IRC log for #padre, 2009-12-23

Time	Nick	Message
00:21		awnstudio joined #padre
00:25	Hyppolit	svn: r9880 \| adamk++ \| http://padre.perlide.org/trac/changeset/9880
00:25	Hyppolit	Lock (both update and refresh) during _time_post_init to prevent useless GUI refreshing during startup.
00:25	Hyppolit	trunk/Padre/lib/Padre/Wx/
00:26		Alias joined #padre
00:26	Alias	Can people please do a speed comparison of startup before and after that change
00:49	kthakore	Alias: sure
00:49	kthakore	Alias: with dev.pl?
00:52	kthakore	Alias: before real 0m6.535s after real 0m6.221s
01:39	garu	patspam++
01:39	garu	which plugin are you developing, btw? :)
02:05	Alias	Third of a second, not bad
02:06	Alias	Not incredible, but not bad
02:21		kthakore joined #padre
02:22	kthakore	garu: hiya!
02:22	kthakore	garu: going to sleep see ya!
02:22	Alias	kthakore: Thanks for the test
02:23	Alias	Was that with just the default new file open?
02:26		Casan joined #padre
03:12	Hyppolit	svn: r9881 \| patspam++ \| http://padre.perlide.org/trac/changeset/9881
03:12	Hyppolit	Padre-Plugin-Plack version 0.01 release
03:12	Hyppolit	trunk/ trunk/Padre-Plugin-Plack/ trunk/Padre-Plugin-Plack/lib/ trunk/Padre-Plugin-Plack/lib/Padre/ trunk/Padre-Plugin-Plack/lib/Padre/Document/ trunk/Padre-Plugin-Plack/lib/Padre/Plugin/ trunk/Padre-Plugin-Plack/lib/Padre/Plugin/Plack/ trunk/Padre-Plugin-Plack/share/ trunk/Padre-Plugin-Plack/share/dot-psgi/ trunk/Padre-Plugin-Plack/share/dot-psgi/frameworks/ trunk/Padre-Plugin-Plack/share/dot-psgi/static/ trunk/Padre-
03:17	* patspam	wanders off to make a cup of hot chocolate
03:17	patspam	got a blog post written for that new plugin, but need to wait for the next version of Padre first ;)
03:18	* Alias	makes new locking API aware of when Padre is shutting down, and extends it into more niches
03:18	Alias	Startup and shutdown should be (somewhat) noticably faster
03:20	Hyppolit	svn: r9882 \| adamk++ \| http://padre.perlide.org/trac/changeset/9882
03:20	Hyppolit	Added support for a "shutdown" mode to the locking API.
03:20	Hyppolit
03:20	Hyppolit	Moved parts of startup and shutdown to the new locking API, and integrated the refresh of the directory browser.
03:20	Hyppolit	trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Wx/
03:21	Ryan52	Padre-Plugin-Plack?
03:22	Ryan52	ah.
03:40		ppant joined #padre
04:14		szabgabx joined #padre
04:39	Alias	hrm
04:39	Alias	Hitting Enter has gotten slow
04:39	Alias	Is someone doing something more complex than they used to?
04:42		yakudzo joined #padre
04:42	yakudzo	Hello everyone.
04:42	szabgabx	gm
04:42	yakudzo	Can padre hightlight JS??
04:43	szabgabx	it should, I think
04:43	szabgabx	but maybe the default is broken
04:45	yakudzo	what I see now is very poor highlighting. I thought maybe I did not added some thing or is there posibility to move hightlight configs from other application ?
04:45	Alias	Define "poor"
04:46	szabgabx	you can try to install the Kate plugin and see use the Kate highlighter for Js
04:46	yakudzo	its only three colors. comments/strings/all other stuff
04:47	Alias	It's possible that nobody has bothered to set up a decent colour scheme, but the highlighting itself supports more
04:47	szabgabx	probably what Alias says
04:47	szabgabx	check out the the share/styles/default.yml
04:48	szabgabx	I don't see any special definitions for JS
04:48	yakudzo	thanks
04:48	szabgabx	you could patch that to your liking and send give it to us to add it to the core
04:50	szabgabx	Alias, btw recently I saw at least two projects that ask for donations
04:50	szabgabx	Ubuntu and Eclipse
04:50	szabgabx	and they both have quite a lot of money from corporate as well
04:50	Alias	szabgabx: Did you see how much they actually RAISED though?
04:50		jayk joined #padre
04:50	Alias	From the donate thing?
04:51	szabgabx	no
04:51	Alias	I have some numbers :)
04:51	Alias	Wikipedia has 340 million users per month
04:52	Alias	Users, not pages
04:52	Alias	They raise, with some effort, $10,000,000 per year in donations
04:52	Alias	Based on my observations, it usually seems to take them several months per donation run
04:53	Alias	So for something that people find useful and use every day, all over the world, they raise approximately 3 cents per user
04:53	Alias	Annually, for 3-4 months of fairly grindy fundraising work
04:54	Alias	Looking at that screamer program I use, which raised $2,500 over 5-10 years, I can imagine 100,000 users
04:54	szabgabx	we talked about this already I just found this to be another point
04:54	Alias	yup
04:55	Alias	I'd meant to mention the wikipedia thing earlier as well
04:55	Alias	Firefox makes in the single dollars per user, for the default Google search, if I recall
04:55	yakudzo	I think not all 340 million people donated to wikipedia. almost 95% of them did not even thinked about it
04:56	Alias	yakudzo: The same would almost certainly be true for Padre
04:56	Alias	yakudzo: Thus the comparison
04:56	Alias	Although we are perhaps more valuable than wikipedia in the lives of our users, so maybe we could average a little higher
04:57	szabgabx	Alias, what is this "Apply diff to file" menu option?
04:57	szabgabx	Apply a patch ?
04:57	Alias	It's a feature someone promised to write if I added the menu entry
04:57	Alias	But yes, apply a patch file to a file, and apply a patch file to a project
04:57	Alias	Because, right now, nobody can send people on Windows a patch file and have it easily applied
04:58	Alias	I wanted a simple sanity-checking wrapper for diff/patch that would make sure the patch is applied properly
04:58	Alias	Without having to type patch -d -O34 -Pwhatever blah
04:58	Alias	While being in the "right directory"
04:58	Alias	Apply patch -> Select File -> done
04:59	Hyppolit	svn: r9883 \| adamk++ \| http://padre.perlide.org/trac/changeset/9883
04:59	Hyppolit	A few more minor additions to the locking system.
04:59	Hyppolit
04:59	Hyppolit	Renamed set_title to refresh_title, since it is really a refresh method
04:59	Hyppolit	trunk/Padre/lib/Padre/Wx/
05:00		waxhead joined #padre
05:00	* Alias	is really happy with the new locking stuff
05:00	Alias	Opening and closing lots of files noticably faster
05:00	Alias	More work to go though
05:01	szabgabx	and I am quite happy with the comments of the menu items showing up on the status bar
05:02	Hyppolit	svn: r9884 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9884
05:02	Hyppolit	add comments to the menu entries of the Edit menu
05:02	Hyppolit	trunk/Padre/lib/Padre/Wx/Menu/
05:02	szabgabx	it creates more work to the translators :-)
05:02	* waxhead	is make test -ing latest trunk...
05:02	szabgabx	oh waxhead is here!
05:03	waxhead	szabgabx, so I see.. from 4 green perfects last night to none...
05:03	Hyppolit	svn: r9885 \| adamk++ \| http://padre.perlide.org/trac/changeset/9885
05:03	Hyppolit	Updating changes for the locking stuff
05:03	Hyppolit	trunk/Padre/
05:03	waxhead	now I know what's happened.. :)
05:03	waxhead	are we OK for a release tonight?
05:03	waxhead	no major issues anyone is working on/aware of?
05:03	Alias	I would like to see the Slow-Enter bug fixed, but otherwise yes
05:03	Alias	Open a large file, put the cursor somewhere in the middle of the file, and hit enter to add a new line
05:04	Alias	LAG
05:04	Alias	Everything else is fast... deleting files is fine...
05:04	Alias	Just hitting enter
05:04	Alias	er, deleting lines
05:04	Alias	It only appeared today, I think
05:04	Alias	oops
05:04	szabgabx	I don't feel that lag
05:04	Alias	And I think I just made saving a file crash Padre :)
05:05	Alias	But I know where that is
05:05	waxhead	cool.. that's a handy feature
05:06	Alias	It's a security feature
05:06	Alias	Prevents corrupting files
05:06	Hyppolit	svn: r9886 \| adamk++ \| http://padre.perlide.org/trac/changeset/9886
05:06	Hyppolit	Thinko bug in refresh_title
05:06	Hyppolit	trunk/Padre/lib/Padre/Wx/
05:06	szabgabx	wait, we could connect that with the donation system
05:07	Alias	"Please donate another $5 to enable saving files"
05:07	szabgabx	:)
05:07	Alias	I think they used to call that Shareware, right? :)
05:08	Alias	I wonder if I should add SQLite integration to the locking system
05:08	Alias	I wonder if we actually use transactions anywhere
05:10	szabgabx	I am taking the kids to school and then try to add comments to the View menu as well and then see if I can convert them to actions
05:10	Alias	Be careful
05:11	szabgabx	but this should not be an issue with the release
05:11	Alias	Those ticked menu entries can be finicky
05:11	Alias	Menu entries that don't enable/disable should be fine
05:11	szabgabx	I'll check if they still work after I make the changes
05:15	waxhead	what about long comments? Do they present a problem?
05:16	Alias	Gah, recursive transaction fail
05:16	Alias	OK, definitely need to add Padre::DB to the locking system
05:17	* Alias	is still annoyed at stupid screen flicjer
05:17	Alias	flicker
05:17	Alias	I can't find a way to kill it off, but at least the whole thing is faster now
05:18	waxhead	Alias, I don't see the flicker... that I recall
05:18	waxhead	is it a windows only issue?
