Camelia, the Perl 6 bug

IRC log for #padre, 2010-09-13

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

All times shown according to UTC.

Time Nick Message
00:29 Alias joined #padre
00:29 Alias Sewi: Ping
00:30 Alias There's a problem with Format::Human::Bytes
00:30 Alias You've accidentally added a dependency on Module::Instal
00:30 Alias Module::Install
00:39 Alias Which has added a dependency on M:I for all of Padre
01:00 jazzanova joined #padre
01:14 asarch joined #padre
01:15 asarch That was really hurt
01:15 asarch :-(
01:15 asarch I won't delete/install Perl again
01:15 asarch The system goes crazy with Perl
01:41 asarch *without
01:42 Alias :)
01:56 asarch WOW! Padre went fine!
01:56 asarch \o/
01:56 asarch Yahoo!
01:59 asarch It seems that the SlackBuild package for wxWidgets is not compatible with the current version of Padre
02:00 asarch However... aMule needs wxWidgets
02:05 chorny left #padre
02:09 zenog_ joined #padre
02:13 zenog left #padre
02:20 Alias__ "Complimentary"
02:20 Alias__ er, wrong window
02:20 asarch left #padre
02:33 asarch joined #padre
02:42 asarch Is there any symbols inspector in Padre?
02:56 kthakore Alias__: oh hai!
02:56 Alias__ As in docs, or as in watch?
02:58 kthakore Has anyone noticed a weird warning overwrite in output?
03:05 Hyppolit #1007: warn output overwrites file name and other warns in Linux (new defect) [ http://padre.perlide.org/trac/ticket/1007 ]
03:05 patspam joined #padre
03:07 Alias__ In 5.10, it was indeed lowercased
03:08 kthakore Alias__: see it?
03:08 Alias__ So possibly an update to the unicore dirs resulted in a change of dir naming policy
03:08 Alias__ Which unix things is an add+remove
03:08 kthakore oh you are not talking to me ...
03:08 kthakore ok
03:08 Alias__ But Windows gets completely confused by
03:08 Alias__ oh wait
03:08 kthakore garu: ping?
03:08 kthakore garu: see Ticket #1007 does it happen to you?
03:08 Hyppolit #1007: warn output overwrites file name and other warns in Linux (new defect) [ http://padre.perlide.org/trac/ticket/1007 ]
03:10 * garu looks
03:10 Alias__ oops, wrong window
03:10 * Alias__ away
03:13 kthakore garu: I am not crazy right?
03:13 kthakore it is seriously bugging me
03:17 garu oO
03:17 garu I see it
03:17 garu it works ok if you tell padre to run the script on an external window
03:17 garu but not on the output window
03:18 kthakore yeah but the bug exist right?
03:18 kthakore I am not crazy
03:19 kthakore just tell me I am not crazy
03:19 kthakore please!
03:19 garu you're not crazy
03:19 garu the bug is crazy
03:19 garu I see it alright
03:19 kthakore ok good
03:19 kthakore how do I set external windo run thin?
03:20 kthakore oh I see it
03:20 kthakore great!
03:20 kthakore no more crazy bug
03:21 kthakore I wonder how it sets the tab name like that. ... oh well
03:21 garu tab name?
03:23 garu wow
03:23 garu another bug
03:24 garu hmm, can't reproduce this other one
03:26 kthakore which one?
03:26 kthakore mines?
03:28 Hyppolit #1007: warn output overwrites file name and other warns in Linux (new defect) [ http://padre.perlide.org/t​rac/ticket/1007#comment:1 ]
03:28 garu kthakore: the other one, regarding the name
03:28 garu I totally saw it, but now I can't reproduce it
03:29 kthakore just follow the steps
03:29 kthakore reopen apdre
03:29 kthakore padre
03:29 kthakore it only happens at the start I think
03:29 garu ah, now I see
03:29 kthakore when the *
03:29 kthakore is added
03:29 kthakore when is that added?
03:29 kthakore or rather where?
03:29 kthakore svn?
03:29 Hyppolit svn is http://padre.perlide.org/tr​ac/wiki/DownloadRepository and also http://svn.perlide.org/padre/trunk/ and also http://padre.perlide.org/trac/browser
03:31 garu kthakore: what?
03:33 kthakore well when a file is changed
03:33 shadowpaste "kthakore" at 217.168.150.38 pasted "lib/Padre/Project.pm:304: $nam" (2898 lines) at http://paste.scsys.co.uk/50639
03:33 kthakore a '*' is added to the tab name
03:33 kthakore where is that * added?
