Camelia, the Perl 6 bug

IRC log for #padre, 2010-07-01

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

All times shown according to UTC.

Time Nick Message
00:01 kthakore jq: hi!
00:16 teejay joined #padre
00:17 kyanardag joined #padre
00:35 waxhead joined #padre
00:35 waxhead morning
00:39 kthakore waxhead: evning
00:39 awnstudio joined #padre
00:42 waxhead kthakore, how goes it?  Busy in the SDL world I see...
00:42 waxhead I was reading your blog on PDL and SDL...
00:42 waxhead I'll get around to trying that out I hope...
00:45 Xaero joined #padre
00:45 kthakore waxhead: cool
00:51 jazzanova joined #padre
00:52 Hyppolit svn: r11726 | waxhead++ | http://padre.perlide.org/trac/changeset/11726
00:52 Hyppolit Add missing Copyright
00:52 Hyppolit trunk/Padre/lib/Padre/
01:05 Hyppolit svn: r11727 | waxhead++ | http://padre.perlide.org/trac/changeset/11727
01:05 Hyppolit Perl Tidy for Release 0.65
01:05 Hyppolit trunk/Padre/ trunk/Padre/lib/ trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Document/ trunk/Padre/lib/Padre/Document/Perl/ trunk/Padre/lib/Padre/Plugin/Devel/ trunk/Padre/lib/Padre/Plugin/PopularityContest/ trunk/Padre/lib/Padre/Role/ trunk/Padre/lib/Padre/Task/ trunk/Padre/lib/Padre/Wx/ trunk/Padre/lib/Padre/Wx/Dialog/ trunk/Padre/lib/Padre/Wx/Directory/ trunk/Padre/lib/Padre/Wx/Menu/ trunk/Padre/lib/Padre/Wx/Role/ tr
01:09 Hyppolit svn: r11728 | waxhead++ | http://padre.perlide.org/trac/changeset/11728
01:09 Hyppolit Set version number for release 0.65
01:09 Hyppolit trunk/Padre/lib/ trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Action/ trunk/Padre/lib/Padre/Browser/ trunk/Padre/lib/Padre/Config/ trunk/Padre/lib/Padre/DB/ trunk/Padre/lib/Padre/DB/Migrate/ trunk/Padre/lib/Padre/Document/ trunk/Padre/lib/Padre/Document/Perl/ trunk/Padre/lib/Padre/Document/Perl/QuickFix/ trunk/Padre/lib/Padre/File/ trunk/Padre/lib/Padre/Plugin/ trunk/Padre/lib/Padre/Plugin/Devel/ trunk/Padre/lib/Padre
01:18 Xaero_ joined #padre
01:22 Xaero joined #padre
01:54 Hyppolit svn: r11729 | waxhead++ | http://padre.perlide.org/trac/changeset/11729
01:54 Hyppolit tag Padre-0.65
01:54 Hyppolit tags/
02:00 kthakore waxhead: already?
02:00 kthakore waxhead: want me to test?
02:00 kthakore svn?
02:00 Hyppolit svn is http://padre.perlide.org/tr​ac/wiki/DownloadRepository
02:05 waxhead ok.. something odd with the Vista install, but a padre --reset fixed it
02:05 waxhead http://peshel.dnsalias.net/pa​dre-release/Padre-0.65.tar.gz
02:05 waxhead test away...
02:06 waxhead I'll push it to cpan a bit later this afternoon.. I have to head out now
02:06 waxhead kthakore, yep.. sure ^^
02:06 waxhead installed on Ubuntu and Vista fine.. apart from some screen corruption on vista when it started.. as I said padre --reset fixed that.
02:06 waxhead back later!
02:07 kthakore waxhead: it worked
02:07 kthakore waxhead: oh package
02:07 kthakore let me test that
02:10 waxhead kthakore, thanks
02:54 kentnl joined #padre
03:38 kthakore waxhead: context help is blank for me
03:54 |Sno| joined #padre
04:13 kaare joined #padre
04:30 Xaero_ joined #padre
04:59 waxhead garu, OK, 0.65 has been uploaded to Pause.
