--- Log opened Fri Apr 06 00:00:56 2007
00:09 -!- tokyo [n=tokyo@c-67-175-134-199.hsd1.il.comcast.net] has joined #synfig
00:10 -!- zotz [n=zotz@24.244.163.157] has quit ["Leaving"]
01:51 -!- zloc [n=zloc@24.244.156.212] has quit [Read error: 110 (Connection timed out)]
01:54 -!- zloc [n=zloc@24.244.156.212] has joined #synfig
02:58 -!- teak [n=teak@24.244.163.157] has joined #synfig
03:12 -!- pxegeek [n=Miranda@c-71-59-140-184.hsd1.or.comcast.net] has joined #synfig
03:13 -!- pxegeek_ [n=chatzill@c-71-59-140-184.hsd1.or.comcast.net] has joined #synfig
03:13 -!- pxegeek [n=Miranda@c-71-59-140-184.hsd1.or.comcast.net] has quit [Client Quit]
03:13 -!- pxegeek_ is now known as pxegeek
03:14 -!- crazy_bus [n=philip@203.51.41.45] has joined #synfig
03:41 -!- crazy_bus [n=philip@203.51.41.45] has quit [Read error: 104 (Connection reset by peer)]
03:56 < teak> is anyone here having network problems tonight?
04:18 -!- zotz [n=zotz@24.244.163.157] has joined #synfig
04:28 -!- zotz [n=zotz@24.244.163.157] has quit ["Leaving"]
04:29 -!- teak [n=teak@24.244.163.157] has quit ["Leaving"]
05:01 < pxegeek> Not particularly...
05:44 -!- crazy_bus [n=philip@203.51.41.45] has joined #synfig
06:46 -!- crazy_bus [n=philip@203.51.41.45] has quit [Remote closed the connection]
07:32 < dooglus> pxegeek: how's the stability in the new windows build?
07:32  * pxegeek is embarrassed, but hasn't installed it yet
07:32 < dooglus> pxegeek: and those tango icons come as standard in xubuntu 6.06 by the way (ubuntu + xfce4)
07:33 < pxegeek> Work's been a bit hectic...
07:33 < pxegeek> Nice!
07:33 < dooglus> I don't remember whether you already had a version with my fix for inverted rectangles or not
07:33 < pxegeek> You want to copy the new icons down and see how they all look?
07:33 < pxegeek>  Most of the right hand side is done now
07:33 < dooglus> ok
07:34 < dooglus> got them in a single .zip file or something?
07:34 < pxegeek> I think 375 didn't have the inverted rectangle fix
07:34 < pxegeek> But I've not used that method since, and things seem much happier :_
07:34 < pxegeek> :)
07:35 < pxegeek> Can do... please hold...
07:35 < dooglus> it was fixed in r398
07:35 < dooglus> I've not had a crash for a long time now
07:36 < dooglus> but I have had a crash since r398
07:36 < pxegeek> I still think the Linux builds are more stable...  
07:36 < pxegeek> Quit distracting me - I'm trying to concentrate...
07:37 < pxegeek> You need the versions of the new icons with the old filenames...
07:38 < dooglus> the filenames have been changed to protect the innocent?
07:42 < pxegeek> Tango uses different icon names than Synfig
07:42 < pxegeek> I hadn't copied a coupl across into my synfig pixmap dir yet
07:43 < dooglus> I just noticed that the tango icons in the top 2 rows are bigger than the synfig icons:
07:43 < dooglus> http://dooglus.rincevent.net/random/1175838103.png
07:44 < dooglus> and also, there's space for 5 columns of icons, not 4
07:46 < pxegeek> I notice the top two rows are wider or spaced wider than the synfig icons... ?
07:46 < dooglus> I think the spacing is the same
07:46 < dooglus> but the tango icons are bigger
07:46 < dooglus> the 'save all' and 'about' icons are smaller though
07:46 < pxegeek> Try this - http://pxegeek.home.comcast.net/synfig/sf_icons.zip
07:47 < dooglus> so the spacing around them is more
07:48 < dooglus> thing is, I rebuild synfig over and over
07:48 < dooglus> and each time, it reinstalls the icons
07:48 < dooglus> so .sif files would be better
07:49 < pxegeek> You can grab them from the web page...  
07:49 < pxegeek> Or I can zip them up this week end
07:49 < pxegeek> I need to get to bed
07:49 < pxegeek> Up early tomorrow
07:50 < pxegeek> I've been rendering these at 128x128 like the existing icons
07:50 < pxegeek> Synfigstudio scales them down for display on the dialog
07:50 < dooglus> http://dooglus.rincevent.net/random/1175838600.png
07:50 < pxegeek> Can't you control how they're scaled?
07:51 < pxegeek> :)  Nice!