03:33 kthakore see ^ that is ack '\*'
03:34 garu there is an event hook for when the panel changes
03:35 kthakore lib/Padre/Wx/StatusBar.pm:195:my $modified       = $self->{_last_modified} ? '*' : ' ';
03:35 kthakore is that it ?
03:37 kthakore yup that is it!
03:37 garu yup
03:37 kthakore my $title          = $modified . $text;
03:37 kthakore it is this line
03:38 kthakore why is it concated to the $text ?
03:38 garu see how $text is built
03:42 kthakore it is coming from $filename and the Base thing
03:42 kthakore File::Basename::basename($filename)
03:42 kthakore where is that?
03:43 garu File::Basename should be in your system
03:43 kthakore oh!
03:43 kthakore it is using $_
03:43 kthakore I think
03:43 kthakore without local
03:44 kthakore whore!
03:44 garu good catch
03:44 garu can you fix it?
03:45 kthakore my($basename, $dirname, $suffix) = fileparse( $path, map("\Q$_\E",@_) );
03:45 kthakore yup!
03:45 kthakore whore!
03:45 garu test it up
03:45 garu then svn commit it :)
03:46 kthakore map("\Q$_\E",@_)
03:46 kthakore what is that?
03:46 kthakore $_ is empty outside of basename
03:46 kthakore weird ness
03:46 garu \Q and \E are used to escape everything in $_
03:47 garu hmmm
03:47 garu kthakore: the warn bug appears to be yet another problem with Wx::Perl::ProcessStream
03:48 Sno| left #padre
03:48 kthakore garu: I am looking at basename ... but don't know where it emerges from
03:49 * Alias plan is to eventually remove Wx::Perl::ProcessStream
03:49 Alias By moving anything that does shell execution to background threads
03:49 kthakore oh well
03:49 garu Alias: people will love you for that, I'm sure :)
03:50 kthakore the bug is in File::Basename::basename($filename);    useage
03:50 kthakore dunno how to fix it
03:50 garu kthakore: do you know how to reproduce it minimally?
03:50 garu without padre?
03:50 Alias I'm starting with an ack -> task replacement
03:50 Alias It's about half written
03:50 patspam left #padre
03:50 garu Alias: the ack panel (find in files) looks way faster than it was
03:51 Alias hrm
03:51 Alias There's still the underlying problem that we can only support the feature set of ack
03:51 garu yeah, absolutely
03:51 Alias I'm mainly doing it for that reason
03:51 Alias So we can do regex subsearch etc
03:51 garu I like ack, but it does have some shortcomings
03:51 Alias With the ultimately goal of project-wide regex search and replace
03:52 Alias ultimate
03:52 Sno| joined #padre
03:52 garu cool
03:52 Alias Ultraedit does it, and it's one of it's killer features
03:52 Alias It's also the last feature in the "Get Adam K on Padre full time" milestone
03:53 Hyppolit #1007: warn output overwrites file name and other warns in Linux (new defect) [ http://padre.perlide.org/t​rac/ticket/1007#comment:2 ]
03:53 garu :)
03:53 kthakore garu: added the comment
03:53 kthakore dunno how to fix it
03:53 garu kthakore: do you know how to reproduce it minimally? without padre?
03:53 kthakore maybe not use FIle::Basename::basename ...
03:53 kthakore garu: welll it is a padre bug
03:53 kthakore so how do I not ... use padre?
03:53 garu if the problem is with File::Basename, it's not a padre bug per se
03:53 kthakore oh you mean the @_ overflow?
03:54 kthakore no clue
03:54 kthakore I don't know where in FIle::Basename it comes from
03:54 garu I'll have a look, let me just try to figure W:P:ProcessStream out
03:54 kthakore ok
03:54 kthakore well I gtg to sleep
03:54 kthakore class tommorow
03:55 garu g'night
03:56 edenc garu: hmm, what's an example of an ack shortcoming?
03:56 Alias edenc: You don't have rigurous control over which subdirs to recurse into or not
03:57 Alias edenc: Regex search...
03:57 edenc doesn't it regex search oob?