05:03 garu waxhead++
05:03 garu waxhead: awesome - I'll make sure the plugin works and push it to pause as well (with padre 0.65 as a dependency)
05:33 Hyppolit svn: r11730 | waxhead++ | http://padre.perlide.org/trac/changeset/11730
05:33 Hyppolit Post release 0.65 updates.
05:33 Hyppolit
05:33 Hyppolit Whoops fixup... 0.65 won't have the right details in this file.
05:33 Hyppolit trunk/Padre/
05:34 waxhead garu++ can't wait.. :)
05:39 Hyppolit svn: r11731 | adamk++ | http://padre.perlide.org/trac/changeset/11731
05:39 Hyppolit Golden Ratio: You're (I'm) Doing It Wrong
05:39 Hyppolit trunk/Padre/lib/Padre/Wx/
05:40 Hyppolit svn: r11732 | adamk++ | http://padre.perlide.org/trac/changeset/11732
05:40 Hyppolit Updating Module::Install version
05:40 Hyppolit trunk/Padre/
05:43 garu waxhead: weird.. I think the cpan version is the same as the one in trunk
05:44 waxhead Padre or your plugin?
05:44 waxhead trunk will be showing 0.65 now
05:44 garu my plugin
05:44 waxhead oh..
05:44 garu it shouldn't work
05:45 garu well, maybe it doesn't
05:45 garu but it loads :)
05:45 garu I'll give it a thorough run in the morning
05:45 garu (3am here)
05:45 garu g'night!
05:46 waxhead night..
05:46 waxhead I'll give it a go here...
05:46 bayashi left #padre
05:47 pece joined #padre
06:02 Hyppolit svn: r11733 | adamk++ | http://padre.perlide.org/trac/changeset/11733
06:02 Hyppolit Noting default sizing improvements
06:02 Hyppolit trunk/Padre/
06:04 Hyppolit svn: r11734 | adamk++ | http://padre.perlide.org/trac/changeset/11734
06:04 Hyppolit Now we have smart display-integrated defaults, we don't need to hand-code specific default sizing.
06:04 Hyppolit trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Wx/
06:05 pece joined #padre
06:05 danlucraft joined #padre
06:07 Hyppolit svn: r11735 | adamk++ | http://padre.perlide.org/trac/changeset/11735
06:07 Hyppolit Nudges
06:07 Hyppolit trunk/Padre-Plugin-FormBuilder/lib/Padre/Plugin/
06:08 Topic for #padreis now Latest Padre is 0.65 - unstable Task API give it a go release, 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
06:10 Hyppolit svn: r11736 | adamk++ | http://padre.perlide.org/trac/changeset/11736
06:10 Hyppolit Strip the "wxWidgets" string off the wxWidgets string version
06:10 Hyppolit trunk/Padre/lib/Padre/Wx/
06:25 tgape joined #padre
06:27 Hyppolit svn: r11737 | adamk++ | http://padre.perlide.org/trac/changeset/11737
06:27 Hyppolit Subtle tweaks to the versions and text in the About dialog
06:27 Hyppolit trunk/Padre/lib/Padre/Wx/
06:28 Alias_ joined #padre
06:37 Alias_ The current release appears to be unusable
06:37 Alias_ Might be the fault of whoever used highly unix-specific code in the recent files menu refresh code
06:37 Alias_ Which is throwing exceptions all over the place and breaking the locking subsystem
06:46 Sewi gm
06:47 Sewi Alias: The recent files list source is a problem because it could lock down whole Padre if a file is no longer accessible (= on a disconnected network device)
06:50 Hyppolit svn: r11738 | adamk++ | http://padre.perlide.org/trac/changeset/11738
06:50 Hyppolit Fixed the non-blocking IO problems
06:50 Hyppolit trunk/Padre/ trunk/Padre/lib/Padre/Wx/Menu/
06:51 Alias Sewi: Conversely, now you can't open files on Windows because the POSIX implementation there doesn't have the concept of non-blocking
06:51 Alias waxhead: You can't release 0.65 as is
06:52 Sewi Sorry for this, so it simply failed the open because O_NONBLOCK wasn't known?