07:51 < dooglus> I expect I can control it.  I'll look into it
07:51 < pxegeek> My task for the w/e - figure out why scale renders darker than the others.  
07:52 < dooglus> the circle, rectangle, gradient icons look a bit as if they're disabled ('greyed out')
07:52 < pxegeek> They all use the same colour scheme
07:52 < dooglus> I'm not familiar with tango specs
07:52 < pxegeek> That's Tango icons for you - the whole look is 'cleaner'
07:52 < pxegeek> or in my view 'bland'
07:52 < pxegeek> ;)
07:52 < dooglus> maybe having half and half isn't the best
07:53 < dooglus> the saturated colours in 'draw' and 'sketch' make the tango icons look a bit washed out
07:53 < pxegeek> Yes.
07:53 < dooglus> don't let me keep you up :)
07:53 < pxegeek> There may be some optimization we can do for the small icon size
07:54 < pxegeek> I know the Tango ones are rendered at ~22x22 and hand tweaked
07:54 < pxegeek> pixel by pixel
07:55 < pxegeek> Do you have an ETA for your HD replacement?  Or are you enjoying running from the live CD
07:55 < dooglus> the laptop is only 4 months old
07:55 < dooglus> it's got a 1 or 2 year warrantee
07:55 < dooglus> so there's no rush
07:55 < dooglus> and yes, I quite like the live CD
07:56 < dooglus> it seems so much faster this way - everything I use is constantly in RAM
07:56 < dooglus> synfigstudio starts up so much faster, so does everything else
07:57 < dooglus> it seems to update the memory stick when the PC is idle, so I don't notice any delay from that
07:57 < dooglus> I'm just wondering how long the flash drive will last though - they have quite limited lifetimes I think
07:59 < pxegeek> depends on the filesystem and whether it knows how to play nicely w/ flash
08:00 < pxegeek> Well, work PC is shutdown, it's time to put this one to bed too...
08:00 < dooglus> I deliberately didn't use a journalled filesystem,
08:00 < pxegeek> 'nite
08:00 < dooglus> thinking that would be using the same area over and over
08:00 < pxegeek> Smart move!
08:00 < dooglus> I used ext2
08:00 < dooglus> the basic linux filesystem
08:00 < dooglus> I doubt it's smart
08:00 < dooglus> anyway, nite :)
08:00 < pxegeek> ;)
08:06 < dooglus> http://en.wikipedia.org/wiki/JFFS
08:12 -!- pxegeek [n=chatzill@c-71-59-140-184.hsd1.or.comcast.net] has quit [Read error: 148 (No route to host)]
10:35 -!- pabs3 [n=pabs@dsl-58-7-131-153.wa.westnet.com.au] has joined #synfig
11:36 < pabs3> whoa, atrus posted a flood of patches
11:36 < pabs3> dooglus: are you able to look at them? not sure I'll have time this weekend
11:37 < dooglus> sure
11:37 < dooglus> I was in the process of fighting gtkmm
11:37 < pabs3> yeah?
11:37 < dooglus> trying to convince it that I want dialogs to be 'transient for' both the toolbox and the canveses
11:37 < dooglus> but it was winning
11:38 < dooglus> para 3 of http://standards.freedesktop.org/wm-spec/1.3/ar01s07.html looks promising
11:38 < dooglus> if I could set the 'transient' property to 'None' for dialogs, they'd pop to the front when I select the toolbox or a canvas.
11:38 < dooglus> but I can't :(
11:39 < pabs3> hmm
11:39 < dooglus> it takes a Gtk::Window argument
11:39 < dooglus> how do you specify 'None' as type Gtk::Window?
11:40 < pabs3> NULL? dunno
11:41 < dooglus> computer says no
11:41 < dooglus> dockdialog.cpp:110: error: no matching function for call to 'studio::DockDialog::set_transient_for(NULL)'
11:41 < dooglus> /usr/include/gtkmm-2.4/gtkmm/window.h:607: note: candidates are: void Gtk::Window::set_transient_for(Gtk::Window&)
11:42 < pabs3> perhaps there is another function? otherwise some raw GTK+
11:49 < pabs3> new win32 version, atrus says it crashes on pirates.sif
11:49 < dooglus> oh bugger
11:50 < dooglus> this compiles, but doesn't seem to work: "gtk_window_set_transient_for(gobj(), NULL);"
11:51 < pabs3> perhaps there is a RootWindow class or something
11:52 < dooglus> the pirates crash here, too
11:52 -!- Atrus[work] [n=Atrus@87.228.121.204] has joined #synfig
11:53 < pabs3> dooglus: in linux SVN? that sounds like a regression
11:53 < dooglus> yes
11:54 < Atrus[work]> hi
11:54 < pabs3> pirates.sif used to open fine
11:54 < dooglus> I know
11:54 < Atrus[work]> Ahh, I see. ;-)
11:54 < dooglus> hi :)
11:55 < pabs3> hmm, still works fine in the version I've built
11:55 < dooglus> revision #?