03:57 Alias edenc: You can't do proper search abort
03:57 garu it's blocking
03:57 Alias With a Task you can tell it to stop without raw killing the process
03:58 Alias Also, communication with ack is via strings
03:58 garu it doesn't search within compressed files
03:58 Alias A Task could use Perl structures
03:59 Alias Theoretically, a Task could read the unsaved file state in a Padre buffer instead of the disk version
03:59 Alias So that unsaved changes would supercede the file on disk
04:00 garu yup. more controle to searching within padre's unsaved buffers also comes to mind
04:00 garu *control
04:00 Alias There's a shitload of potential cool-ness we might be able to do with the deeper integration
04:44 lgtaube left #padre
04:54 asarch left #padre
04:59 kaare joined #padre
05:35 garu wow, this is a really weird bug
05:35 garu somehow related to ProcessStreams's ->GetLine()
05:36 garu if we make my $stderr = $_[1]->GetLine(), then AppendText($stderr), it works
05:37 garu but the current way, AppendText( $_[1]->GetLine() ) gives us that error
05:37 garu bizarre
05:44 Sewi joined #padre
05:47 garu sigh
05:47 garu that's not it either
05:47 garu it's some buffering issue
05:49 garu AppendText($_[1]->GetLine . "\n") shows me the bug
05:50 garu AppendText($_[1]->GetLine . " \n") has no bug
05:52 * garu hates voodoo bugs
06:04 garu ok, if we flush the entire stderr and stdout buffers on each run, it works
06:13 Hyppolit svn: r12465 | garu++ | http://padre.perlide.org/trac/changeset/12465
06:13 Hyppolit fixed output window buffering issues (ticket #1007)
06:13 Hyppolit trunk/Padre/ trunk/Padre/lib/Padre/Wx/
06:16 Hyppolit #1007: warn output overwrites file name and other warns in Linux (new defect) [ http://padre.perlide.org/t​rac/ticket/1007#comment:3 ]
06:22 garu kthakore: when you wake up, please try to reproduce *both* bugs
06:22 garu I can't reproduce the second one now
06:23 garu maybe, just maybe, that buffering issue was also causing the bogus filename problem
06:23 garu (it would explain why the warn string was being passed as filename, I guess)
06:24 garu kthakore: if you can't reproduce it either, please close that ticket
06:24 garu kthakore: either way, please let me know how things go :)
06:39 garu sigh
06:40 garu how thoughtful of ProcessStream, appending a \n to the stdout buffer everytime it reads a \n somewhere else
06:50 Hyppolit svn: r12466 | garu++ | http://padre.perlide.org/trac/changeset/12466
06:50 Hyppolit fixing "\n" swarm on output panels
06:50 Hyppolit trunk/Padre/lib/Padre/Wx/
06:51 garu I love padre... code is so clean, it's a joy even to debug it :)
06:52 garu ( padre, not wx::processtream :/ )
06:52 Alias :)
06:52 Alias We try :)
06:52 pece joined #padre
06:53 lgtaube joined #padre
06:53 Alias That's what a lot of the "tidy everything, nobody imports functions" etc stuff is about
06:53 Alias Keeping the debugging cognitive load down as much as possible
06:53 garu absolutely
06:53 Alias More core, but easier code
06:54 garu and it really pays in the long run
06:54 Alias I'm glad it's working for you
06:54 Alias It's all been largely speculative on my part, and imposed by force of will
06:54 garu heh
06:54 Alias Well, and some experiences from trying to hack other people's code
06:54 garu well, I do know where each thing goes already
06:54 garu but I haven't touched padre core in a while now
06:54 Alias "What the fuck is some_function, and where the hell is it defined"
06:55 garu and could still figure it out rather easily
06:55 Alias There's a few fairly twisty bits, like the locking system
06:55 Alias But mostly those are hidden away behind "at least it's easy to use" APIs
06:55 garu heh
06:56 Alias The one problem that keeps trapping me still is when a refresh operation trigger by an unlock throws an exception
06:56 Alias Because we decided to just trap those exceptions and continue
06:56 Alias (Otherwise Padre just UPDATE locks forever and is dead)
06:56 garu ouch
06:56 Alias So now, stuff just silently doesn't happen
06:56 Alias Which is weird differently
06:57 Alias Although I think now that the unlock exceptions are at least sent to TRACE
06:57 Alias So you've got a chance to find them :)
06:58 garu one thing happened to me while I was debugging... I got a warning on the s/// in line 1304 of Padre::Wx::Main
06:58 garu but couldn't reproduce it steadily enough to debug it :/
06:58 Alias It happens
06:58 Alias garu: Oh, you should ask Gabriel to look at the directory tree, now it's stable
06:59 Alias And see if it lives up to his expectations
06:59 Alias (Even though there's some features missing)
06:59 garu hahahah will do
06:59 garu he'll probably get excited about it enough to work on it again
06:59 Alias Looks especially awesome when you use it on a project on a slow remote directory
06:59 Alias I have a massive work project on a samba drive that Padre has historically just choked on
06:59 Alias Now it works great
07:00 garu hmmm
07:00 garu wow
07:00 garu it loads my $HOME
07:00 Alias Try the search :)
07:01 Alias It's still a teeeensy bit rough around the edges, and I'm working on that
07:01 Alias But it's now usable
07:01 garu hahaha
07:01 garu awesome
07:01 garu I see the "Searching... $path" going crazy
07:01 Alias Yup :)
07:02 garu and it almost doesn't block :P
07:02 Alias (I admit I'm showing off there, slightly)
07:02 Alias The blocking is just too many events
07:02 Alias I'm going to make it chunk the directory tree updates into quarter second increments
07:02 Alias Which should mostly fix it
07:02 garu Alias++
07:02 Alias But like I said, it works
07:02 Alias The rest is just polish
07:02 garu it's definitely worth showing off :D
07:03 garu Alias: the old one GabrielVieira did, it had an option to avoid searching hidden dirs, no?