06:52 Alias No
06:52 Alias It dies because O_NONBLOCK wasn't known
06:52 Alias Which blows up the entire locking subsystem
06:53 Alias Which leaves the GUI in a permanently locked state
06:53 Alias Which means the gui never renders after that point
06:53 Alias Partly my fault for making the locking sub-system vulnerable to trivial exceptions in the lock-release sequence
06:53 Alias It also locks the database, which breaks the local configuration
06:54 Alias I've fixed the immediate problem
06:54 Sewi Alias: I was thinking about moving the check into a one-time background-task
06:54 Sewi The menu would always show everything in the history list and the background task could clean up the database.
06:55 Alias You can't talk to the database in the background
06:55 Alias Or at least, it's never been done and there's some concerns about threading
06:56 Alias Plus, it might be possible to dead-lock stuff if we're not careful
06:57 Alias If the non-blocking fix you added makes it fast enough now, it's going to be tricky to find that last bit of performance
06:58 Sewi It's no question of performance, the O_NONBLOCK saves Padre from waiting on a file lock but it still can't fight the network-path-is-no-longer-connected thing which still freezes Padre.
06:59 Alias What's causing it?
06:59 Alias Where's the block coming from?
06:59 Sewi The disk-IO is blocking, it's no Padre or Perl fault
06:59 Sewi But if it would block the bg task, Padre should not freeze
07:00 Alias The disk io from what operation?
07:00 Alias Which line of code?
07:02 Sewi Every try to access the file, doesn't matter if it's a -e, -f, open or sysopen
07:05 Hyppolit svn: r11739 | adamk++ | http://padre.perlide.org/trac/changeset/11739
07:05 Hyppolit When calling arbitrary delayed methods during a lock-release we become vulnerable to exceptions, trap and ignore them so that the failure doesn't cascade fatally into the rest of the locking system.
07:05 Hyppolit trunk/Padre/lib/Padre/
07:05 Alias So the non-blocking code blocks?
07:06 Alias That would seem to be counter to the principle of being non-blocking :)
07:08 Sewi It doesn't block if the file is locked by file-locking (flock & others), but if the OS is waiting for the IO call which should complete fast but doesn't do, it waits.
07:09 Alias Can things like POE and such tolerate the same kind of errors?
07:09 Alias How do they deal with non-blocking open?
07:10 Sewi It's on OS level. so everything doing a "fopen" or "stat" system call on that file will block. Maybe AIO would work, but it's highly unix-specific :-)
07:11 Sewi I asked the perlmonks about this: http://perlmonks.org/?node_id=846056
07:12 Sewi I also tried that usafe signals thing but it didn't help, too
07:12 Alias I wonder what the Wx filesystem layer does
07:12 Alias We're not using it atm
07:12 Alias But Wx being an async loop has an async file api
07:12 Alias (That we don't use)
07:14 Alias wxFile::Access
07:14 Alias That would be the most likely option
07:14 Alias Try it and see if it blocks too
07:16 Alias When does this problem occur for you?
07:16 Alias At the instant that the network filesystem drops?
07:17 Sewi For example if I unplug the cable from my netbook and switch over to wlan: It tried to access the mounts using the new local IP (can't use the wired IP any longer) and this is failing
07:23 Alias Well, whatever you are doing to fix this problem, try to keep it in unix-specific blocks
07:23 Alias Take a look at what I did to fix your code
07:23 Alias It's highly unlikely anything tricky will translate to Windows
07:25 Sewi I'd move the check into a task which could use -e and run in the background
07:25 Sewi The DB usage might be a small problem but should be solveable
07:35 Xaero joined #padre
07:41 daxim joined #padre
07:43 waxhead broken release?
07:43 waxhead :(
07:44 waxhead OK.. I'm heading out now... if there's a fix for the broken bits then I'll do another release tonight to limit the damage...
07:44 waxhead 0.65 is already uploaded to cpan...
07:47 Sewi What about the memory problems?