05:18	waxhead	or lots of files or big files?
05:18	Alias	It's entirely possible it's Windows-only
05:19	Alias	Flicker-type stuff glitches are often GUI backend specific
05:19	waxhead	oh.. sweet.. szabgab++ for comments in status bar..
05:19	Alias	I see it when I do File -> Open -> 5 or 10 or 20 files
05:19	Alias	One flicker per file
05:19	* waxhead	is thinking about getting a nokia N900
05:19	waxhead	when you open the file?
05:20	Alias	right
05:20	Alias	Happens even with one
05:20	waxhead	ok, I"ll see what happens.. are they big?
05:20	Alias	But more noticable if you multi-select 20 files in the open dialog
05:20	Alias	Size doesn't seem to matter
05:20	Alias	Size makes loading slower, but the flicker doesn't seem to be longer
05:21	waxhead	OK, I'll give it a crack...
05:23	waxhead	Alias, I openned nearly all the files in lib/Padre/ no flicker.. just a long time openning...
05:24	waxhead	Padre faded to grey to show it was busy/not responding... but all files opened in tabs at the same time...
05:26	submersible1	Alias: I see some odd syntax flicker when using a dark theme. Like a block redraw of white before the highlighted block is written over it
05:26	waxhead	hey submersible1
05:27	Alias	submersible: I'm not entirely convinced that Wx's WindowUpdateLocker correctly prevents painting of everything
05:27	Alias	I suspect the AUI layout stuff
05:28	Alias	I seem to remember that it was pretty smooth this time last year, when we were using vanilla tabs
05:28	Alias	I've also got a flicker issue on my machine, where I can see the new tab at 100x100 pixels on the top left of the editor panel, before quickly resizing itself to full size
05:28	Alias	Which I swear should not be happening
05:29	Alias	Maybe we need to do a Hide/Show fidget in that case
05:29	waxhead	awesome...Can't locate object method "document" via package "Padre::Wx::Main" at /home/pete/Programming/Perl/padre/trunk/Padre/lib/Padre/Wx/Main.pm line 1045.
05:29	Alias	You're a revision behind man
05:29	Alias	svn update
05:32	waxhead	cool... that was what I figured was the problem...
05:34	waxhead	'check one thing and use another'
05:35	waxhead	there's something weird about the crtl-tab too
05:35	waxhead	lots of files open no flicker
05:35	Alias	Yeah, control flow in ctrl-tab is weird
05:36	waxhead	it was simply last open, but got changed to last used...
05:37	waxhead	which is ok, but it's that really odd crtl tab the new tab comes up then it seems to tab somewhere else before it moves throught he tabs again...
05:42	Alias	I agree, whatever the new behaviour is, I don't understand it
05:51	Alias	OK, more pointless refresh code removed
05:52	Alias	Curiously, even though I think I've now expunged much of the useless stuff, things are still slower than I would live
05:52	Alias	like
05:52	Alias	I wonder if DProf will work with PAdre
05:59	Alias	I'm guessing it probably won't :)
06:00	Alias	Wow
06:00	Alias	OK, so I'm profiling
06:00	Alias	I open Padre, immediately open 10 files in one dialog, then hit exit
06:00	Alias	Guess what consumes the most CPU (other than the event loop and dialog modal stuff)
06:01	waxhead	no idea what?
06:01	Alias	DBI::db::commit
06:02	Alias	1.2 seconds
06:02	Alias	Followed by Text::FindIndent::_check_vim_modeline
06:02	Alias	and Text::FindIndent::parse
06:02	Alias	whoa
06:02	Alias	_check_vim_modeline gets called 13,000 times
06:02	Alias	:)
06:03	Alias	Regexp::DESTROY(xsub) called 65,000 times, wtf
06:06	Alias	I missed DBI::st::execute(xsub)
06:06	Alias	3.something seconds
06:07	Alias	So for around 16 second runtime of Padre, with about 6 of that in the event loop waiting for input, we spend 4-5 seconds in SQLite
06:07	Alias	And almost a second in Text::FindIndent
06:12	waxhead	tha'ts a lot really
06:12	Alias	# 11 times (0s+1.17s) by Padre::DB::History::insert
06:13	Alias	That's 100ms per file
06:13	Alias	Which seems like a hell of a lot
06:13	Alias	100ms is half the entire user-noticability-time budget
06:13	Alias	# 11 times (0s+828ms) by Padre::DB::LastPositionInFile::insert
06:14	Alias	# 22 times (15.6ms+0s) by Padre::DB::SessionFile::insert
06:14	Alias	THAT is interesting
06:14	Alias	Because I'm not even using sessions
06:14	Alias	Not to mention it's double the number of files I opened
06:22	Alias	A fairly large chunk of startup is spent doing Padre::DB compilation as well ...
06:22	Alias	Like, a second? That seems wrong
06:22	Alias	I might need to take a look at caching the ORLite-generated code
06:26	Hyppolit	svn: r9887 \| adamk++ \| http://padre.perlide.org/trac/changeset/9887
06:26	Hyppolit	Tweaks to locking, so we don't multiply update the Recent Files menu during multiple-file opens
06:26	Hyppolit	trunk/Padre/lib/Padre/Wx/
06:36	yakudzo	When I select some text in KDE 4.3 (e.g in browser) that go back to padre and paste it with middle button on mouse it pastes me two buffers one that I selected and one that was saved by ctrl+C
06:38	waxhead	yakudzo, can you raise a ticket for it?
06:38	yakudzo	I'll try (:
06:38		azawawi joined #padre
06:39	waxhead	yakudzo, http://padre.perlide.org/trac/
06:40	waxhead	hey azawawi
06:42	azawawi	good morning
06:43	azawawi	waxhead: im afraid i will not be able to finish what i was working on for 0.53... maybe 0.54 new year edition :)
06:43	Alias	hi
06:43	Alias	azawawi: What were you working on?
06:43	azawawi	Alias: Perl 6 refactoring...
06:44	azawawi	Alias: exact refactoring based on STD
06:45	azawawi	Alias: and i finished some of my remaining TODO (already committed) for open resource and quick menu access
06:45	yakudzo	seems that in the middle of disscussion of http://padre.perlide.org/trac/ticket/390 people already found this bug. but did not noticed
06:46	azawawi	Alias: anything special for new year?
06:47	waxhead	yakudzo, that is an old bug.. now that we have some Debugger support in Padre maybe we'll get someone to take a look at this again...
06:48	waxhead	azawawi, no worries... as long as we have a working trunk we'll be fine...
06:48	yakudzo	then I'll just add the comment that bug still exists (:
06:48	waxhead	we might do a 0.54 before new year too...
06:48	waxhead	yakudzo, thanks.. that will be great.. I'll see if I can keep it on my todo list to prod some of the devs to take another look at htis...
06:49	Alias	azawawi: For Padre or my social life?
06:49	Alias	Mostly I'm working on the new locking API
06:49	Alias	Trying to make Padre faster
06:49	Alias	It's feeling too sluggish for my tastes
06:49	azawawi	Alias: for Padre :)
06:50	Alias	It's starting to get a teensy bit more complex than I'd like already
06:50	Alias	But this is exploratory coding for me, so that's ok
06:50	Alias	I'll refactor it again in 6 months I guess
06:50	Alias	And we're definitely seeing speedups, even from modest usage
06:53	szabgab	yakudzo, you could also fix it :-)
06:55	Hyppolit	#805: Command line and gui integration with RAD (new enhancement) [ http://padre.perlide.org/trac/ticket/805 ]
06:55	submersible1	yakudzo: or bug someone else to fix it :)
06:55	waxhead	ARGH!!!
06:56	yakudzo	((:
06:57	submersible1	the whole middle button paste thing is getting to annoy me enough to go in there and look.
06:57	waxhead	submersible1, good luck... !! :)
06:58	Alias	ugh
06:58	waxhead	it's spooky in there...
06:58	Alias	The whole event handler spagetti needs reworking
06:58	Alias	Or at least, the main window/editor keystroke/mouse stuff
06:58	yakudzo	I need to register to paste to tickets
06:58	submersible1	waxhead: I'm thinking more along the lines of examining it's genesis.. ie finding the commit that introduced it
06:58	azawawi	szabgab*: debugger blocks UI on win32
06:58	waxhead	this auto complete method thing is frustrating the daylights out ouf me...
06:58	yakudzo	in trac
06:59	waxhead	yakudzo, thanks.. we'll see if someone can take a look at it.. clearly it's going to be someone who finds it annoying enough to get motivated enough to track it down..
06:59	szabgab	trac?
06:59	Hyppolit	trac is http://padre.perlide.org/trac and also http://padre.perlide.org/trac/register
06:59	* waxhead	doesn't use middle mouse button for pasting
06:59	azawawi	waxhead: which one? the one that completes the sub ?
06:59	szabgab	yakudzo, go ahead register I opened the registration
07:00	waxhead	azawawi, yeah.. it seems something has changed and you only type on char and it auto completes
07:00	waxhead	might have been something I did here too..
07:00	azawawi	waxhead: yeah it is definity annoying... the best thing for autocompletion, is to show a list if the user wants it
07:01	azawawi	waxhead: automatic auto-completion can sometimes screw things up... :)
07:01	* Alias	hates auto-completion
07:01	waxhead	azawawi, I've added in an option in the preferences to turn it on for "scripts" as well
07:02	* waxhead	likes it
07:02	yakudzo	szbalint: thanks registered.
