06 El_Che pretty important: if all goes well, I can stay paid at home for a year
15:06 El_Che szabgab: all the presentations I have in the drawers are about linguistics :)
15:07 szabgab jq, I have another issue I wanted to ask you, about git
15:07 El_Che for your perl training I send a mail here to the people interested
15:07 szabgab I started to use git svn on my notebook
15:07 szabgab instead of SVK
15:07 El_Che I really hope bureaucracy won't block it
15:07 jq great. git++
15:08 szabgab and now I am wondering to try to push it out to github as well
15:08 jq you need to create a new origin
15:08 szabgab if I do that I can let people fork it there and I can pull in their commits from there
15:08 jq and i think it's the best to have a dedicated branch to push on github
15:09 szabgab oh, I don't know what does that mean
15:09 jq so let's say that your local branch will be 2github
15:09 jq first create it:
15:09 jq git checkout -b 2github
15:09 szabgab one sec
15:09 jq then create your repo on github
15:09 jq then create the remote 'github' that will point to github
15:10 PacoLinux joined #padre
15:10 szabgab I am worried what will happen if my copy is gone?  Is the mapping between svn and the git repo kept in the git repo ?
15:10 szabgab will you also be able move commits from github to our svn repo ?
15:10 jq git remote add github git@github.com:szabgab/padre
15:10 jq yes, the mapping is kept in the tracking branch
15:11 jq which explain why i think it's best to have a branch dedicated to github
15:11 jq so, once the github branch is created (and is the active branch) and the project is created on github
15:11 jq you need to push the branch one github
15:12 szabgab so I pull the whole svn repo to my local machine, create a branch there and add the remote to this branch
15:12 jq git push github 2github
15:12 jq yes
15:12 jq but beware
15:13 jq you will have to be careful when merging work pushed by others
15:13 jq 1) either you rebase their work in order to have a clean history (which svn will prefer)
15:13 szabgab should I create a repo by pulling from /trunk only, right ?
15:13 jq right
15:14 jq 2) or you don't and all the work will appear as one merged commit in svn
15:14 jq 1) will mess your repo on github
15:14 szabgab so let me forget about the livecd for now and try to understand the github thing
15:14 jq 2) will squash commits and loose history
15:14 jq none of the problem is a big one
15:14 jq but you'd better know what you're doing
15:17 szabgab the problem is that I still don't understand git
15:17 szabgab I need to learn a bit more about it
15:18 szabgab now fetching svn, it will take a while
15:19 szabgab jq, does the work 'origin' have special meaning ?
15:20 szabgab and  'master' ?
15:20 jq szabgab: no, it's just the default name of the first origin added to your repo
15:20 jq and no, it's just the default name of the first branch in your repo
15:20 szabgab like trunk, tags, branches in svn, right ?
15:20 jq so you can name them whatever you want
15:20 szabgab just arbitrary but accepted names
15:21 jq yes
15:32 szabgab jq you suggested to write "git remote add github git@github.com:szabgab/padre"  while after creating the repo on github it suggest to write "git remote add origin git@github.com:szabgab/padre.git"
15:32 szabgab origin instead of github
15:32 szabgab why?
15:32 szabgab or is that just that I call the remote repo 'github' instead of 'origin' ?
15:33 jq yes
15:33 jq it's just the name that changes
15:33 jq i don't know if there's already an origin declared by git-svn
15:33 jq run:
15:33 jq git remote show
15:34 jq to see the list of your remotes
15:36 szabgab I am still waiting for the initial fetch to complete
15:36 szabgab I am at 2208
15:40 jq yes, svn is sloooow
15:43 Alias joined #padre
15:50 awnstudio joined #padre
16:32 cognominal joined #padre
16:36 Hyppolit svn: r9841 | ddn123456++ | http://padre.perlide.org/trac/changeset/9841
16:36 Hyppolit Updated dutch translation.
16:36 Hyppolit trunk/Padre/share/locale/ trunk/Padre-Plugin-Catalyst/share/locale/
17:12 szabgab jq, soon it will finish (at 9000 now)
17:12 szabgab but it is not clear to my when I am pushing out to github why does it matter which branch am I pushing out
17:13 szabgab I thought the main thing would be where do I pull to (when I pull from github)
17:13 szabgab or is that just to make some convention for myself?