07:49 Sewi The about-dialog seems to be broken, too, it doesn't do any rendering
07:50 Alias wxahead: We'll need to do a 0.66 immediately
07:50 Alias waxhead: ^^^
07:50 Alias trun kis ready to go
07:51 danlucraft joined #padre
07:53 danlucraft1 joined #padre
07:58 wallberg joined #padre
11:14 waxhead Alias, trunk is right to release off?
11:16 Hyppolit svn: r11740 | waxhead++ | http://padre.perlide.org/trac/changeset/11740
11:16 Hyppolit Getting ready for bug fix release 0.66
11:16 Hyppolit trunk/Padre/
11:18 Hyppolit svn: r11741 | waxhead++ | http://padre.perlide.org/trac/changeset/11741
11:18 Hyppolit project tidy for release 0.66
11:18 Hyppolit trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Wx/ trunk/Padre/lib/Padre/Wx/Menu/
11:19 rindolf joined #padre
11:19 rindolf Hi all.
11:19 waxhead Alias, any last issues to go?
11:20 waxhead hey rindolf
11:20 rindolf waxhead: hi. What's up?
11:21 Hyppolit svn: r11742 | waxhead++ | http://padre.perlide.org/trac/changeset/11742
11:21 Hyppolit Bumping version to 0.66
11:21 Hyppolit trunk/Padre/lib/ trunk/Padre/lib/Padre/ trunk/Padre/lib/Padre/Action/ trunk/Padre/lib/Padre/Browser/ trunk/Padre/lib/Padre/Config/ trunk/Padre/lib/Padre/DB/ trunk/Padre/lib/Padre/DB/Migrate/ trunk/Padre/lib/Padre/Document/ trunk/Padre/lib/Padre/Document/Perl/ trunk/Padre/lib/Padre/Document/Perl/QuickFix/ trunk/Padre/lib/Padre/File/ trunk/Padre/lib/Padre/Plugin/ trunk/Padre/lib/Padre/Plugin/Devel/ trunk/Padre/lib/Padre
11:21 rindolf So, is 0.66 out soon?
11:25 waxhead yeah...
11:25 waxhead shorted ever release cycle I think...
11:25 waxhead 0.64->0.66 in less than a day
11:26 waxhead just doing release testing now.. then start building and then a release to test
11:31 waxhead Alias, ping
11:32 shadowpaste "waxhead" at 217.168.150.38 pasted "Tests failing for release." (11 lines) at http://paste.scsys.co.uk/45373
11:32 waxhead Alias, ^^^^
11:32 waxhead Sewi, ^^^^
11:34 Sewi waxhead: Did you try again?
11:34 waxhead try what again?
11:34 Sewi run the tests
11:34 * Sewi updating & testing...
11:34 waxhead Sewi, prove -lv t/02-new.t
11:36 waxhead oh.. these look like tests against config...
11:37 waxhead didn't Alias do something to improve the config ?
11:37 waxhead - Improved the quality and integration of the default window size (ADAMK)
11:37 waxhead I wonder if that's anything to do with it... when you look at the tests it looks like it's check the GUI stuff...
11:38 szabgab I recall seeing some reports about .padre being owned by root
11:38 Sewi 02 is still running here
11:39 szabgab I guess at least one of the tests creates it
11:40 waxhead Sewi, it's passing?
11:41 waxhead hmmm.. the last change I show here for that line is Adam
11:41 waxhead he's set it to -1
11:41 waxhead and the test is testing for 40
11:41 waxhead Can't raise him though to ask if that's what he meant...
11:43 waxhead we could set it back to the tests values?
11:43 Sewi It's waiting for STDIN input and throwing out strange error before the first test.
11:45 waxhead OK... setting Padre::Config back to the values the tests are expecting fixes this test...
11:45 waxhead however what I'm not sure about is if that is what Adam really meant...
11:45 szabgab that sounds bad
11:45 szabgab Alias, Alias Alias !!!!!  ^^^^^
11:46 szabgab Alias, wake up !!!!
11:46 waxhead szabgab, well, the thing is, if adam meant -1 then the tests need to be changed to -1...
11:46 waxhead so it's a bit of a 50/50 call...
11:46 szabgab we should try to understand what he wanted
11:46 waxhead change the test to the code, or the code to the tests????