07:02	Alias	It just encourages people to write longer and uglier/pointlessly-long method names
07:02	Alias	foo_bar_method
07:02	waxhead	as opposed to fbm
07:02	Alias	Note: People actually put pointless words in like "method" more often that you'd think
07:02	waxhead	yeah, like people prefix their database tables with tbl_
07:02	Alias	I think I shall dub it "Two Word Method Syndrome"
07:03	Alias	waxhead: Nah, that's consistent
07:03	Alias	The thing about TWMS is that they put in RANDOM extra words, and randomly either at the front or back
07:03	Alias	initiate_foo
07:03	waxhead	thanks for the useless 4 extra chars I have to type.. 5 keystrokes in total to query a TABLE I already know is a table
07:03	Alias	whatever_enable
07:04	Alias	But yeah, tbl_ can be annoying
07:04	waxhead	it is, I kick myself every now and then for doing it myself
07:04	Alias	Having done some code generators in my time, I suspect a lot of it is to make things like Microsoft Access (with lots of auto-generated stuff) easier
07:04	waxhead	not any more though
07:05	Alias	If you make a table called a "name" and an index/whatever called a "name", and the database wants unique resource names, wth are you supposed to do
07:05	waxhead	I think it's more to do with languages like vb/script that you see the convention oObject or strString
07:05	Alias	Clearly you don't want Access-thingy to go "You can't call that index 'name', you called the table 'name'"
07:05	waxhead	Alias, idx_name
07:06	waxhead	because it's the computer that deals iwth the index not ME
07:06	waxhead	I type select * from table; more than I do alter table drop index idx_name;
07:07	Alias	Right, but "more" doesn't come into it when you are building a tool for arbitrary numbers of other people
07:07	Alias	It either works, or it blows up for some subset of users
07:07	Alias	Anyways
07:08		marcela joined #padre
07:08	Alias	I agree though, the shortname prefix stuff sucks, and it would be nice if it wasn't exposed to the user
07:09	Hyppolit	svn: r9888 \| adamk++ \| http://padre.perlide.org/trac/changeset/9888
07:09	Hyppolit	Integrated Padre::DB transactions into the Padre::Locker API.
07:09	Hyppolit	We can now do basic (commit or die, no partial rollback) nested transactions.
07:09	Hyppolit	trunk/Padre/ trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Config/ trunk/Padre/lib/Padre/DB/ trunk/Padre/lib/Padre/Wx/ trunk/Padre/lib/Padre/Wx/Dialog/
07:10	Hyppolit	svn: r9889 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9889
07:10	Hyppolit	add comments to the View menus
07:10	Hyppolit	trunk/Padre/lib/Padre/Wx/Menu/
07:12	Hyppolit	svn: r9890 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9890
07:12	Hyppolit	enable warning when a menu does not have a comment
07:12	Hyppolit	trunk/Padre/lib/Padre/Wx/
07:12	Hyppolit	#390: Copy and Paste onto highlighted text (accepted defect) [ http://padre.perlide.org/trac/[…]et/390#comment:10 ]
07:16	Alias	Holy shit
07:16	Alias	I think I found Padre's missing performance
07:17	waxhead	Alias, cool.. how?
07:17	szabgab	removed the sleep (1) call?
07:17	waxhead	Alias, want to blog about it to give numpties like me a clue on how you do such a thing?
07:18	Hyppolit	svn: r9891 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9891
07:18	Hyppolit	remove old commented out code
07:18	Hyppolit	trunk/Padre/lib/Padre/Wx/
07:18	szabgab	crap tests fail
07:20	Alias	waxhead: Yeah, I will. I think these changes justify a dedicate post in addition to the release post
07:21	Alias	dedicated
07:21	Alias	Shutdown is so fast you may never see the file closing progress bar again :)
07:22	waxhead	wow.. serious???
07:22	waxhead	what was it?
07:22	Alias	Padre::DB->begin
07:22	Alias	It's the database stuff to save the "last session" and "file position"
07:22	Alias	SOME of it was using transactions
07:23	Alias	But normal Padre::DB->begin/commit stuff doesn't nest
07:23	szabgab	Variable declared in conditional statement at line 68, column 2. Declare variables outside of the condition. (Severity: 5)
07:23	Alias	So by adding it to the locking API we get basic nesting
07:23	szabgab	Perl::Critic found these violations in "blib/lib/Padre/Config/Host.pm":
07:23	Alias	Which means I just add a giant ->lock('DB') at the beginning of ->on_close_window
07:23	Alias	And whenever the hell that guard object expires, it commits
07:24	Alias	Which is AFTER the ::Main->Hide
07:24	Alias	So Padre both shuts down faster, and APPEARS to shut down superfast
07:24	Alias	If the shutdown process is user-cancelled, the guard object will just commit the parts that WERE completed
07:24	szabgab	Alias, please fix that conditional assignment
07:25	Alias	szabgab: Let me get this speed thing committed
07:27	Hyppolit	svn: r9892 \| adamk++ \| http://padre.perlide.org/trac/changeset/9892
07:27	Hyppolit	Wrap extra (nested) transactions around major startup/shutdown/open-many/close-many actions.
07:27	Hyppolit	Padre not only IS faster at these operations, but shutdown in particular APPEARS to be super fast.
07:27	Hyppolit	trunk/Padre/lib/Padre/Wx/
07:28	Alias	Righto
07:28	Alias	svn update
07:28	Alias	And tell me what you think
07:29	Alias	Try opening 10 or 20 files, then hitting window close
07:30	Alias	Text::FindIndent is still slowing file opening down
07:30	Alias	But we can address that next
07:30	waxhead	what do i use to get svn to tell me what files have changed?
07:30	Alias	svn diff -R1234:1235 or something
07:30	Alias	man svn
07:30	Alias	:)
07:31	Hyppolit	svn: r9893 \| azawawi++ \| http://padre.perlide.org/trac/changeset/9893
07:31	Hyppolit	Two translation fixes
07:31	Hyppolit	trunk/Padre/lib/Padre/Action/
07:31	waxhead	svn status does it
07:31	Alias	oh, changed between when you did an update?
07:31	Alias	As in locally changed?
07:31	* Alias	runs full test suite to keep szabgab happy
07:32	Hyppolit	svn: r9894 \| waxhead++ \| http://padre.perlide.org/trac/changeset/9894
07:32	Hyppolit	Added a configuration option for "auto complete scripts"
07:32	Hyppolit	trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Document/ trunk/Padre/lib/Padre/Wx/Dialog/Preferences/
07:33	Alias	auto-complete scripts?
07:33	Alias	Wassat?
07:33	waxhead	it's an option you might want when working in a class, but not a plain of script
07:33	waxhead	s/of/old/
07:33	waxhead	need to do this auto pod thing too.
07:35	Alias	hmm
07:35	Hyppolit	svn: r9895 \| azawawi++ \| http://padre.perlide.org/trac/changeset/9895
07:35	Hyppolit	Another translation fix
07:35	Hyppolit	trunk/Padre/lib/Padre/Wx/Menu/
07:35	Alias	You know, if I worked out a way to disable last-session tracking entirely when not needed, I bet Padre would get faster file opening
07:36		kaare joined #padre
07:36	Alias	oooorrr....
07:36	Alias	hrm
07:36	* Alias	digs
07:39	Hyppolit	svn: r9896 \| waxhead++ \| http://padre.perlide.org/trac/changeset/9896
07:39	Hyppolit	Remove the pod insertion until an option is available for it.
07:39	Hyppolit	trunk/Padre/lib/Padre/Document/
07:39	* szabgab	attempting to convert file menu to be actions
07:39	szabgab	waxhead, how long till you kick in the release process?
07:39	Alias	I thought it was....
07:40	* Alias	wants another half an hour or something for speed work
07:40	waxhead	szabgab 1000 UTC
07:40	waxhead	that's 9pm here
07:40	waxhead	someone asked about it on the mailing list so I set a time
07:40	szabgab	2:20 more ?
07:40	waxhead	I'm happy to leave it until tomorrow if you like
07:40	szabgab	nope
07:41	waxhead	I have the day off tomorrow.. and all the days until the 4th of Jan... :)
07:41	Alias	Be better to do it earlier
07:41	waxhead	ok.. then it's 1000 UTC
07:41	szabgab	I just want to have a feeling of when to make the commits that ruin everything ?
07:41	Alias	Menu -> Action conversions have a history of bugs
07:41	Alias	I wouldn't mind seeing a release before too much more is converted
07:41	waxhead	which is 9pm .au time.. actually that DST time in .au
07:42	waxhead	not now...
07:42	Alias	waxhead: Yeah, some people and their non-delta time discussions
07:42	waxhead	9pm...
07:42	szabgab	none of theses timestamps help me, I need an ETA
07:42	waxhead	it gives the translators some extra time...
07:42	waxhead	ETA?
07:42	Alias	Clearly such people haven't really understood the concept of interweb time
07:42	Alias	waxhead: "We're releasing at +2:20h"
07:42	szabgab	or some other 3 letter word
07:42	waxhead	oh right..
07:43	Hyppolit	svn: r9897 \| azawawi++ \| http://padre.perlide.org/trac/changeset/9897
07:43	Hyppolit	Another translation fix
07:43	Hyppolit	trunk/Padre/lib/Padre/Wx/Menu/
07:43	waxhead	2 hours and 20 minutes
07:43	Alias	"See you in the morning" becomes "See you in +10h"
07:43	Alias	etc
07:43	waxhead	got it..
07:43	szabgab	we could add a clock to Padre showing the minutes left till the release starts :-)
07:43	* waxhead	is off to get the laundry in
07:44	Alias	You spend enough time in AU with most of the world somewhere between +6 and +19 hours different and you adapt to delta fastt :)
07:44	Alias	Everything is a third of a day different
07:44	szabgab	crap, azawawi is changing the file beneath my change
07:44	Alias	szabgab: Remind me at some point to have a discussion about special event code :)
07:44	szabgab	not good
07:45	Alias	I've been pondering how the hell to do a "Christmas Edition" or whatever
07:45	Alias	And it's HARD without being able to show 5 meg of audio files into the tarball :)
07:45	Alias	shove
07:45	azawawi	szabgab: which one File.pm?
