--- Log opened Mon Jan 07 00:00:30 2008
00:20 < pxegeek> morning - take your time...
00:20  * pxegeek not feeling that great all arund
00:28 < pxegeek> https://pastebin.ca/844482
00:31 < pxegeek> when libtools are linked with -avoid-version, it tries to compile libmod_particle.dll instead of libmod_particle-0.dll
00:31 < pabs3> I think that was the point
00:32 < pxegeek> Windows uses libsynfigapp-0.dll, libsynfig-0.dll...
00:34 < pabs3> yeah, looks like the nsis scripts simply need updating - doing that now
00:35 < pxegeek> Ah - now we'll really see how good the uninstall/reinstall is 
00:37 < pabs3> the modules are written out withough the -0 in them though
00:38 < pabs3> really all these .nsh files shouldn't be nessecary
00:40 < pabs3> pxegeek: try this patch: http://bonedaddy.net/pabs3/files/tmp/fix-win32-installer.diff
00:41 < pxegeek> on 1272?
00:42 < pxegeek> Rhetorical - just opend the diff
00:42 < pabs3> yeah, on latest svn
00:56 -!- genete [n=Genete@84.122.34.88.dyn.user.ono.com] has joined #synfig
00:59 -!- dyloxin [n=dyloxin@c-71-198-19-209.hsd1.ca.comcast.net] has quit [Read error: 104 (Connection reset by peer)]
01:01 -!- dyloxin [n=dyloxin@c-71-198-19-209.hsd1.ca.comcast.net] has joined #synfig
01:02 < pxegeek> patched - compiling...
01:23 < KiBi> pabs3: BTW, I don't think rore is hitting this gettext issue.
01:23 < KiBi> pabs3: At least, the .mo is created and installed where it belongs.
01:23 < KiBi> But maybe the i18n support wasn't activated, indeed.
01:24 < pabs3> hmm, ok. check if ENABLE_NLS is set in config.h, otherwise the textdomain/etc functions aren't called 
01:24 < KiBi> pabs3: Will check that with her once she's up, a bit lazy right now.
01:24 < KiBi> Yup, reminding that.
01:24 < pabs3> cool
01:24 -!- ZanQdo [n=Daniel@201.201.2.22] has joined #synfig
01:24 < KiBi> Well, actually... /me ssh erwin.
01:25  * KiBi curses autopoint.
01:25 < KiBi> pabs3: That's it. Will look tomorrow for a workaround and document it.
01:26 < KiBi> rore:  ^^^^^
01:26 < pabs3> KiBi: looks like you/I wrote one on http://synfig.org/Translation
01:27 < pabs3> oh, wait that won't work
01:27 < KiBi> pabs3: yep, that's what she was doing.
01:27 < KiBi> But that doesn't activate NLS support.
01:27 < KiBi> Hence her problem.
01:27 < KiBi> Thus, rather?
01:28 < pabs3> hmm, hence I think
01:29 < KiBi> ok
01:29 < pabs3> hmm, probably should forward my email to bug-gnu-gettext@gnu.org
01:30  * pabs3 does that now
01:32 < genete> dooglus: the following configuration seems not to work:
01:32 < genete> Duplicate
01:32 < genete> Inline canvas
01:32 < genete> *Translate (i)
01:33 < genete> *Layer
01:33 < genete> if "i" is the duplicate index
01:33 < genete> it seems that duplicates the circle "i" times and then translate it "i" times. 
01:34 < pxegeek> Hi genete!
01:34 < genete> Although Translate layer have a random origin based on a "i" seed
01:34 < genete> hi pxegeek 
01:34 < genete> :)
01:36 < genete> oops IT works!
01:38  * pxegeek looking forward to trying the duplicate layer out
01:38 < pxegeek> Also looking forward to seeing your T-Shirt!
01:39 < genete> the T shirt on the oven... :)
01:39 < pxegeek> ?
01:39 < pxegeek> The January challenge
01:39 < genete> I know
01:40 < genete> it is cooking in my mind
01:43 < pxegeek> Ah- "on the backburner"
01:43 < genete> yes, sorry for my bizarre English
01:44 < pxegeek> Hey - it's better than my Spanish
01:45 < genete> I'm now working with dooglus on the new Dup layer and don't want to loose the inertia
01:48 < pabs3> KiBi: do you think hacking config.h after the configure would be enough? http://synfig.org/Translation#Testing
01:48 < KiBi> pabs3: Might be enough, indeed.