07:03 Alias Yeah, but it never really worked properly
07:03 Alias There's a concept similar to MANIFEST.SKIP built into the ::Project API
07:04 garu maybe we can make it smarter then, it could even share code with the new find in files option
07:04 Alias So the project can inform anything that wants to know which parts of a project are hidden
07:04 Alias Gabriel promptly ignored that API and duplicated something similar but different
07:04 Alias The new one ditches all the old stuff and just uses the "official" ignore rules from the project
07:05 Alias I haven't yet restored the option to make it ignore the ignore rules
07:05 garu cool
07:05 Alias I want to finish the performance hacks first
07:05 garu sounds like a plan
07:05 garu how does the ::Project figure out what to skip?
07:06 Alias That's up to the project
07:06 Alias I think the null project just does hidden/dot-dir
07:06 garu doesn't look like it :/
07:06 Alias ::Perl will do some smart things with MANIFEST or MANIFEST.SKIP if your project has one
07:06 garu search is scanning the hell out of my dotdirs
07:06 Alias And if not, it guesses using some heuristics
07:07 Alias hrm
07:07 Alias aaanyways
07:07 garu hehe
07:07 garu no worries
07:07 garu it's still awesome :D
07:07 Alias Look in the Project classes, there's an ->ignore_dirs_rule method or something
07:07 garu waaaaaaaaay better than the earlier version
07:07 garu I guess the new Task is really paying off as well
07:07 Alias But it means it's up to the project to work it out, and the directory tree code doesn't make any assumptions of it's own
07:07 Alias Which is better
07:07 garu yup
07:08 Alias The tree asks for a rules object from the project, and passes it into the file scanning task
07:08 Alias And the task applies the rule
07:08 Alias So only directories that pass the rule are sent to the tree for display
07:08 Alias And yes, the new Task API is helping a lot
07:08 garu I'll take a look when I wake up, sounds like I can tweak it to ignore hidden/dot by default
07:08 Alias That shiny status bar trick is one line of code
07:08 Alias Well, two
07:09 Alias The second one ditches the final directory name at the end
07:09 garu cool
07:09 garu hmmm
07:09 garu found a bug!
07:09 garu :P
07:09 Alias I'm sure
07:10 garu Alias: try searching for anything on a huge dir, and while the status bar is working, delete your search query
07:11 garu it won't update, and keep showing the last entry :)
07:11 garu I think I can fix it real quick, let's see
07:12 Alias hrm
07:12 Alias You need to clear if just after the task_reset
07:12 Alias it
07:12 Alias But tbh, don't bother
07:13 Alias There's a missing piece of functionality in the task api I overlooked
07:13 Alias And it's better off there
07:13 garu oh, ok
07:14 garu what's the missing functionality?
07:14 Alias There's no distinction between a finish hook and an abort hook
07:15 Alias The status stuff is triggered by a message from the task child to the owner of the task
07:15 Alias But the first thing that happens when you initiate a cancel is to task_reset, which severs the relationship between the child and the owner
07:15 Alias So all subsequent messages are discarded
07:15 garu right
07:15 Alias So we can't sent a "blank the status" from the cancel handler in the child
07:15 garu it's an abort
07:16 Alias Oh wait, maybe we can
07:16 Alias I added the status bar thing as a special magic handler
07:16 Alias Which might evade the owner entirely
07:16 garu we always can. question is, should we :)
07:16 Alias Try putting the status blank message at the moment it detects a cancel instruction in the child
07:17 garu yeah, statusbar.pm's methods appear to have no owner knowledge
07:17 * garu tries
07:18 garu where do you want the $self->SetStatusText( "", FILENAME ) ?