07:45	* Alias	has been pondering some kind of "Downloaded Content" directory
07:45	azawawi	szabgab: translation fixes... nothing major
07:45	szabgab	Menu::File.pm
07:46	Alias	Where Padre will proactively start pulling media content a week before a special event
07:46	szabgab	yes, ok, I just was moving the content of that file to another file
07:46	Alias	And stashing it somewhere
07:46	szabgab	but never mind, I'll try to carry your changes over
07:47	Hyppolit	svn: r9898 \| adamk++ \| http://padre.perlide.org/trac/changeset/9898
07:47	Hyppolit	Making Perl::Critic happy again
07:47	Hyppolit	trunk/Padre/lib/Padre/Config/
07:48	szabgab	rapdily approaching rev 10,000
07:48	Alias	yup
07:48	Alias	Should comfortably make it before NYE
07:48	Alias	At the current rate of around 50 per 16 hours
07:49		user_4515 joined #padre
07:52	szabgab	there is certainly more work in the actions code to make it really configurable
07:53	szabgab	and the menu refresh code also makes things a bit hard-coded
07:56	szabgab	hi user_4515 !
07:57	szabgab	can we help you?
07:57	szabgab	:-)
07:57	waxhead	ha.. scared off by your enthusiasism
07:57	waxhead	sheesh.. or my shocking spelling
07:58	waxhead	release in +2:02
07:59	waxhead	szabgab, how does one use the debug client?
07:59	szabgab	open a file
07:59	szabgab	a script
08:00	Alias	szabgab: yeah, the action stuff is rather dangerous
08:00	waxhead	szabgab, good material for a blog post.. :)
08:00	Alias	It's going to take a while to find a model that works, I think
08:00	Alias	Also, there's some speed issues
08:00	szabgab	click on one of the step buttons
08:00	szabgab	and keep clicking on other buttons :-)
08:00	Alias	Populating the actions is actually O(quadratic)
08:00	Alias	It gets slower for each action
08:01	Alias	Actually
08:01	Alias	That generalises to O(n^2)
08:01	Alias	So yeah, not great
08:01	Alias	Mostly because it does duplicate checking by iteration
08:02	waxhead	what login do you use for ohloh? The one it "sees" you as in a project?
08:03	szabgab	you create a login and then associate usernames (of commits) with it
08:03	waxhead	szabgab thanks
08:04	Hyppolit	svn: r9899 \| jquelin++ \| http://padre.perlide.org/trac/changeset/9899
08:04	Hyppolit	updated french translation
08:04	Hyppolit	trunk/Padre/share/locale/
08:11		pece joined #padre
08:15	waxhead	release in +1:45
08:16	yakudzo	today new release ?
08:17	szabgab	for some people that's already tomorrow :-)
08:17	szabgab	or maybe that's the fishes in the Pacific ?
08:18	szabgab	yakudzo, anyway, yes, that's the plan
08:26	Hyppolit	svn: r9900 \| azawawi++ \| http://padre.perlide.org/trac/changeset/9900
08:26	Hyppolit	Translation fix
08:26	Hyppolit	trunk/Padre/lib/Padre/Wx/Menu/
08:28	Hyppolit	svn: r9901 \| azawawi++ \| http://padre.perlide.org/trac/changeset/9901
08:28	Hyppolit	Updated Arabic translation
08:28	Hyppolit	trunk/Padre/share/locale/
08:32	szabgab	ok, I converted the File menu to Actions but I think I'll wait till the release and commit it only after that
08:33	Alias	Unexpected breakages in the File menu would be rather high profile, yes :)
08:34	szabgab	this would be expected breakage :-)
08:34	Alias	heh
08:36	Hyppolit	svn: r9902 \| azawawi++ \| http://padre.perlide.org/trac/changeset/9902
08:36	Hyppolit	Another translation fix and updated Arabic translation
08:36	Hyppolit	trunk/Padre/lib/Padre/Wx/Menu/ trunk/Padre/share/locale/
08:39		pece2 joined #padre
08:43	Alias	wtf
08:43	Alias	11 commits?
08:43	Alias	Dammit
08:43	Alias	Clearly I'm still missing something
08:45	Alias	Whoever added the config entry, please also fix 04-config.t
08:47	Hyppolit	svn: r9903 \| azawawi++ \| http://padre.perlide.org/trac/changeset/9903
08:47	Hyppolit	Updated Arabic translation
08:47	Hyppolit	trunk/Padre/share/locale/
08:51	Alias	aha
08:51	Alias	There it is
08:51	submersible1	que?
08:51	Alias	There's no transaction around Padre->new
08:52	Alias	You'd be AMAZED how much time Padre has been spending spinning pointlessly in SQLite
08:52	azawawi	submersible1: why the 1? :)
08:52	submersible1	Alias: you never know... I might be just barely astounded
08:52	Alias	home vs work
08:52	submersible1	nah - libpurple is just futzing me about
08:53	* submersible	submersible dooms all other zombie submersibles
08:53	azawawi	:)
08:55	Hyppolit	svn: r9904 \| azawawi++ \| http://padre.perlide.org/trac/changeset/9904
08:55	Hyppolit	04-config.t now passes...
08:55	Hyppolit	trunk/Padre/t/
08:55	azawawi	Alias: not me... but i fixed it :)
08:56	azawawi	Alias: what was Padre doing "endlessly" with SQLite?
08:57		pece joined #padre
08:59	Alias	connect, query, commit, disconnect
08:59	azawawi	Alias: where? Padre::DB?
08:59	Alias	All over the place
09:00	Alias	And since both ORLite and SQLite don't natively allow arbitrarily nested transactions, we couldn't really wrap them in transactions
09:00	Alias	Not properly
09:00	Alias	Because if anything in a transaction calls anything ELSE in a transaction, boonm
09:00	Alias	So now Padre::Locker has added a nesting wrapper over the actual transaction handling
09:01	submersible	so anyone can ->rollback the current transaction ?
09:01	Alias	->rollback is not supported
09:01	Alias	It's commit or bust
09:01	Alias	Which is why it was so easy to do the nesting
09:01	submersible	bust ? like - sry , your database is hosed?
09:02	Alias	No, like Perl die, resulting in Program crashing, resulting in rollback during END-time for DBI
09:02	Alias	Perl-level crash of course, not segfault (hopefully)
09:02	Alias	The only thing that makes nested transactions inherently hard is partial rollback
09:03	Alias	Remove that and everything is simple :)
09:04	Alias	Oh dear
09:04	Alias	Action bugs :(
09:05		marcela joined #padre
09:07	Alias	It appears that we don't set the Checked state of the View menu from config at constructor time any more
09:07	Alias	Which means that when the main->show_functions fires, it THINKS that it's being enabled from the menu
09:07	Alias	Which means it saves the new config setting into the database
09:07	Alias	At constructor time
09:11		marcela joined #padre
09:19		pece2 joined #padre
09:20		elgatito joined #padre
09:21		elgatito left #padre
09:22	Alias	ok, trying a simple (slightly ugly) fix for the View stuff
09:23		marcela joined #padre
09:26		pece joined #padre
09:30		pece2 joined #padre
09:30	Hyppolit	svn: r9905 \| azawawi++ \| http://padre.perlide.org/trac/changeset/9905
09:30	Hyppolit	Translation fixes and updated Arabic translation
09:30	Hyppolit	trunk/Padre/lib/Padre/Action/ trunk/Padre/lib/Padre/Wx/Menu/ trunk/Padre/share/locale/
09:30	Alias	Oh man, I've really screwed up the startup process
09:30	Alias	Must have done it 6 months ago at least
09:31	Alias	azawawi: I _REALLY_ want this fixed before release
09:31	Alias	Going as quickly as I can :)
09:38	azawawi	&
09:40	Alias	oh wow
09:42		kentnl joined #padre
09:56	Hyppolit	svn: r9906 \| adamk++ \| http://padre.perlide.org/trac/changeset/9906
09:56	Hyppolit	Huge improvement to startup speed, by auditing startup-time database connections.
09:56	Hyppolit	Found some that needed transactions wrapped around them, and some bugs in the GUI layout code that was accidentally causing the HOST configuration database to be written back into the SQLite db 4 times.
09:56	Hyppolit	trunk/Padre/lib/ trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/DB/ trunk/Padre/lib/Padre/Wx/
09:57	Alias	OK, I'm pretty much happy to see that revision released
09:57	Alias	But I'm going to take a quick look at Text::FindIndent and see if we're calling it needless (since it costs a fortune to run)
09:59		Sewi joined #padre
10:00	Sewi	hi all
10:03	szabgab	hi Sewi
10:03	szabgab	I actionified the File menu but have not committed yet, postponing it after the release
10:03	Sewi	thanks
10:03	Alias	Sewi: So, I did a teeensy bit of optimisation work today
10:03	Alias	svn update
10:03	Alias	Let me know what you think
10:04	szabgab	Alias, on startup I get lots of debuggin output
10:04	Alias	szabgab: oh rly?
10:04	Alias	nopaste
10:04	szabgab	package Padre::DB::SyntaxHighlight
10:04	szabgab	sub select
10:04	szabgab	etc...
10:04	Alias	paste it
10:05	szabgab	lots of package this/that and then names of subs
10:05	Alias	paste it
10:05	Alias	:)
10:05	szabgab	I seem to have lost all the pastebot links
10:05	Alias	nopaste
10:05	szabgab	we don't have a link to a pastebot in padre?