01:48 < KiBi> pabs3: I'll check later on, playing a teg game.
01:49 < KiBi> (pabs3: I don't want to mess with her local copy with her permissions, have to get another checkout)
01:49 < pabs3> ok
01:50 < KiBi> pabs3: maybe emphasize *before* running make.
01:52 < KiBi> pabs3: heh, no INSTALL, no autogen.sh :)
01:52 < pabs3> there used to be a hacky bootstrap script. it was moidered though. muahahahha
01:53 < KiBi> I'm just amused there's nothing at all now :)
01:54 < pabs3> thats what the build instructions page is for :)
01:55 < pabs3> and automake/autoconf gives us INSTALL for released tarballs 
01:55 < KiBi> ok
01:56 < KiBi> Is there something like git-clean in svn?
01:57 < KiBi> (that is not: svn st| grep ^? | awk ... | xargs rm)
01:57  * KiBi does that in the meanwhile.
01:58 < pabs3> hm
01:58 < pabs3> make maintainer-clean works for me
01:58 < KiBi> tututu, I said svn :)
01:58 < pabs3> :)
01:59 < KiBi> ok, running your line before make isn't OK.
01:59 < pabs3> indeed :)
01:59 < KiBi> (no .mo installed)
02:00 < KiBi> oh, stupid me.
02:02 < KiBi> Actually, ok for the .mo file.
02:02 < KiBi> Let's "make" now, 'later.
02:04 < KiBi> hmm
02:05 < KiBi> http://paste.debian.net/46311
02:07 < pabs3> pxegeek: how did the patch go?
02:08 < pxegeek> You psychic?
02:08 < pxegeek> It just finished compiling
02:08 < pxegeek> (I don't know why it took so long....)
02:09 < pxegeek> looks promising
02:10 < pxegeek> until I try and run it....
02:10 < pxegeek> Setup says
02:10 < pxegeek> Error opening file for writing:
02:11  * pabs3 wonders if there is ccache on windows
02:11 < pxegeek> C:\Program Files\synfig\bin\libsynfig-0.dll
02:11 < pxegeek> abort/retry/ignore....
02:11 < pabs3> do you have studio open or something?
02:12 < pabs3> did you remove the win32build dir?
02:13 < pabs3> is src\synfig\.libs\libsynfig-0.dll available
02:18 < pxegeek> D'oh - Oh yes. that'll do it....
02:18 < pxegeek> Maybe need a check for that when running the installer.
02:19 < pxegeek> Much happier.  
02:21 < pxegeek> And duplicate shows up as a layer!
02:22 < pabs3> pxegeek: what version of gettext do you have?
02:22 < pxegeek> $ gettext --version
02:22 < pxegeek> gettext (GNU gettext-runtime) 0.16.1
02:22 < pxegeek> Copyright (C) 1995-1997, 2000-2006 Free Software Foundation, Inc.
02:22 < pxegeek> This is free software; see the source for copying conditions.  There is NO
02:23 < pxegeek> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
02:23 < pxegeek> Written by Ulrich Drepper.
02:23 < pxegeek> Looks like I just have runtime?  No headers?
02:23 < pabs3> hmm, why did I put 0.13 in the configure.ac then
02:23 < pxegeek> I upgraded since then :)
02:25 < pabs3> KiBi: apparently AM_GNU_GETTEXT_VERSION([0.13]) means install gettext.m4 from gettext 0.13 instead of install gettext.m4 from >= 13
02:26 < pabs3> I
02:26 < pabs3> 'll upgrade the requirement to one that doesn't have the problem
02:27 < pxegeek> http://dooglus.rincevent.net/synfig/logs/2007/%23synfig-2007-11-29.log - about a third of the way down...
02:27 < KiBi> pabs3: Reply from upstrema?
02:27 < KiBi> *eam
02:27 < pabs3> ya
02:29 < pxegeek> I now have gettext in /bin rather than /C/GTK/bin
02:29 < pxegeek> (don't know if it works mind you)
02:31 < pxegeek> I'm still trying to figure out how this build process works.
02:32 < pxegeek> make package is the step that creates the Windows installer, right?  
02:32 < pabs3> iirc, yes
02:32 < pxegeek> Why does the script then turn around and 'make  clean' right after that?
02:35 < pabs3> dunno, thats how atrus wrote it I guess
02:50 < KiBi> pabs3: Any idea for my trace, btw?