17:14 jq well, you can be in whatever branch you want when you push
17:14 jq but if you're in the correct branch, you can skip its name
17:15 jq eg, to push branch foo to remote branch rfoo on remote myremote:
17:15 jq git push myremote foo:rfoo
17:15 jq but if you already pushed that once, and that you are on branch foo, you can just type:
17:16 jq git push
17:16 jq which is easier, faster to type, and less error-prone
17:18 Alias God I wish git had better terminology
17:22 szbalint It would have been nice for someone to write git 5 years before it was written, so that we get used to the terminology
17:27 szabgab ok, initial fetch was done
17:28 Alias szbalint: Well, it uses SIMILAR words to the things that were written 5 years before, but not always QUITE the same
17:28 szabgab git remote show     gives nothing
17:33 szabgab now pushing out to github
17:40 szabgab jq, there you go, I pushed it out to github
17:40 szabgab now what ? :-)
17:41 daxim I'll clone and have a look
17:41 daxim and compare with my local git-svn repo
17:42 szabgab great, thanks
17:42 daxim clone was done in 10 seconds
17:42 szabgab the whole thing is really so small
17:44 daxim all the tags and branches are missing.  is this your first conversion from svn?
17:44 szabgab yes, and as you could read in the backlog that's what jq suggested
17:46 daxim can't find it
17:46 daxim anyway, http://blog.woobling.org/20​09/06/git-svn-abandon.html is the way to go
17:46 daxim I converted Xacobeo with it and potyl was delighted
17:48 szabgab who is potyl ?
17:48 daxim http://search.cpan.org/~potyl
17:49 szabgab I'll read what nothingmutch writes but do you remember that our repository went through several changes where directories were moved around?
17:50 szabgab I am not sure how that will be handled or why, really why should we care of what happened in the first 10,000 commits
17:50 Alias To work out how the fuck bits of crazy mess got that way
17:51 Alias Although I don't imagine we need to care about the time before we got the current layout quite as much
17:51 Alias Cause it was still mostly just you and me at that point
17:52 daxim why should we not care? :)
17:53 daxim I suggest filtering the whole thing according to its top directories afterwards
17:53 daxim in other words Debug-Client becomes it own repo and so on.
17:53 * Alias dislikes that approach
17:53 Alias Now I have to download 100 different repos?
17:54 daxim the all-in-one approach that svn fails to discourage always strikes everyone as unnecessary hacky
17:54 Alias I like it more for it's sense of encouragement
17:54 Alias All the code is there, in one place
17:54 daxim yes, 81 repos
17:54 Alias You can do global search/replace, you get an implicit permissions to work on things
17:55 daxim automation
17:55 Alias Which nobody will use of course
17:55 daxim what makes you think that?
17:55 daxim embrace the new toolchain
17:56 Alias What new toolchain
17:57 Sewi joined #padre
17:57 Sewi hi all
17:57 szabgab mind you guys we are not switching to git yet, I am just experimenting with it and wanted to see if this would allow others to contribute
17:57 daxim git and its extensions for dealing with several repos
17:57 daxim alright, then I pledge myself to do a proper conversion with author file and all.
17:57 daxim tonight starts my end-of-year vacation, I'll have time
17:58 szabgab so maybe it should be called padre-svn on github
17:58 daxim do you want to run parallel?
17:58 szabgab for now svn is the authoritive repo
17:59 daxim I understand
17:59 szabgab yes, I thougt keeping them in parallel
18:00 szabgab so maybe I'd better change the name on github so later if and when we switch to github we can do a more proper import - if I understand it by that time
18:00 daxim then another important question, do you want patchsets to flow from github to svn?
18:00 szabgab what do you mean by "patchsets" and "flow" >
18:00 szabgab ?
18:00 szabgab I though I'll have to pull them from github
18:00 szabgab and then push them to svn
18:01 szabgab manually
18:01 Alias I personally have no objection if patches come in from github to us
18:01 daxim hmmm, I mean people commit to the git mirror
18:01 szabgab yes, though I guess people will have to fork it on github first
18:01 szabgab then I won't have to deal with giving them commit bit
18:02 daxim this is correct
18:02 daxim however, some git metadata cannot be accurately represented in svn
18:03 daxim so my point with the question earlier was to assess whether you are okay with that
18:04 szabgab what metadata?
18:04 Alias I would say we're probably ok with it
18:05 daxim of the top of my head:  svn has just a committer identity.  git has author and committer, and the id is different
18:06 daxim how does one deal with a git committer id, then, how to map it to an svn id?
18:07 szabgab you tell me, I never tried it
18:08 daxim there is no good solution because it involves loss of precision or information.  example: create an svn user github and shove/mold everything coming from the mirror into it
18:08 dapatrick joined #padre
18:09 daxim for this reason most people opt for a clean cut switch-over or one-way mirrors
18:10 Alias If we did it one way, we'd lose information anyway
18:10 Alias Or at least, that informatino would just never exist
18:11 daxim you mean the patchsets independently applied on github?