07:18 garu err, nevermind, found it
07:19 Alias You shouldn't be doing that yourself
07:19 Alias Should be something like $self->handler->message( STATUS => '' ) ?
07:19 Alias There's one in the task already
07:26 danlucraft joined #padre
07:34 garu hehe
07:34 garu EOMGIBROKEPADRE
07:38 garu Alias: it's strange... that call's already there after the while(@queue) on Padre::Wx::Directory::Search
07:39 garu I inadvertedly tried placing it under sub clear on Padre::Wx::Directory, but I got a Can't locate object method "handle" via package "Padre::Wx::Directory"
07:39 Alias oh
07:39 garu that's what I get for trying to debug things at 4:40 am :)
07:39 Alias That code was $self == task
07:42 garu should I clear it on Padre::Role::Task::task_reset then? Doesn't feel right to me
07:45 garu I see task_request, but doesn't look like a way to get the running task object so I can call ->handler on it
07:45 garu well, I'm gonna sleep on it tonight :)
07:49 Alias__ That ->handler should be done inside the child, not the parent
07:52 user_7917 joined #padre
07:52 user_7917 left #padre
07:58 Steffen joined #padre
08:10 daxim joined #padre
08:33 awnstudio left #padre
08:33 [1]awnstudio joined #padre
08:33 [1]awnstudio is now known as awnstudio
09:22 awnstudio left #padre
10:06 jazzanova left #padre
10:38 pece left #padre
10:44 pece joined #padre
10:50 pece left #padre
10:51 waxhead Alias, see this? http://blogs.wandisco.com/2010/09/08/why-we-got​-so-heavily-involved-in-the-subversion-project/
10:59 pece joined #padre
11:06 pece left #padre
11:44 zenog_ Alias__: If you want make changes to the Ack code in Padre, I will wait with the recursive search and replace feature ...
12:26 kthakore garu: you fixed both of the bugs. But basename still wil have the @_ overflow ... I suspect it has to do with \n or what not.
12:54 patspam joined #padre
13:15 kaare left #padre
13:37 bayashi is now known as bayashiaway
13:43 patspam left #padre
14:06 patspam joined #padre
14:23 lgtaube left #padre
14:33 williamt left #padre
14:42 Steffen left #padre
14:53 dapatrick joined #padre
14:53 Alias__ zenog_, wait for me. I'll drop the Find in Files code when it's working, and we can work on deriving Replace in Files for that
15:08 asarch joined #padre
15:20 zenog_ Alias__: great!
15:23 kthakore win 6
15:33 kthakore hi Alias__ still awake?
15:34 Alias__ just
15:39 Hyppolit #1007: warn output overwrites file name and other warns in Linux (assigned defect) [ http://padre.perlide.org/t​rac/ticket/1007#comment:4 ]
15:39 kthakore Alias__: whatcha up to?
15:39 Hyppolit #1007: warn output overwrites file name and other warns in Linux (closed defect) [ http://padre.perlide.org/t​rac/ticket/1007#comment:5 ]
15:40 kthakore garu++
15:40 kthakore holy smokes search is insane
15:41 kthakore does right work anymore in directory broswer?
15:41 Alias__ I culled some features
15:42 kthakore awww but I can't set the project folder no more
15:42 Alias__ Advanced settings
15:42 kthakore what do I look at?
15:42 Alias__ main_directory_root
15:42 Alias__ or something similar
15:43 kthakore where is advanced settings?
15:43 Alias__ Preferences dialog
15:43 Alias__ Then at the bottom, see Advanced...
15:43 kthakore ah ok
15:44 Alias__ btw, insane in a good way right? :)
15:45 kthakore yeah it is good
15:45 kthakore but it hangs the window a bit
15:45 Alias__ Yeah, I'm working on that
15:45 kthakore when it efreshes
15:45 Alias__ Fires a few too many events
15:45 Alias__ I just need to put in some chunking
15:45 kthakore yeah
15:45 kthakore oh ok
15:46 Alias__ Bunch up the reported files and only render 4 or 5 times a second max
15:46 Alias__ Should be enough
15:46 asarch How can I know if a plugin is already installed on my system?