10:05	Alias	dipsy, nopaste
10:05	Alias	dipsy, nopaste?
10:06	Alias	nopaste?
10:06	szabgab	http://www.pastebot.net/paste/tNNcBGFrnVv/
10:06	szabgab	nopaste?
10:06	Alias	I was hoping for bot recognition
10:06	szabgab	Hyppolit, nopaste is http://www.pastebot.net/
10:06	Hyppolit	nopaste was unknown
10:06	Hyppolit	nopaste is now http://www.pastebot.net/
10:06	Alias	Without looking to see if we had any actual bots :)
10:06	szabgab	nopaste?
10:06	Hyppolit	nopaste is http://www.pastebot.net/
10:06	szabgab	now we have it
10:06	Alias	oh!
10:06	Alias	Typo, I left a debug variable on
10:06	Alias	one sec
10:07	szabgab	but should really have a Link in Padre/Help to a pastebot
10:07	szabgab	agree?
10:07	Hyppolit	svn: r9907 \| adamk++ \| http://padre.perlide.org/trac/changeset/9907
10:07	Hyppolit	Remove debugging dump
10:07	Hyppolit	trunk/Padre/lib/Padre/
10:07	Alias	Right, try again
10:08	Alias	Generate URL for Current Document or something
10:08	Alias	We can probably do it without any new dependencies
10:08	Alias	There's a generic HTTP task already
10:08	szabgab	yes, they are gone
10:09	szabgab	I would start by adding a menu item 'Open browser on nopaste'
10:09	szabgab	then think about it a bit more
10:09	Alias	Meaningless for newbies
10:09	Alias	"TinyURL Current Document"
10:09	Alias	or something
10:09	szabgab	I did not mean to do it with current document, at least not yet
10:09	Alias	Or to really sync with the Zeitgeist, "Twitter This Document"
10:10	Alias	:)
10:10	Sewi	What about a Textbox for pasting something which goes to nopaste and which returns a copy-able link
10:11	Alias	Help -> Share -> three or four options...?
10:11	Alias	OK, I'm done and ready for the release
10:11	Alias	All performance hacks are in
10:12	Alias	New most expensive functions
10:12	Alias	297ms547msEncode::Guess::::guess Encode::Guess::guess
10:12	Alias	297ms297msWx::Menu::::AppendString Wx::Menu::AppendString(xsub)
10:13	Alias	188ms266msPadre::Util::::get_project_dir Padre::Util::get_project_dir
10:13	Alias	That last one is interesting, because to open 15 or so documents, it got called 300 times
10:13	Alias	And it does filesystem operations
10:13	Sewi	It should be cached.
10:13	Alias	You'd have thought so
10:13	Alias	Or at least cached for a while
10:13	Alias	Like, say, a second
10:14	Alias	:)
10:14	* szabgab	actionifying the Edit menu as well
10:14	Sewi	Think what happens on remote fs, like nfs, samba, sshfs, ftp...
10:14	Alias	Sewi: It's not inconceivable that a file will change projects because of Makefile.PL pokery during the lifetime of a Padre instance
10:14	Alias	But yeah, remote stuff will get slaughtered by that
10:14	Sewi	Yes, but caching for a few seconds should be ok
10:15	waxhead	release is now...
10:15	Alias	I'll look into in a subsequent profiling session
10:15	Alias	Everything is great for right now
10:15	Alias	Padre is much much faster
10:15	Alias	Life is good
10:15	Alias	Sewi: And session support is now free if you don't use it :)
10:15	waxhead	any blockers, changes to be committed desperately??
10:15	Sewi	waxhead: I know that you announced it long enough, but would you let me another 10 minutes for translation? :-)
10:15	Alias	Maintaining the padre-last session was costing a fucking fortune
10:15	waxhead	Sewi, yep.. sure
10:16	waxhead	nothing in Changes has changed much
10:16	Alias	Sewi: So now padre-last only gets maintained if your main_startup settings are "last" or "session"
10:16	waxhead	Sewi, I think now that szabgab has done all his status bar changes you should be right
10:18	Hyppolit	svn: r9908 \| adamk++ \| http://padre.perlide.org/trac/changeset/9908
10:18	Hyppolit	trunk/Padre/
10:18	Alias	(just dumped in one last Changes file entry)
10:19	waxhead	man, the tool tip thingy needs to respect escape...
10:20	Sewi	Wow, I updated this just before the last release and now... 160 new! :-)
10:21	Alias	Yeah, the pre-christmas rush :)
10:21	szabgab	waxhead, escape? tooltip?
10:22	Hyppolit	svn: r9909 \| adamk++ \| http://padre.perlide.org/trac/changeset/9909
10:22	Hyppolit	Setting release date in Changes
10:22	Hyppolit	trunk/Padre/
10:23	waxhead	szabgab, when you have autocomplete always on, when it picks out a work and provides the list you can't hit escape to tell it to "go away not now".
10:23	waxhead	s/work/word/
10:23	Alias	I wonder how hard it would be to make a modal intersticial cpan installation dialog
10:23	Alias	Would probably cause some issues downstream, I imagine
10:23	Sewi	waxhead: Ticket it, please...
10:23		tsee joined #padre
10:24	szabgab	first we should fix the cpan installer of padre, I think it is borked
10:24	Alias	Run -> "Run with Profiling" -> Dialog(Installing Devel::NYTProf...)
10:24	Alias	szabgab: Yup :(
10:24	szabgab	and we need to implement the "prefer os packaging system" in the CPAN client
10:25	szabgab	unrelated to Padre
10:25	Hyppolit	svn: r9910 \| waxhead++ \| http://padre.perlide.org/trac/changeset/9910
10:25	Hyppolit	Update for preferences added
10:25	Hyppolit	trunk/Padre/
10:25	submersible	Aha
10:25	tsee	Good morning.
10:25	submersible	#390 middle button paste
10:25	Hyppolit	#390: Copy and Paste onto highlighted text (accepted defect) [ http://padre.perlide.org/trac/[…]et/390#comment:10 ]
10:25	submersible	hi tsee
10:26	tsee	Long time no see, dear Australians.
10:26	waxhead	hey tsee
10:26	Alias	heya
10:26	tsee	Always sleeping down under.
10:26	Alias	tsee: We need to take a look at Text::FindIndent at some point
10:26	Alias	tsee: Major major performance issues
10:26	tsee	What's wrong with it? Oh!
10:26	tsee	I'm surprised that matters.
10:27	tsee	Does it read the whole document each time?
10:27	Alias	100ms+ per file
10:27	Hyppolit	svn: r9911 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9911
10:27	Hyppolit	add changes
10:27	Hyppolit	trunk/Padre/
10:27	tsee	Ah. It had to check the last few lines for mode lines.
10:27	Alias	It checks EVERY lines
10:27	tsee	But other than that, it could stop after a few tens of lines, usually.
10:27	tsee	Yeah, I recall now.
10:27	Alias	15 files, 60,000 calls to _parse_emacs_blah blah
10:27	tsee	Oops.
10:27	tsee	It's a big ball of regexp, too.
10:27	Alias	And 60,000 Regexp::DESTROY calls
10:28	waxhead	how's everyone going with finishing up?
10:28	tsee	Ugh, it actually calls methods!?
10:28	waxhead	ready for a release yet?
10:28	tsee	I mean regexp destruction?
10:28	Alias	NYTProf apparently can see that now
10:28	tsee	That's fucking horrible.
10:28	tsee	What version of Perl?
10:28	Alias	5.10.1
10:28	tsee	Stuff changed
10:28	tsee	Yes, might have been between 5.10 and 5.10.1
10:29	Alias	Regardless, at the very least a round of micro-optimisation needed
10:29	tsee	Should be fairly easy to shave off a bit of the time.
10:29	Alias	A little reuse of qr// here, a little //o there and maybe inline the call to parse the vim/emacs crap
10:29	Alias	Should resolve half the problem
10:29	tsee	/o might not do much when dealing with qr// anyway.
10:29	Alias	point...
10:29	Alias	Anyways
10:30	tsee	Also, inlining even more (big ball of regexp) may be... counterproductive.
10:30	Alias	Indent checking is now like 80%+ of the entire cost of opening files
10:30	tsee	Holy shit.
10:30	Alias	And that includes SQLite commits :)
10:30	Alias	When you get some time, run dev.pl -p, immediately open 15 files with a single open dialog, then close the window
10:31	Alias	That should get you a pretty clear look at the problem
10:31	submersible	pmsl
10:31	tsee	I'll have a look.
10:31	tsee	But this is a normal working day for me.
10:31	tsee	So don't expect miracles.
10:31	Alias	Until now, the cost of indent checking was being hidden by the insano cost of everything else
10:31	tsee	Yep.
10:31	Alias	Which is all otherwise fixed
10:31	tsee	Congrats.
10:31	Alias	So you've got easily a month before new-and-fast becomes comparitively old-and-slow
10:31	Alias	For now, I've just turned auto-detect off
10:32	tsee	No!
10:32	tsee	That's the wrong solution.
10:32	Alias	I'm mostly working on my own code
10:32	Alias	I turn it on when needed
10:32	tsee	Oh, for yourself. Fair enough.
10:32	Alias	oh, you thought I meant for everyone? :)
10:32	tsee	I thought for the release or something.
10:32	Alias	heh
10:32	Alias	No no
10:32	tsee	Because that would violate DWIM.
10:32	Alias	0.53 is way faster than 0.52
10:32	tsee	Which we've been very good about-.
10:32	Alias	It's only slow compared to 0.54
10:32	Alias	:)
10:34	submersible	I think maybe i don't close and re-open the editor often enough...
10:36	Sewi	We should add more crashs to force people to do so :-)
10:37	submersible	eeeew
10:37	waxhead	Sewi, you done yet?