18:12 Alias Well, compare patches imported from a slave github to patches applied directly to svn
18:12 Alias There's not really a HUGE difference
18:12 daxim I certainly would object if it cost me my ohloh juice ;)
18:13 Alias heh
18:13 Alias Well hey, you can still to the git-svn thing
18:14 Alias Loooots of juice
18:14 daxim yes, I got lucky to get an svn account
18:14 * Alias still remembers jq's 20-patch floods
18:25 virtualsue_ joined #padre
19:05 daxim and off I go.  I'll be back today or tomorrow
19:05 Getty ok now i sit on a windows, so i finally can use padre :D
19:08 Getty sounds evil, or? ;)
19:16 pece2 joined #padre
19:18 Alias ok
19:18 Alias So I'm happy for everything of mine to be released at this point
19:18 Alias Locking stuff is migrated, but not enhanced yet
19:18 Alias Which I'll attempt after 0.53
19:44 robn joined #padre
19:58 Woody2143 joined #padre
20:05 Hyppolit svn: r9842 | zipf++ | http://padre.perlide.org/trac/changeset/9842
20:05 Hyppolit Updated Spanish translation
20:05 Hyppolit trunk/Padre/share/locale/
20:13 cognominal joined #padre
20:24 kyanardag joined #padre
20:49 szabgab Sewi, I wonder if we can tell the menusystem of Padre that when a user opens a menu and the mouse is over an item for a long period of time then a text should appear - the long explanation of that action
20:56 Sewi szabgab: That would be really cool. Windows knows tooltips, but I don't know if Wx could handle them
20:56 szabgab I found a SetHelp on MenuItem but calling it did not seem to make a difference
20:59 [1]awnstudio joined #padre
21:04 szabgab Sewi, as it turns out - I was told on #wxwidgets if we set SetHelp on the menu item, the text shows up in the status bar
21:04 szabgab and as I can see there are default texts for the Undo/Redo etc items
21:05 szabgab but we can set the text from action to the  SetHelp
21:08 kyanardag joined #padre
21:16 Hyppolit svn: r9843 | szabgab++ | http://padre.perlide.org/trac/changeset/9843
21:16 Hyppolit show the comments of the actions in the toolbar when the menu is pointed at
21:16 Hyppolit trunk/Padre/ trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Wx/
21:27 patspam (repeat question from yesterday): any way for plugins to register themselves against file extensions / mime types that padre doesn't know about yet?
21:27 patspam I'm thinking about maybe using editor_enable/disable
21:27 szabgab nope
21:28 szabgab I think we had some reason to do that explained in the docs of padre
21:28 szabgab but I am not sure what
21:28 szabgab and why would that be reasonable
22:06 Hyppolit svn: r9844 | szabgab++ | http://padre.perlide.org/trac/changeset/9844
22:06 Hyppolit add test to be skipped as it is crashing
22:06 Hyppolit trunk/Padre/lib/Padre/Action/ trunk/Padre/t/
22:12 submersible1 patspam: is this regarding the workarounds for mimetype in the WebGui plugin ?
22:12 patspam no, for a new plugin
22:13 submersible1 I had similar trouble when using 'registered_document' plugin hook ... you can only hook known mimetype
22:15 patspam yeah, I think it will be pretty common that people want to build plugins for new file types
22:15 submersible1 patspam: if it blocks plugin development ... I would suggest you patch your mimetypes into trunk and raise a ticket for mimetype registration via pluginmanger
22:15 Sewi Sounds good :-)
22:15 Sewi gn8
22:15 submersible1 in my case I want to invent synthetic mimetypes
22:16 szabgab I think the question will be if two plugins try to register the same mime-type
22:17 szabgab but maybe that's the same with an other thing and Padre just needs to decide and warn
22:17 szabgab and the plugin authors need to work it out or get the mime-type in the core
22:30 Hyppolit svn: r9845 | szabgab++ | http://padre.perlide.org/trac/changeset/9845
22:30 Hyppolit change the comments on the Debugger actions
22:30 Hyppolit trunk/Padre/lib/Padre/Action/
22:39 Potatohead joined #padre
23:31 CromeDome joined #padre
23:57 Getty someone feeling alive and could fix a bug in Padre::Plugin::PHP?
23:57 Getty cause i have no idea of the Makefile.PL concepts of it, there is an error of bareword usage
23:57 Getty (is_padre_plugin)
23:58 Getty what is required so that this works?

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