15:46 Alias__ asarch: Plugin manager
15:46 kthakore asarch: see your plugin manager
15:46 asarch Thank you guys
15:46 asarch I only have three plugins
15:47 Alias__ Those would be the three that ship by default
15:47 mj41 left #padre
15:47 Alias__ One for feedback, one to support development of Padre itself, and one toy plugin so you can try writing your own
15:48 kthakore Alias__: so when do we get back project features?
15:48 Alias__ project features?
15:48 Alias__ Like what?
15:50 Alias__ sorry, bed time
15:50 kthakore the directory reset stff
15:50 kthakore stuff
15:50 kthakore Alias__: ok buy
15:50 Alias__ buy -> bye
15:50 Alias__ ;)
15:50 * Alias__ &
15:51 mj41 joined #padre
16:01 lgtaube joined #padre
16:02 GabrielVieira left #padre
16:02 GabrielVieira2 joined #padre
16:15 danlucraft left #padre
16:47 skeeterbug joined #padre
16:50 asarch cpan> install Padre::Plugin::Nopaste
16:50 asarch "Error GETing http://localhost/#darkstar:47835/: Can't connect to localhost:80 (connect: Connection refused) at t/local/back.t line 54"
16:50 asarch With this package: PETDANCE/WWW-Mechanize-1.66.tar.gz
16:59 daxim tests run fine for me
17:01 daxim left #padre
17:28 danlucraft joined #padre
17:30 kanishka joined #padre
17:38 rindolf joined #padre
17:38 rindolf Hi all.
18:52 GabrielVieira joined #padre
18:53 GabrielVieira2 left #padre
19:37 danlucraft left #padre
19:44 asarch left #padre
19:51 [Sno] joined #padre
20:02 kanishka left #padre
20:10 chorny joined #padre
21:04 rindolf left #padre
21:04 perlite left #padre
21:04 perlite joined #padre
21:13 [Sno] left #padre
21:15 [Sno] joined #padre
21:22 [Sno] left #padre
21:42 [Sno] joined #padre
22:06 [Sno] left #padre
22:16 epac joined #padre
22:52 epac hello.
22:52 epac i would like to register an account with TRAC to input a ticket.
22:53 epac i am having issues with installing padre from cpan (error with file::Spec catdir sub...)
22:54 chorny Do you have issues with Pare dist or with one of dependencies?
22:55 epac all deps seems to have installed fine. a number of the tests are failing...
22:55 epac i just realized that i may have used the 32bit strawberry on a 64bit system though..
22:55 epac i'll poke a bit more...
22:56 epac the first test to fail is t/06-utils.t with "Use of uninitialized value $_[0] at perl/lib/File/Spec/Win32.pm line 152.
22:56 epac quite a few of those throughout a few of the tests.
22:57 chorny Do you use Portable version or normal Strawberry?
22:57 epac portable
22:57 chorny Which version?
22:57 epac and i think the 32bit portable on a x64 system
22:57 epac 5.12.1.0
22:57 epac i downloaded it friday i believe.
22:58 chorny You need to edit portable.perl
22:58 chorny file inside Portable
22:58 epac oki
22:59 chorny Find HomeDir entry and replace it with:
22:59 chorny HomeDir:
22:59 chorny my_home: data
22:59 chorny my_data: data
22:59 chorny my_documents: data
22:59 chorny also create directory "data"
22:59 epac in perl? or in my ~
22:59 chorny in perl
22:59 chorny Did you run cpan client from portableshell.bat ?
22:59 epac yes, i did
23:00 chorny Now you can try again
23:00 epac :)
23:00 epac will do
23:01 epac maybe this is a stupid question, but why is the MSI not relocatable, if the portable edition is?
23:02 epac or rather, would it not be possible to wrap the portable edition in a MSI file, and update the couple of files after install to make it work ?
23:04 chorny it is relocatable from 5.12
23:22 epac chorny: it just finished compiling padre...
23:22 epac looks good.
23:23 epac and i'm running..
23:24 epac so the MSI for 5.12... i could for example, harvest all the files from the portable zip, edit the portable.perl file,  and setup the PATH in the wxs file...
23:25 epac and that would work as a fully relocatable perl msi?
23:25 epac or is the msi you have now for download be installed anywhere?
23:29 chorny standard Strawberry requires installation, can be installed into any directory and can't be moved after this
23:29 epac cool.
23:29 epac thanks for you help chorny...
23:31 epac left #padre
23:33 [1]awnstudio joined #padre

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