10:38		Alias joined #padre
10:39	Sewi	waxhead: Only by accident :-)
10:40	submersible	Alias: come back here!
10:41	tsee	Strangely, T::FI takes only about 8-10ms per file on my laptop.
10:41	tsee	This is the average over the whole of Padre's code base.
10:41	tsee	And this is a 2.5 years old laptop.
10:41	tsee	low-end
10:43	Hyppolit	#806: Auto complete tooltips won't "go away" (new defect) [ http://padre.perlide.org/trac/ticket/806 ]
10:43	szabgab	waxhead, I think you can already update the version number
10:43	szabgab	and let Sewi rush in his translation
10:44	waxhead	yeah, that's it.. we're going for release
10:44	waxhead	szabgab oh... how do I not screw up the --tag like I did last time?
10:44	szabgab	you don't upload the .tar.gz created without --tag
10:44	szabgab	only the one you create with --tag
10:45	tsee	Okay, Padre/Wx/Main.pm takes 160ms for indent-checking.
10:45	waxhead	szabgab, I did that though...
10:46	waxhead	ok... lets see how it goes this time huh? :)
10:48	Sewi	??? is a really cool string to translate :-)
10:48	szabgab	make it !!!
10:49	Hyppolit	svn: r9912 \| waxhead++ \| http://padre.perlide.org/trac/changeset/9912
10:49	Hyppolit	Version update
10:49	Hyppolit	trunk/Padre/
10:50	Hyppolit	svn: r9913 \| waxhead++ \| http://padre.perlide.org/trac/changeset/9913
10:50	Hyppolit	Perl tidy for release 0.53
10:50	Hyppolit	trunk/Padre/lib/ trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/DB/ trunk/Padre/lib/Padre/Document/ trunk/Padre/lib/Padre/Wx/ trunk/Padre/lib/Padre/Wx/Dialog/Preferences/ trunk/Padre/lib/Padre/Wx/Menu/
10:50	tsee	Okay, shaved off 20% of T::FI run time already.
10:53	Hyppolit	svn: r9914 \| waxhead++ \| http://padre.perlide.org/trac/changeset/9914
10:53	Hyppolit	Version Update to 0.53 for release
10:53	Hyppolit	trunk/Padre/lib/ trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Action/ trunk/Padre/lib/Padre/Config/ trunk/Padre/lib/Padre/DB/ trunk/Padre/lib/Padre/DocBrowser/ trunk/Padre/lib/Padre/Document/ trunk/Padre/lib/Padre/Document/Perl/ trunk/Padre/lib/Padre/File/ trunk/Padre/lib/Padre/HelpProvider/ trunk/Padre/lib/Padre/Plugin/ trunk/Padre/lib/Padre/Plugin/PopularityContest/ trunk/Padre/lib/Padre/Project/ trunk/Padre/lib/Pad
10:58	Hyppolit	svn: r9915 \| jquelin++ \| http://padre.perlide.org/trac/changeset/9915
10:58	Hyppolit	updated french translation part 1
10:58	Hyppolit	trunk/Padre/share/locale/
10:59		Alias joined #padre
11:01		Alias joined #padre
11:01	Alias	Has the release happened yet?
11:01	waxhead	Alias, getting there
11:01	Alias	ok
11:01	waxhead	trying to build a test
11:01	* Alias	is going to start chasing the get_project_dir hot-spot
11:01	Alias	But I didn't want to commit until after 0.53
11:02	waxhead	Alias, it's not tagged yet
11:02	Sewi	done
11:02	Hyppolit	svn: r9916 \| Sewi++ \| http://padre.perlide.org/trac/changeset/9916
11:02	Hyppolit	Updated German translation to 100% again - thanks waxhead for waiting!
11:02	Hyppolit	trunk/Padre/share/locale/
11:02	Sewi	waxhead: thanks!
11:04	waxhead	WTF!!
11:05	tsee	Alias: 20% speedup on T::FI already.
11:05	waxhead	Sewi, no prob
11:06	tsee	Alias: Also, for an average .pm file in Padre, the T::FI run takes about 9ms for me as opposed to 100ms.
11:06	Alias	Curious
11:06	Alias	So the get_project_dir load is coming from the Window menu :)
11:06	Alias	Every time it does a refresh, it calls get_project_dir for every currently open tab
11:06	Alias	And since it's not part of the locking API, and it's being called directly by things like ::Main->close, guess how often it's being called
11:06	Alias	For open(15 files) and close(15 files) it's being called O(n)
11:06	Alias	So it's an O(n) with a ton of filesystem stats in it :)
11:06	Alias	I'll start to fix that :)
11:06	Alias	At least down to O(n) anyways
11:07	Alias	tsee: Curious
11:07	Alias	I was mostly taking swathes from Padre::
11:07	tsee	Exactly.
11:07	Alias	So maybe a fair whack of that included things like Main.pm
11:07	tsee	I took the whole of it and divide by the number of files.
11:07	tsee	Gave me an average of 9ms.
11:07	Alias	Plus it was in the profiler
11:07	tsee	That changes the picture, of course.
11:07	Alias	So the per-line sub calls to the vim detector probably cost more than normal
11:08	tsee	Problem is that I'd have to find the hidden vim documentation that says whre the vim mode lines are allowed.
11:08	tsee	I'm not sure it's only at start and end of the document.
11:08	Alias	Or just express it as a single regex
11:08	tsee	I moved it out of the sub.
11:08	Alias	That should help
11:09	Alias	And removed the qr stuff I imagine
11:09	Alias	or reused
11:09	tsee	Reused.
11:09	Alias	I would bet that costs more in the profiler too
11:09	tsee	And now I'm checking on the vim prefix only. If that matches, I check the full modeline regexps.
11:09	Alias	Excellent
11:10	tsee	vim/emacs mode lines or whoever invented their format should be shot
11:10	Alias	At least this is simple
11:10	Alias	After this is fixed, the next heaviest calls are to Encode
11:10	Alias	Which is more troublesome
11:11	Alias	It's one of the classic failures in software development
11:11	Alias	Anyone with an even rudimentary understanding of game theory or economics would have spotted it instantly
11:12	Hyppolit	svn: r9917 \| jquelin++ \| http://padre.perlide.org/trac/changeset/9917
11:12	Hyppolit	updated french translation part 2
11:12	Hyppolit	trunk/Padre/share/locale/
11:12	waxhead	anyone want to run release.pl and see if it builds on your system?
11:13	waxhead	it's the second part and it fails saying that DISPLAY is not set...
11:13	tsee	What's the failure? Encode?
11:13	waxhead	t/04-config-t
11:14	waxhead	DISPLAY isn't set!!!
11:14	waxhead	man..
11:14	szabgab	you can release it with --display if that's what stopping you
11:15	szabgab	I think we carry this problem for some time already
11:15	waxhead	going to give it a go
11:15	waxhead	so how is this possible... last two releases were smooth.. :)
11:16		Alias joined #padre
11:16	szabgab	you can binary search which revision broke it
11:17	szabgab	you can run the release.pl with --rev to build with certain revision and see if that is still working
11:18		Alias joined #padre
11:21	waxhead	woo hoo.. build and installs
11:21	* Alias	commits and puts a stop to that :)
11:21	tsee	Alias: This is about it. I don't have a lot more time and no easy micro-optimization ideas.
11:21	Alias	(j/k)
11:21	Alias	tsee: The basics are fine, dump it to CPAN and I'll take a look later
11:21	Hyppolit	svn: r9918 \| jquelin++ \| http://padre.perlide.org/trac/changeset/9918
11:21	Hyppolit	updated french translation part 3 (french back to 100%)
11:21	Hyppolit	trunk/Padre/share/locale/
11:21	tsee	Wrt. micro-optimization, maybe you should have a look. You're likely to have a better idea about what to micro-optimize.
11:22	waxhead	jq, pity.. just built with part 2 translation... sorry
11:23	waxhead	szabgab can I upload this somewhere, or do you want to get it from my server?
11:23	tsee	Text::FindIndent 0.06 uploaded to CPAN.
11:23	szabgab	I'll take it from your server
11:23	tsee	20-25% faster.
11:23	tsee	Should we depend on it in Padre already?
11:23	tsee	It's a safe and simple performance win.
11:24	szabgab	waxhead, msg me the ur;
11:24	waxhead	tsee,
11:24	szabgab	url
11:24	waxhead	tsee, I don't see any reason not too... Client::Debug kept increasing nearly every day...
11:25	waxhead	szabgab, in a bit.. nearly complete
11:25	tsee	Just asking because it's just before the release.
11:25	waxhead	tsee, too late for 0.53
11:25	Hyppolit	svn: r9919 \| waxhead++ \| http://padre.perlide.org/trac/changeset/9919
11:25	Hyppolit	tag Padre-0.53
11:25	Hyppolit	tags/
11:25	tsee	Ah, okay.
11:25	waxhead	see ? :)
11:25	tsee	Easy decision, then.
11:26	Hyppolit	svn: r9920 \| tsee++ \| http://padre.perlide.org/trac/changeset/9920
11:26	Hyppolit	require faster FindIndent
11:26	Hyppolit	trunk/Padre/
11:27	Hyppolit	svn: r9921 \| jquelin++ \| http://padre.perlide.org/trac/changeset/9921
11:27	Hyppolit	would you please stop changing the string catalog when i'm translating it?! :-)
11:27	Hyppolit	trunk/Padre/share/locale/
11:29	jq	how are we supposed to name the mo files in plugins by now?
11:31		azawawi joined #padre
11:32	jq	fr-fr.po, Padre__Plugin__Foobar-fr-fr.po or Foobar-fr-fr.po ?
11:32	azawawi	Padre__Plugin__Foobar-fr-fr.po
11:36	jq	ok
11:36	jq	thanks
11:42	szabgab	waxhead, http://perlide.org/download/so[…]Padre-0.53.tar.gz
11:43	waxhead	szabgab installed OK for you?