11:46 szabgab otherwise those tests are meaningless
11:46 waxhead szabgab, from the change file:
11:46 Sewi which one is newer, test or code?
11:46 waxhead - Improved the quality and integration of the default window size (ADAMK)
11:47 waxhead code
11:47 waxhead I'd say
11:47 Sewi Update the test.
11:47 waxhead by setting -1 I'd say that it would trigger some sort of initial testing/checking before setting...
11:47 szabgab someone should hit Adam with some heavy material so he will run the tests before committing a change
11:47 Sewi This release needs heavy don't-use-except-for-testing-warnings anyway.
11:47 waxhead it's why we are release as unstable...
11:47 szabgab and make sure his intentions are backed up by the tests
11:49 Sewi Why do we need a newer Wx than 0.96 for rendering HTML correctly? Everything was find with the About page not long ago, now it's broken.
11:49 Sewi 4 tests failing in 02-new, all get -1 and expect other values
11:50 [Sno] joined #padre
11:50 * waxhead sent email to alias just in case it's quicker
11:51 waxhead Sewi, that's the failures I get
11:51 waxhead if you look at the 'blame' for Padre::Config you can see Adam changed those values to -1
11:52 waxhead is he on twitter?
11:52 * waxhead isn't
11:52 Sewi Send him a SMS or call him up (depending on your local time) :-)
11:52 szabgab waxhead, he is in Australia not in twitetr
11:52 waxhead don't have his mobile...
11:52 * waxhead thinks he should though!
11:53 waxhead at least I'm in the same time zone has him..
11:53 waxhead mind you I odn't keep his waking hours!
11:53 waxhead what game is he playing these days?
11:54 kthakore whats wrong folks?
11:55 waxhead failing tests...
11:55 waxhead just unsure what the intent was with the change...
11:56 waxhead either change the code to what the tests were testing for, or change the test to reflect the new code
12:00 jagd joined #padre
12:05 waxhead what's the plan then?
12:06 waxhead change the code or the tests?
12:06 Alias__ joined #padre
12:06 waxhead hey!!!
12:06 waxhead Alias
12:06 waxhead Alias__, can you check the back logs, or want the summary?
12:07 Alias__ summary
12:07 waxhead - Improved the quality and integration of the default window size (ADAMK)
12:07 waxhead prove -lv t/02-new.t
12:08 Alias__ checking
12:08 waxhead the problem appears to be in Padre::Config, where you set the values to -1
12:08 waxhead the tests are checking for 40,20,800,600
12:09 waxhead we were guessing what to do, change tests or change code..
12:11 Alias__ tests
12:11 Alias__ We used those 40,20,etc because the default -1,-1 of Wx sucked
12:12 Alias__ Padre::Wx::Display now provides a far better default than Wx itself
12:12 waxhead ok... you want to commit the changes?
12:12 waxhead or I'll do it if needs be
12:13 Alias__ I will
12:13 waxhead cool...
12:16 waxhead when?
12:16 waxhead :)
12:19 Hyppolit svn: r11743 | adamk++ | http://padre.perlide.org/trac/changeset/11743
12:19 Hyppolit Moving back to a pure default -1 rect
12:19 Hyppolit trunk/Padre/t/
12:22 waxhead ok.. tests pass
12:22 waxhead just testing more now
12:23 Alias__ ditto here
12:25 waxhead getting there...
12:26 waxhead that's better!
12:29 waxhead sweet...
12:31 waxhead anyone want to pip it?
12:32 waxhead still getting Scalar leaked, but installed OK on ubuntu...
12:32 waxhead about to tag it...
12:32 Alias__ Yeah, I expect to see some Scalar leaked
12:32 Alias__ We'll have to chase them down again, whoever did it last time
12:32 Alias__ I tried not to get hung up on details to land the big changes
12:32 waxhead steffen wasn't it?
12:33 Alias__ Can't recall, but probably
12:33 Alias__ I also don't understand WHY they happen
12:34 Hyppolit svn: r11744 | waxhead++ | http://padre.perlide.org/trac/changeset/11744
12:34 Hyppolit tag Padre-0.66
12:34 Hyppolit tags/
12:36 waxhead http://peshel.dnsalias.net/pa​dre-release/Padre-0.66.tar.gz
12:40 Alias__ I get test failures
12:40 Alias__ When installing via pip, but not with the checkout
12:43 waxhead Alias__, really?