11:56 < pabs3> changing my theme so I can look at the text in the about dialog :)
11:56 < pabs3> 394
11:56 < dooglus> this crashes, too: synfig -w 1 -h 1 -o pirates.png pirates.sif
11:56 < pabs3> synfig 394 and studio 369 rather
11:56 < Atrus[work]> I just check last win32 build before upload and found it. But I hove no mingw debug tool at work computer...
11:57 < dooglus> it crashes in Linux, too
11:57 < dooglus> I expect I have broken something
11:58 < Atrus[work]> ok. And also a dont see icon for "All Keyframes locked" button. Check it in evening. May be its just not included into win32 distro.
11:58 < Atrus[work]> :)
11:58 < pabs3> dunno if such a thing as svn-bisect exists, but you could go from synfig 394 to the latest manually
12:00 -!- Atrus[work] is now known as Atrus[gone_for_f
12:00 -!- Atrus[gone_for_f is now known as Atrus[afk]
12:01 < dooglus> r406 is a possible
12:02 < dooglus> that looks like the only possibility.
12:02 < pabs3> no warp layers though
12:04 < dooglus> I'll build r405 anyway
12:04 < dooglus> see whether that has the bug
12:05 < dooglus> this system isn't so quick to rebuild synfig when a few header files have changed.
12:05 < dooglus> the poor memory stick is having a hard time now
12:07 < dooglus> doh.  r405 doesn't build.  some idiot left a prototype out of rectangle.h
12:08 -!- Atrus[afk] is now known as Atrus[work]
12:11 < pabs3> :)
12:13 < pabs3> dooglus: wow, r427 is rather a strange thing to have to fix
12:15 < dooglus> yes
12:16 < dooglus> I don't know what happened there - | and & were used pretty consistently throughout that file
12:16 < dooglus> but not in any other file
12:16 < dooglus> || and && should work faster, because they shortcut
12:16 < dooglus> A && B doesn't evaluate B if A is false
12:16 < pabs3> ya
12:17 < dooglus> r400 is OK; r405 is bad
12:19 < dooglus> I guess it has to be 401 then
12:20 < dooglus> yup
12:20 < dooglus> 401 crashes, 400 is OK
12:21 < pabs3> what about 401 plus the r421 patch?
12:22 < dooglus> oh, without the patch it won't compile...
12:22 < dooglus> I don't think r401 is bad - I think it's probably showing up a bug that was hiding before
12:23 < dooglus> all 401 does is says that stuff hiding behind opaque rectangles doesn't need rendering
12:23 < pabs3> hmm, ok
12:24 < dooglus> I'll look into it
12:27  * pabs3 bbiab
12:47 -!- crazy_bus [n=philip@203.51.41.45] has joined #synfig
12:55 < pabs3> got a g++ 4.3 ETL ftbfs patch, committed it
12:56 < dooglus> ok
12:56 < dooglus> I've stripped pirates down to just 'Darrin', 'Goat' and 'Rectangle040', and it still crashes
13:01 < dooglus> problem seems to be something to do with rectangles which are smaller than 1 pixel when rendered
13:01 < dooglus> if is_solid_color() is true for them, and they're < 1 pixel, then it crashes
13:01 < dooglus> I extended the range of rectangles for which is_solid_color() is true,
13:02 < dooglus> and in doing so, extended the visibility of the bug
13:02 < dooglus> anyway, that's my story and I'm sticking to it
13:02 < pabs3> :)
13:02 < dooglus> it was like that when I found it
13:03 < dooglus> here's a test...
13:03 < dooglus> in your old version of studio, can you change pirates:
13:04 < dooglus> "Image" -> "Dead Heads" -> "Darrin" -> "Rectangle040"
13:05 < dooglus> change the compose type to 'Straight'
13:08 < pabs3> didn't seem to do anything
13:09 < dooglus> zoom out a bit
13:09 < pabs3> how far?
13:09 < dooglus> until it crashes :)
13:09 < Atrus[work]> =)
13:09 < pabs3> hasn't crashed at 6.2%
13:10 < dooglus> oh
13:10 < pabs3> can't seem to zoom out any more, no crash
13:10 < dooglus> ok
13:11 < pabs3> hmm, the - button doesn't do anything, but it isn't disabled
13:11 < pabs3> (ie, no change in the ruler bars)
13:12 < dooglus> is the last top-level layer enabled?
13:12 < dooglus> (base image)
13:12 < pabs3> nope
13:12 < dooglus> can you enable it?
13:13 < pabs3> enabling it in the unmodified image at default on-load zoom level caused a crash 
13:14 < pabs3> synfigstudio: /home/pabs/devel/misc/synfig/install/include/ETL/_handle.h:733: T* etl::loose_handle<T>::operator->() const [with T = synfig::Canvas]: Assertion `obj' failed.