11:43	szabgab	now installing on ubuntu and on windows
11:44	szabgab	ubuntu 9.10 installed
11:45	waxhead	Sewi, what's your blog URL???
11:45	waxhead	szabgab, ok when you're good I'll upload it to cpan
11:45	szabgab	go ahead
11:45	szabgab	my windows machine is a bit slow
11:46	szabgab	lunch &
11:48	waxhead	ok..
11:49	Sewi	waxhead: http://padre-ide.de/blog/
11:50	waxhead	Sewi, thanks.. googled you!
12:02	szabgab	windows install was ok
12:02	szabgab	I have not tried to actually run padre :-)
12:02	szabgab	tsee, where is the new prereq ? the smokers complain
12:04	kaare	Seems to work fine on OpenSUSE :-)
12:04	kaare	0.53, that is.
12:12	waxhead	kaare, thanks.. I'll have it uploaded to cpan shortly...
12:12	tsee	szabgab, on its way to PAUSE
12:13	tsee	Will be there within a few minutes to hours to days
12:13	tsee	$there being your CPAN mirror
12:13	tsee	Depending on it's update frequency.
12:13	yakudzo	sabayon works too
12:23	waxhead	ok, file has been submitted to cpan
12:26	kaare	F2 is b0rked. Repeats a dialog (indefinitely?) with "Could not find a help provider for Text"
12:27	Sewi	kaare: Ticket this, please
12:28	kaare	Trac ticket system is b0rked. You can't enter tickets for 0.53 ;-)
12:28	Sewi	It's also on trunk :-)
12:29	waxhead	szabgab, can you update trac with the new version??
12:29	Sewi	kaare: Sure, you could choose 0.52 :-)
12:29	Hyppolit	#807: F2 is broken (new defect) [ http://padre.perlide.org/trac/ticket/807 ]
12:29	waxhead	kaare, use trunk too.. it's likely broken there as well.
12:29	Sewi	waxhead: Just added it
12:30	waxhead	Sewi, thanks
12:30	waxhead	cpan has indexed Padre 0.53
12:32	kaare	hm, how does the debugger work?
12:32	kaare	Say I want to debug Padre, what to do?
12:33	kaare	(definitely NOT press F2 ;-))
12:34	waxhead	kaare, ask Gabor.. I think the more we ask him the sooner he'll write up a blog/howto about it! :)
12:35	Sewi	Open the Perl script and klick on one of the debugger icons (toolbar right)
12:35	waxhead	Sewi, but it's a perl script, not padre ?
12:35	Sewi	You might not get what you want if you use the STOP button :-)
12:35	szabgab	:-)
12:35	Sewi	Ask szabgab if you could debug a module.
12:35	szabgab	padre is just a perl script
12:35	szabgab	you can debug modules
12:36	waxhead	szabgab, that it is.. but a little more complex that a simple perl script
12:36	Sewi	szabgab: Does debugging work with multiple open files?
12:36	szabgab	but the breakpoints are not kept correctly between runs
12:36	Sewi	waxhead: Use dev.pl to run the debugger :-)
12:44	azawawi	waxhead++
12:44	szabgab	waxhead, thanks for the release !
12:44	waxhead	my pleasure...
12:44	waxhead	:)
12:45	waxhead	anyone see Steffen's email in the dev list?
12:46		yakudzo left #padre
12:46	tsee	I fixed the indentation issue.
12:46	waxhead	tsee, tops!! makes life simpler for all I expect...
12:46	waxhead	I like you email in the mailing list...
12:47	waxhead	getting things right in the editor is key... sometimes it's a hard slog though..
12:47	tsee	I know.
12:47	tsee	I was trying to show my appreciation of everyone's work.
12:48	waxhead	works for me.. glad to see some Padre envy... ;)
12:48	tsee	My feelings exactly.
12:48	tsee	Speaking of appreciation. I think my girl-friend will appreciate if I go shopping like I promised. :)
12:48	Hyppolit	svn: r9922 \| jquelin++ \| http://padre.perlide.org/trac/changeset/9922
12:48	Hyppolit	french translation for padre::plugin::plack
12:48	Hyppolit	trunk/Padre-Plugin-Plack/ trunk/Padre-Plugin-Plack/share/locale/
12:48	waxhead	tsee, absolutely!
12:49	waxhead	tsee, I'm thinking we need a release that acknowledges our better halves... I get to do this because my partner is rather forgiving of my time sink...
12:50	waxhead	she even said she was proud of what I do once she discovered this 'release manager' gig...
12:55	waxhead	Sewi, did you get the new ticket for the CRTL-Tab issue?
13:02	Sewi	waxhead: #804? Yes.
13:02	Hyppolit	#804: CRTL-Tab Last seen order not quite right (new defect) [ http://padre.perlide.org/trac/[…]ket/804#comment:1 ]
13:07	waxhead	htat's the one...
13:07	waxhead	really odd behaviour
13:11	Hyppolit	svn: r9923 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9923
13:11	Hyppolit	change the File menu to be built by Actions
13:11	Hyppolit	trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Action/ trunk/Padre/lib/Padre/Wx/Menu/
13:16	Hyppolit	svn: r9924 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9924
13:16	Hyppolit	change the Edit menu to be built by Actions
13:16	Hyppolit	trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Action/ trunk/Padre/lib/Padre/Wx/Menu/
13:17	Sewi	szabgab: thanks
13:17	szabgab	you are welcome
13:18	szabgab	I think one left, I'll try to do later
13:18	szabgab	and then we need to think over how to eliminate the code part from the Padre::Wx::Menu::* modules and keep only the hierarchy of the menues
13:22	Sewi	Why not keep the original idea and move it to config?
13:23	Hyppolit	svn: r9925 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9925
13:23	Hyppolit	add the two new Action modules to the badcode exceptions as they are using Wx related code
13:23	Hyppolit	trunk/Padre/xt/
13:23	Sewi	Is Wx::gettext Wx-related?
13:24	szabgab	actually I think wx::gettext should be replaced by _T everywhere
13:24	szabgab	but that's not the Wx related stuff
13:24	szabgab	the Wx related is that we have a use ... Wx in the module
13:25	szabgab	If I understand the idea is to move the menu hierarchy to a configuration file?
13:25	szabgab	if so then it cannot contain any code that needs to be executed, right?
13:25	waxhead	doesn't kephra do that?
13:25	szabgab	so what I am saying is just another step in that direction
13:25	* waxhead	can't remember what it was he was going to do
13:25	szabgab	yes Kephra does that too
13:26	waxhead	has anyone ever found out why we always get 1 scalar leaked when padre closes?
13:26	Sewi	All code should be in the actions - the menu structure should be a reference to actions and special menu items (like "list of open files")
13:26	jq	waxhead: "she even said she was proud of what I do once she discovered this 'release manager' gig..." ==> i understand you volunteer for the release manager hat for the whole year 2010 :-)
13:26	szabgab	waxhead, I get 2 scalars leaked
13:26	jq	congratulations! \o/
13:27	waxhead	that's not quite what it means... :)
13:27	waxhead	I'll keep at it until someone else wants to take it on...
13:27	jq	oh? that's what i understood ;)
13:27	waxhead	share the glory and the love.. :)
13:29	waxhead	I'll tell you what.. when I get a release where all languages are 100% for Padre, I'll retire.. ;)
13:30	* jq	ponders stopping updating french translation :-)
13:33	Hyppolit	svn: r9926 \| waxhead++ \| http://padre.perlide.org/trac/changeset/9926
13:33	Hyppolit	updated events for release 0.53
13:33	Hyppolit	trunk/template.padre.perlide.org/data/
13:34	waxhead	jq, how long would you generally need to get the translation done for Padre only?
13:34	waxhead	say a day or so?
13:34	waxhead	oh.. and how do you know what needs to be translated?
13:34	waxhead	is it everything with Wx::gettext() ?
13:35	* waxhead	is totally ignorant of what it takes for the translations
13:36	tsee	waxhead, szabgab, the scalars leaked issue is a combination of Wx, threads, and the crazy ways in which we use both.
13:36	jq	a day is largely enough
13:36	tsee	If you drill all the way down, it's a general issue in core perl:
13:36	tsee	The stacks aren't reference counted.
13:37	jq	and i check padre translation status from time to time to know if i need to translate stuff
13:37	jq	also, announcing the release in advance on the ml really helps
13:37	tsee	(For those without internals fu, this essentially means that the normal garbage collection doesn't apply to @_, but it works differently.)
13:37	waxhead	jq, no worries.. I'll keep posting release times then..
13:37	waxhead	jq, how do you know what in padre needs translation?
13:38	tsee	The bug isn't that the items on the stack aren't part of the reference counting, but that it requires all sorts of contortions to deal with that fact. It's the way it is for performance reasons, as far as I can tell.
13:38	tsee	And what we're seeing is just one manifestation of the issue.
13:38	jq	waxhead: http://perlide.org/translations/
13:38	waxhead	tsee, wow... thanks...
13:38	tsee	I tried to tebug it a while ago, but it's way beyond me.
13:39	tsee	*debug
13:39	* azawawi	&
13:39	tsee	Apart from being annoying and unsettling for users, the warnings a virtually harmless.
13:39	waxhead	jq, not what's missing, but what in padre is not translated.. do you run a script or something?
13:40	tsee	Some day, someone should sweet-talk Mattia to try to get to the bottom of it, but even for him, it's probably a big effort with unclear outcome.
13:40	tsee	Specifically because it's likely not his (or Wx's) fault.
13:40	waxhead	jq, I see now.. a little better.. just reading the translationintro page
13:41	waxhead	tsee, so it's actually perl throwing that warning out?