12:43 Alias__ xt/actions.t ....................... 1/? Use of uninitialized value $devpl in concatenation (.) or string at xt/actions.t line 47 Use of uninitialized value $devpl in concatenation (.) or string at xt/actions.t line 47.
12:43 waxhead I'm still waiting on the vista box..
12:44 waxhead 0.66 installed fine on my windows box
12:44 Alias__ That said, it's an xt test
12:44 Alias__ So it will make no difference to an end user install
12:45 Alias__ I just happened to still have RELEASE_TESTING on
12:45 waxhead ok...
12:45 waxhead ah..
12:45 waxhead that does it!
12:45 waxhead So what was the problem with 0.65?
12:45 waxhead file locking on win32?
12:47 Alias__ Unix-specific filesystem code that crashed on Windows, right in the middle of the unlocking phase of Padre::Locker
12:47 Alias__ Which cascaded into the rest of the unlock process, and left the gui completely locked
12:48 kaare joined #padre
12:48 Alias__ Permanently, once you load any file
12:48 waxhead interesting..
12:48 waxhead ok.. it's in the release announcement...
12:48 Sewi Throw away this sysopen thing if it's a problem.
12:49 waxhead uploaded to pause
12:49 Alias__ Sewi: I wrapped if ( WIN32 ) { -f } else { NONBLOCK }
12:49 Sewi It helps in case of flocked/fcntl-locked files but didn't help on the problem I was investigating
12:49 Sewi Alias: I saw, but feel free to revert back to -e if there are any further problems.
12:52 Alias__ It's ok
12:52 Alias__ Although people might find it odd
12:53 Alias__ If locked files dissapear periodically from recent files
12:53 Alias__ But I'm happy to leave it for now
12:53 Alias__ (Since it doesn't run on my machine anyway)
12:55 Sewi Do you use any *ix?
12:57 Alias__ Not on the desktop
12:57 Alias__ Windows desktop, Linux servers
12:57 Alias__ For the most part
12:57 Hyppolit svn: r11745 | waxhead++ | http://padre.perlide.org/trac/changeset/11745
12:57 Hyppolit post release update
12:57 Hyppolit trunk/Padre/
13:00 Topic for #padreis now Latest Padre is 0.66 - If you have 0.65 upgrade now, 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
13:01 waxhead that's everything I think...
13:03 waxhead szabgab, did you get the url for 0.66?  ^^^
13:10 waxhead OK.. lets hope this release lasts a little longer...
13:21 Alias__ Hopefully it should be
13:25 waxhead Ok, to bed...
13:25 waxhead nite
13:35 kthakore Alias__: oh hai!
13:36 kthakore Alias__: how do we (frozen-bubble) make a desktop file for windows and linux?
13:36 Alias__ Did you see my use.perl post?
13:37 Alias__ Also, look at Padre::Desktop for the weirdo Win32 stuff you need to do
13:40 kthakore Alias__: yeah I did
13:40 kthakore Alias__: padre::desktop?
13:40 kthakore Alias__: oh also Frozen-Bubble is complete for windows now
13:41 Alias__ I played it this morning
13:41 Alias__ Only problem left is launching the thing
13:43 kthakore Alias__: ok
13:43 Alias__ kthakore: Fancy an online game?
13:43 * Alias__ is logged into the net
13:43 kthakore Alias__: fb?
13:43 kthakore sure
13:44 Alias__ I'm on the german server
13:44 kthakore Alias__: sure
13:46 Alias__ What happened?
13:46 kthakore Alias__: join a game
13:46 kthakore Alias__: my game
13:46 kthakore bah
13:46 kthakore Alias__: ok where?
13:46 Alias__ Start your game
13:47 kthakore done
13:48 kthakore bah
13:48 kthakore Alias__: how do I get balls onto you thing?
13:48 Alias__ Pretty sure you need to drop balls that aren't the matched colour
13:48 kthakore ah!