02:50 < KiBi> (http://paste.debian.net/46311?
02:50 < pabs3> KiBi: committed a change to make it use 0.14.4
02:50 < KiBi> s/?/)/
02:50 < KiBi> OK, will try that.
02:51 < pabs3> KiBi: not sure, possibly related to the recent changes by daniel macks
02:52 < KiBi> pabs3: Anyway, I'll see whether that helps WRT config.h
02:52 < KiBi>    language is requested. */
02:52 < KiBi> #define ENABLE_NLS 1
02:52 < KiBi> \o/
02:52 < KiBi> blah, svn st doesn't show hidden files...
02:53 < pabs3> cool
02:53 < KiBi> (hence an incomplete cleanup last time, and possibly the above-mentioned issue)
02:59 < pabs3> pxegeek: committing that patch now
03:02 < CIA-37> synfig: pabs * r1273 / (2 files in 2 dirs): Install gettext 0.14.4 macros since earlier ones have a bug that disables NLS on amd64
03:02 < CIA-37> synfig: pabs * r1274 /synfig-core/trunk/src/modules/ (21 files in 21 dirs): Fix building the Win32 installer: update Win32 install scripts to remove versions from the module dll filenames because r1261 added -avoid-version which changed all the filenames of the modules
03:18 < KiBi> pabs3: Works out of the box.
03:18 < pabs3> cool
03:18 < KiBi> pabs3: autoreconf -vfi && ./configure --prefix=/usr/localkibi && make && sudo make install # that's it.
03:19 < KiBi> (Once po/fr.po added and ALL_LINGUAS set, of course.)
03:19  * pabs3 uses --prefix=~/opt
03:20 < KiBi> Point.
03:26 < pabs3> genete: 6 fuzzy translations, 1 untranslated message.
03:26 < genete> hi pabs3, for core?
03:26 < pabs3> yeah, don't have your studio translation yet :)
03:27 < genete> ;)
03:27 < genete> I'll rebuild tomorrow and translate then.
03:27 < genete> now it is too late...
03:27 < genete> bed time
03:28 < genete> BTW, just achieved the particles thing with the new duplicate layer
03:28 < genete> now the files are much smaller
03:29 < pabs3> actually, it is only a few strings, so maybe wait till we are closer to the release
03:29 < genete> ok
03:30 < pabs3> dooglus: btw, have you been adding new files with strings for translation in them to the po/POTFILES.in list?
03:36 < genete> night all
03:36 -!- genete [n=Genete@84.122.34.88.dyn.user.ono.com] has quit ["Abandonando"]
03:53 < dooglus> pxegeek: I think what happens is it makes the .exe installer for core, cleans up, then builds everything again with different paths, for the following studio build to link to.  and atrus says "it was like this when I got here, guv" (not an exact quote)
03:54 < pxegeek> yes - it puts synfig in a temp dir
03:54 < pxegeek> so it can build the icons
03:54 < dooglus> pabs3: I've not been doing that, no.  I didn't know about that.
03:55 < pxegeek> Some low hanging fruit for optimization
03:56 < dooglus> genete: how's the speed using the duplicate layer compared to manually duplicating the layers?  I guess the load time is much quicker now, but what about rendering?
03:58 < pabs3> dooglus: yeah, xgettext needs to know which files to extract strings from
04:04 < pxegeek> dooglus - do you have a description for duplicate for http://synfig.org/Layer
04:04 < pxegeek> ?
04:04 < dooglus> pxegeek: I can do
04:04 < dooglus> pabs3: ok
04:06 < pxegeek> Ok - from scratch.  I have a circle, how do I use duplicate to get two circles?
04:07 < dooglus> put duplicate layer over the circle
04:07 < dooglus> expand duplicate layer's parameter
04:07 < pxegeek> yup
04:07 < dooglus> set the 'to' parameter to 2
04:07 < dooglus> export the 'index' parameter
04:08 < dooglus> in the circle layer, link to it somehow
04:08 < dooglus> link it to the radius, and set the amount to 0.5
04:08 < dooglus> you'll get 2 circles of different sizes on top of each other
04:09 < pxegeek> I'm missing the 'link to it somehow' step
04:09 < dooglus> ok, specifically
04:09 < dooglus> select the circle layer
04:09 < dooglus> go to the 'children' dialog
04:09 < pxegeek> convert radius to?