13:41	tsee	waxhead, yes.
13:41	waxhead	need another pragma
13:41	tsee	If you compile it with all the right flags, it even gives you some hints as to what scalar was leaked.
13:41	jq	waxhead: when i see that there are some non-translated strings, i then fire the poedit program that makes it easy to update a po file with a catalog and listing untranslated/fuzzy strings
13:41	waxhead	use scalar::leak qw/shutup/;
13:42	tsee	waxhead, it's a valid warning.
13:42	waxhead	jq, cool... just trying to get a feel for the work involved...
13:42	waxhead	tsee, I guess so.. but never looks good when Padre closes down...
13:43	tsee	It's just that considering Padre's footprint and likely occasional memory leaks, 1-2 scalars leaked at thread termination (!= ::Task terminations, threads are reused) is negligible.
13:43	waxhead	can we catch it and simply stop it from being send to stdout ?
13:44	tsee	Depends. Probably difficult (global destruction and all). Let me check something.
13:44	tsee	It's likely STDERR, by the way.
13:44	waxhead	tsee, of course.. it's an error being reported.. makes sense..
13:44	waxhead	it's late here.. ;)
13:45	tsee	There's a bunch of open bug reports on related issues.
13:45	tsee	bug reports against the core.
13:46	tsee	There's also TODO tests.
13:46	tsee	But those issues aren't necessarily exactly ours.
13:46	tsee	I don't even know WHAT exactly is going wrong despite all the investigation.
13:47	tsee	The warning is coming from perl.c line 1233.Perl_warner(aTHX_ packWARN(WARN_INTERNAL),"Scalars leaked: %ld\n", (long)PL_sv_count);
13:47	tsee	It's in perl_destruct.
13:48	tsee	Which is called when a perl interpreter (== a thread or the main thread) is destroyed.
13:48	tsee	But the leak could have been anywhere.
13:48	tsee	If anyone wants to dig into this, check out "perldoc Padre::Manual::Hacking"
13:49	tsee	Under "Heavy-duty debugging"
14:05	waxhead	wow.. tht's what it takes to debug at that level...
14:06	tsee	Consider what it took my clueless self to get to the point of that document.
14:06	waxhead	tsee, you're hardly clueless...
14:09	waxhead	far out...
14:09	waxhead	what's the potential to break your normal perl isntall with doing this?
14:09	tsee	Low.
14:09	tsee	Do it as a user, not as root.
14:10	waxhead	i might have a look at it myself over the break
14:10	tsee	When it comes to C and debugging threads issues... I learned it all for THIS.
14:10	waxhead	didn't you do the task threading stuff for Padre?
14:10	tsee	Yes.
14:10	waxhead	hardly clueless then!
14:10	tsee	Well, I learned Perl threads for that.
14:11	tsee	Then learned debugging them (to some extend) to fix it.
14:11	tsee	I got a lot of help from the clueful, too.
14:11	waxhead	the perl debugger still scares me.. hence i'm waiting on Debugging in Padre.. :)
14:11	waxhead	that always makes a difference, having help around
14:12	waxhead	ok time for bed.. it's very late here
14:12	waxhead	nite
14:12	tsee	Good night.
14:13		patspam joined #padre
14:20	Hyppolit	svn: r9927 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9927
14:20	Hyppolit	add script by Tux listing the installed module based on perllocal
14:20	Hyppolit	trunk/Perl-Dist-XL/script/
14:21	Hyppolit	svn: r9928 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9928
14:21	Hyppolit	svn:ignore
14:21	Hyppolit	trunk/Padre-Plugin-Git/
14:34		pece joined #padre
14:44	tsee	Alias: For when you return (in case you read logs). I just applied a change to T::FindIndent that makes the total speedup 33%.
14:44	tsee	For small files, it'll be a little slower. For large files that contain emacs mode lines at the end of the file, it should be hugely fasttr.
14:45		szabgab joined #padre
14:47		Topic for #padreis now Latest Padre is 0.53, for installation see http://padre.perlide.org/trac/wiki/Download \| Log: http://perlide.org/irclogs/ and http://irclog.perlgeek.de/padre/today \| Pasting: http://scsys.co.uk:8001
14:54	tsee	Okay. Done for today.
14:54	tsee	Now 40% faster: 12.9s benchmark down to 7.6s.
14:55	tsee	If vim didn't allow fucking modelines everywhere...
15:03	szabgab	tsee++
15:12	Hyppolit	svn: r9929 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9929
15:12	Hyppolit	[XL] update perl development version to 5.11.3
15:12	Hyppolit	trunk/Perl-Dist-XL/lib/Perl/Dist/ trunk/Perl-Dist-XL/script/
15:38	Hyppolit	svn: r9930 \| patspam++ \| http://padre.perlide.org/trac/changeset/9930
15:38	Hyppolit	Padre::Plugin::Plack version 0.02 release
15:38	Hyppolit	trunk/Padre-Plugin-Plack/ trunk/Padre-Plugin-Plack/lib/Padre/Document/ trunk/Padre-Plugin-Plack/lib/Padre/Plugin/ trunk/Padre-Plugin-Plack/lib/Padre/Plugin/Plack/ trunk/Padre-Plugin-Plack/t/
16:14	Hyppolit	svn: r9931 \| patspam++ \| http://padre.perlide.org/trac/changeset/9931
16:14	Hyppolit	Padre::Plugin::Plack Translation support
16:14	Hyppolit
16:14	Hyppolit	Added translation support
16:14	Hyppolit	Added README generation
16:14	Hyppolit	Reviewed i18n messages
16:14	Hyppolit	trunk/Padre-Plugin-Plack/ trunk/Padre-Plugin-Plack/lib/Padre/Plugin/ trunk/Padre-Plugin-Plack/lib/Padre/Plugin/Plack/ trunk/Padre-Plugin-Plack/share/locale/
16:14	patspam	jq: I'm really sorry, I only just learnt about message.pot files
16:15	patspam	the one that was previously in Padre::Plugin::Plack was a remnant from the Catalyst plugin that I stole from
16:15	patspam	I've now updated it with the correct i18n messages
16:21	Hyppolit	svn: r9932 \| jquelin++ \| http://padre.perlide.org/trac/changeset/9932
16:21	Hyppolit	updated french translation
16:21	Hyppolit	trunk/Padre-Plugin-Plack/share/locale/
16:22	Hyppolit	svn: r9933 \| patspam++ \| http://padre.perlide.org/trac/changeset/9933
16:22	Hyppolit	Padre::Plugin::Plack removed github references
16:22	Hyppolit	trunk/Padre-Plugin-Plack/ trunk/Padre-Plugin-Plack/lib/Padre/Plugin/
17:48		CromeDome joined #padre
18:09	PerlJam	Is there a way to not show certain files in the directory browser?
18:09	PerlJam	Or is something like that even on the horizon?
18:10	PerlJam	Basically I have a bunch of files that are generated and I don't want to see them in the list because I'm not ever going to edit them.
18:22		patspam1 joined #padre
18:22	tsee	PerlJam, that seems like something that should be configurable.
18:22	tsee	Ideally, one would be able to configure this globally (.svn, get lost!) and project-based.
18:22	tsee	I don't think it's currently possible.
18:23	tsee	But it shouldn't be that hard to do.
18:23	tsee	Not saying I will do it :)
19:39		robn joined #padre
19:56	szabgab	anyone understands this:
19:56	szabgab	Use of inherited AUTOLOAD for non-method Module::Install::Msgfmt::install_share() is deprecated at inc/Module/Install/Msgfmt.pm line 21.
19:57	szabgab	when running Makefile.PL of Padre on a perl 5.11.3
19:59	Hyppolit	svn: r9934 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9934
19:59	Hyppolit	try to build padre on perl 5.11.3
19:59	Hyppolit	trunk/Perl-Dist-XL/lib/Perl/Dist/ trunk/Perl-Dist-XL/script/
20:00		kaare joined #padre
20:02	Hyppolit	#808: warning in Makefile.PL on perl 5.11.3 (new defect) [ http://padre.perlide.org/trac/ticket/808 ]
20:04	Hyppolit	svn: r9935 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9935
20:04	Hyppolit	match the expected warning on perl 5.11 and later
20:04	Hyppolit	trunk/Padre/t/
20:33		dapatrick joined #padre
20:38		kent\n joined #padre
20:44		patspam joined #padre
20:49	Hyppolit	svn: r9936 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9936
20:49	Hyppolit	fix the link to the latest blog
20:49	Hyppolit	trunk/template.padre.perlide.org/data/
20:52	Hyppolit	svn: r9937 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9937
20:52	Hyppolit	fix blog link, add link to Plack blog
20:52	Hyppolit	trunk/template.padre.perlide.org/data/
21:04		pece joined #padre
21:14		jjore joined #padre
21:27	Hyppolit	svn: r9938 \| szabgab++ \| http://padre.perlide.org/trac/changeset/9938
21:27	Hyppolit	link to blog about stand alone Padre for linux release
21:27	Hyppolit	trunk/template.padre.perlide.org/data/
21:34		pece joined #padre
21:46		pece joined #padre
21:52		waxhead joined #padre
21:53		pece joined #padre
21:58		pece2 joined #padre
22:01		pece joined #padre
22:04		pece2 joined #padre
22:15		pece joined #padre
22:17		pece2 joined #padre
22:31		pece joined #padre
22:35		pece2 joined #padre
23:01		[1]awnstudio joined #padre
23:09		[1]awnstudio joined #padre
23:46	kthakore	hiya!
23:46	kthakore	anyone around?
23:48	kthakore	Alias that was just whatever project I had open at the time (4 files I think)

← Previous day | Channels | #padre index | Today | Next day → | Search | Google Search | Plain-Text | plain, newest first | summary