13:49 Alias__ Kinda funny you port it for months, and can't win  :)
13:49 kthakore yeah
13:49 kthakore ...
13:49 kthakore I suck
13:51 Alias__ Whoa, wth :)
13:51 kthakore I suck sooo much
13:52 Alias__ This is why I never can get around to writing games
13:52 Alias__ I much prefer playing other people's :)
13:53 kthakore hehe
13:53 kthakore Alias__: yeah I am trying to fix a nasty PERL_DESTRUCT_LEVEL bug
13:54 kthakore Alias__: well you can help me port the windows fb-server
13:54 kthakore Alias__: it is in C right now and can't compile
13:55 kthakore Alias__: good game!
13:55 kthakore Alias__: the only thing that doesn't work in windows is lan play
13:55 kthakore Alias__: becuase of that server
13:56 Alias__ ah
13:56 kthakore Alias__: bah he opened 3 more bugs
13:59 Alias__ :)
13:59 kthakore Alias__: gah ...
13:59 kthakore Alias__: I just wanna release this!
13:59 kthakore fracking bugs
14:00 Alias__ BTW, alpha blending breaks down for me if I turn graphics off the highest level
14:00 Alias__ But, tbh, I don't care :)
14:00 Alias__ I just want to work out how to wrap this in something simple enough for gamers to launch
14:01 kthakore Alias__: I fixed it!
14:01 kthakore Alias__: see github
14:01 Alias__ For strawberry professional, it should be easy to make a simple "Games for Perl" sub-menu
14:01 kthakore yeah
14:01 kthakore Alias__: I will leave that in your hands. I don't understand how desktop icons work
14:01 kthakore Alias__: I prefer a terminal
14:02 Alias__ Nobody uses desktop icons any more anyway really
14:02 Alias__ It's a bit of a holdover
14:03 Alias__ Most dialogs that offer to install one, default it off
14:03 Alias__ So I prefer just doing the start menu
14:03 kthakore we can put in start menu?
14:04 kthakore how?
14:04 kthakore Alias__: also is there a gui (for windows) CPAN installer?
14:04 Alias__ In Strawberry Professional, we would add a special entry for you
14:04 kthakore Alias__: yay!
14:04 Alias__ For the CPAN install, it's a way bigger problem
14:04 kthakore ok
14:05 Alias__ I know how to theoretically do it
14:05 Alias__ What I don't know is how to do it right and scalably
14:06 kthakore ok
14:06 kthakore Alias__: leave that for later I guess
14:07 Alias__ Yeah
14:07 Alias__ As far as I'm concerned, the REAL power here is in being in the default install
14:07 kthakore Alias__: it looks like besides web and oop stuff in perl. The rest of the scope is not much explored in perl land.
14:07 Alias__ In the same way as Windows Solitaire is
14:09 kthakore Alias__: YAYA!
14:10 kthakore Alias__: SDL now leaks nothing!
14:10 kthakore before
14:10 kthakore indirectly lost: 2,682,169 bytes in 53,383 blocks
14:10 kthakore and now
14:10 kthakore indirectly lost: 0 bytes in 0 blocks
14:10 kthakore MUAHAHAHA
14:10 kthakore before  definitely lost: 21,889 bytes in 92 blocks
14:10 kthakore and now
14:11 kthakore definitely lost: 14,119 bytes in 180 blocks
14:11 kthakore yay!
14:11 kthakore Alias__: bah ...
14:13 sorear I don't beleive in "definitely lost"
14:13 sorear Or has Rice finally been overturned?
14:15 kthakore sorear: well the definately lost is all in perl_yyparse
14:15 kthakore sorear: does that mean nonweakened references?
14:16 sorear what it *means* is that valgrind has solved the halting problem
14:16 sorear I have a suspicion it's lying
14:16 kthakore sorear: what?
14:26 kthakore sorear: yay! you were right!
14:27 kthakore sorear: valgrind lies on preallocated X11 buffers
14:27 kthakore I am mem leak free! MUAHAHHAA
14:28 sorear a mem leak isn't worth fixing unless it's >1MB or >O(1), imo
14:28 kthakore 14 kb is ok then?