13:14 < dooglus> yes, for me too
13:14 < dooglus> I guess that's a different bug
13:15 < dooglus> it might be the same bug as when you add a new 'paste canvas' layer from the context menu
13:16 < pabs3> ah, that is in the tracker iirc
13:18 < pabs3> heh, the dead heads in that pic must be the diff voria employees
13:19 < dooglus> I guess
13:21 < Atrus[work]> So... I I understand right - there is a two crash bug found or just one? ;-)
13:21 < Atrus[work]> s/I /If /
13:22 < pabs3> for me just the one when enabling base image
13:23 < pabs3> anyways, gotta go
13:23 -!- pabs3 [n=pabs@dsl-58-7-131-153.wa.westnet.com.au] has quit ["Don't rest until all the world is paved in poems."]
13:26 < dooglus> Atrus[work]: seems to be 2
13:26 < dooglus> Atrus[work]: but one of them is probably one we already knew about
13:27 -!- teak [n=teak@24.244.163.157] has joined #synfig
13:27 -!- crazy_bus [n=philip@203.51.41.45] has quit [Read error: 104 (Connection reset by peer)]
13:28 < Atrus[work]> Well, funny. I just want make sure what synfig work perfectly before upload it and open random example... =)
14:06 -!- zotz [n=zotz@24.244.163.157] has joined #synfig
14:11 < dooglus> Atrus[work]: I've found the bug that was causing pirates to crash
14:11 < dooglus> I'll check in a fix as soon as I've tested it
14:12 < Atrus[work]> ok
14:14 < dooglus> ok, it's revision 430.
15:17 < dooglus> I just checked in your patches.  Thanks a lot for them.
15:17 < dooglus> (r431-r440)
15:18 < Atrus[work]> good. ;-)
15:19 < Atrus[work]> I think - may be make new repo for single synfig installer for windows in future?
15:25 < dooglus> a single installer is a good idea
15:25 < dooglus> but a new repo?  what do you mean?
15:26 < Atrus[work]> I mean: in current situation (core and studio - separate projects) there no way to make single installer, based only at one of it.
15:26 < dooglus> oh, really?
15:26 < dooglus> they're in the same svn repository
15:27 < Atrus[work]> I think it may be to make a set of shell (or makefile) scripts which will make core and studio both together and produce the single installer (may be with Gtk+/Gtkmm integrated into)
15:27 < Atrus[work]> yes, I understand, but they hmm... they looks like separate project.
15:28 < Atrus[work]> I mean, what studio installer required builded and installed core, which is not so easy in windows
15:29 < Atrus[work]> right now core builded twice: onece for installer and onece for temorarry installation into mingw for make graphics for studio
15:31 < Atrus[work]> this is a reason for a studio has personal installer title picture but core does not, because when core is making it cant render its own projects before installation.
15:31 < Atrus[work]> Right now I'm trying to find the way - how to make them together.
15:33 < Atrus[work]> And the one of question is where to keep the single installer scripts, besause it not a part of core and not a part of studio, but required both sources of them.
15:35 < Atrus[work]> Heh, looks like I'm overload you... ;-)
15:52 < dooglus> sorry
15:52 < dooglus> I got distracted
15:52 < dooglus> keep them with studio I think
15:53 < dooglus> because synfig-core should be able to be built with studio,
15:53 < dooglus> but studio is never built without core
15:53 -!- zotz [n=zotz@24.244.163.157] has quit ["Leaving"]
15:53 -!- teak [n=teak@24.244.163.157] has quit ["Leaving"]
15:53 < dooglus> I found a way to get studio and core checked out as a single project
15:54 < dooglus> using 'svn switch'
15:55 < dooglus> do a 'svn co http://svn.voria.com/code/' and as soon as you see it checking out 'trunk/...', hit control C and use 'svn switch' to tell it not to get all the branches
15:55 < dooglus> I'll make a wiki page about it
15:55 < Atrus[work]> ok, I'll see.
15:55 < Atrus[work]> I mean I'll read it. =)
16:20 < dooglus> here: http://wiki.synfig.com/Subversion
16:20 < dooglus> it's a big of an annoying process, but you only need to do it once
16:20 < Atrus[work]> ok
16:36 -!- Bombe [n=droden@wing.pterodactylus.net] has quit [Read error: 110 (Connection timed out)]
16:40 -!- Atrus[work] [n=Atrus@87.228.121.204] has quit ["?????"]
23:22 -!- zotz [n=zotz@24.244.163.157] has joined #synfig
23:30 -!- zipola [n=zipola@zip.kortex.jyu.fi] has quit ["Abiit"]
--- Log closed Sat Apr 07 00:00:30 2007