14:28 sorear yes
14:28 kthakore sorear: well it was intdirectly leaking 2.3 MB before
14:28 kthakore sorear: without the destruct_level
14:28 sorear especially for a big app like Padre
14:28 kthakore sorear: this is for SDL
14:28 kthakore sorear: sowwy
14:43 patspam joined #padre
15:14 dapatrick joined #padre
15:16 rindolf joined #padre
15:19 didess joined #padre
15:27 rindolf joined #padre
16:04 Steffen joined #padre
16:06 danlucraft joined #padre
16:07 Hyppolit svn: r11746 | adamk++ | http://padre.perlide.org/trac/changeset/11746
16:07 Hyppolit The skeletal beginnings of a task that will let us replace Ack with our own search. It should give us better control over search, and let us remove some dependencies as a bonus.
16:07 Hyppolit trunk/Padre/lib/Padre/Task/
16:08 Hyppolit svn: r11747 | adamk++ | http://padre.perlide.org/trac/changeset/11747
16:08 Hyppolit trunk/Padre/lib/Padre/Task/
16:40 danlucraft oh cool
16:40 danlucraft Alias: what will you replace Ack with?
16:41 danlucraft Alias__: an indexing service? or will yo keep with a scanning method?
17:17 Alias__ Nope, a manual search
17:17 Alias__ Indexing is painful and a bit wasteful
17:17 Alias__ Plus, who knows what root they want to use
17:17 Alias__ So you'd have to index everything in the world
17:18 Alias__ And that way lies trouble
17:21 danlucraft Alias__: indexing a project wouldn't be too bad, would it?
17:23 kthakore danlucraft++
17:23 kthakore Alias__: atleast give us that!
17:29 * PerlJam wishes padre wasn't so "project specific" even.
17:30 PerlJam but, on demand indexing wouldn't be so bad (at least that way the user would be cognizant of the index enough to know when it needs to be rebuilt)
17:34 pece joined #padre
17:36 Sewi PerlJam: It starts to be bad once Padre starts indexing your .mp3 collection as text :-)
17:36 PerlJam or maybe even, on first search of a dir ... "I haven't seen this directory before, should I index it before continuing the search?"  or some such
17:36 PerlJam Sewi: there could be a predefined set of rules about which file types to not index
17:37 PerlJam It gets bad quick though whenever you do have indexing as there are all sorts of questions
17:38 Sewi PerlJam: You can't match everything. I got a directory with some 300.000 files organized in sub-dirs. It's application-specific neccessary to keep them this way, but Padre would start to run into things he shouldn't.
17:39 Sewi Next one: If a program starts scanning your hard disk, what would you think in terms of privacy?
17:39 PerlJam indeed.
17:40 PerlJam I see both sides though.  And removing the ack crutch from padre is a good first step to allow that part to be configurable :)
17:40 Sewi The new directory tree cache is already scanning masses of things it shouldn't - for the price of Padre eating much memory and cpu resources.
17:41 Sewi What about using any running indexer service? The user should have configured this (usually). There is one on Windows and I think there is one on Ubuntu, too.
17:41 Sewi Yep, it's called beagle
17:41 Sewi Simply re-use existing data
17:58 szabgab http://svok.blogspot.com/2010/07/anno​uncing-perl-5-documentation-team.html
18:07 * PerlJam still wishes that "perldoc blah" would just DWIM instead of "perldoc -f foo" or "perldoc -O '='" or "perldoc -q 'is a number'" etc.
18:36 szabgabx joined #padre
19:17 szabgabx weir,  perl Makefile.PL seem to be stuck after Writing makefile for Padre
19:18 szabgabx on Win 7   PAdre 0.66
19:35 szabgabx on trunk it seems to work well (after installing Module::Install)
19:47 danlucraft joined #padre
19:49 kgish joined #padre
20:04 kaare joined #padre
20:17 danlucraft joined #padre
20:20 patspam joined #padre
20:49 szabgabx joined #padre
20:55 patspam joined #padre
21:07 patspam joined #padre
21:46 patspam joined #padre
22:09 kaare left #padre

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