04:10 < pxegeek> OK
04:10 < dooglus> expand the valuenodes
04:10  * pxegeek nods
04:10 < dooglus> select the index parameter you already exported
04:10 < dooglus> right-click the circle's 'radius' and 'connect'
04:10 < dooglus> that will connect it to the selected child
04:10 < dooglus> make the amount 0.5 to see the change
04:10 < pxegeek> Ah - wasn't selecting the exported value first
04:11 < pxegeek> connect wasn't an option
04:11 < dooglus> right
04:11 < dooglus> nothing to connect to
04:11 < dooglus> you see 2 circles?
04:11 < pxegeek> virtually ;)
04:11 < pxegeek> they're right on top of each other.
04:11 < dooglus> in the dup layer, play with from/to/step
04:11  * pxegeek try moving centers instead
04:12 < dooglus> step 0.5 will make twice as many (nearly)
04:12 < dooglus> right-click center
04:12 < dooglus> convert to composite
04:12 < dooglus> link x to the exported index
04:12 < pxegeek> Because vector is the wrong type...
04:12 < pxegeek> got it
04:12 < dooglus> yeah
04:13 < dooglus> where do you want the circle?   2.3  -- doesn't work as an answer
04:13 < dooglus> what should the X coordinate by?  2.3  -- that's fine
04:13 < dooglus> (without the typo it's fine, anyway)
04:14 < pxegeek> eas :)
04:14 < dooglus> ndede
04:14 < dooglus> also, randomness: make each a different color:
04:14 < dooglus> convert the color to random
04:14 < dooglus> and link the seed to the index
04:15 < dooglus> (except the seed is integer and the index is real, so I'm not too sure)
04:15 < dooglus> genete has a trick for that if necessary
04:15 < pxegeek> Or have a line of circles of varying shades of red...
04:15 < pxegeek> I'm going to have to play a lot more with exporting values and linking them....
04:16 < dooglus> I'm setting up the same example
04:16 < dooglus> what from/to/step did you use?
04:17 < pxegeek> 1 to 2 step 1.  It's not very exciting
04:17 < dooglus> ok, I wouldn't call that "varying shades" :)
04:18 < pxegeek> Well, no, sot yet - just trying to figure scale out
04:18 < dooglus> are they overlapping?
04:18 < pxegeek> yes.  
04:18 < pxegeek> Why does 1 = 60pt?
04:18 < dooglus> because you're using pts
04:18 < dooglus> it's 1 inch
04:18 < dooglus> no
04:18 < dooglus> um
04:18 < dooglus> 1 unit
04:18 < dooglus> :)
04:19 < pxegeek> 0 to 2 in step1 gives a black and two red circles :)
04:19 < dooglus> from the toolbox > file > setup > misc
04:19 < dooglus> you can choose the unit system to use
04:20 < pxegeek> (or a black and a red and a very red:) )
04:20 < dooglus> disconnect the radius
04:21 < dooglus> so you can get separate circles
04:21 < dooglus> set the amount back to 1
04:21 < dooglus> so their colours don't mix
04:21 < dooglus> then you'll see what you really have
04:21 < pxegeek> So I changed the duplicate layer to 0 to 1 in step 0.1
04:22 < pxegeek> Gives ~10 circles black through red
04:22 < dooglus> amount?
04:22 < dooglus> (on the circle)
04:22 < pxegeek> 1
04:23 < dooglus> hmm
04:23 < dooglus> you linked the index to the random seed?
04:23 < pxegeek> You can see the gradient on the left hand edge
04:23 < dooglus> hmm - made studio crash for me "same_type_as(x)' failed"
04:24 < pxegeek> no
04:24 < dooglus> so why are they different colors I wonder
04:24 < pxegeek> deliberately not.
04:24 < pxegeek> I linked the red channel to the exported value
04:24 < pxegeek> It's behaving as I'd expect - don't worry!
04:25 < dooglus> aah, that's fine then
04:26 < pxegeek> So if I want to increase the separation of the centers, do I first have to convert the x-axis to scale and use my exported value as the argument for the scalar?
04:26 < dooglus> you can do that
04:26 < dooglus> or you can play with the dup layer parameters from/to/step
04:26 < dooglus> but that will mess up your red going from 0 to 1 of course
04:26 < pxegeek> But that would then mess up the colors
04:26 < pxegeek> jinx!
04:27 < dooglus> so yes.  wot you said...
04:27 < dooglus> you might want to take 0.5 off the value before scaling it
04:27 < dooglus> so it's centered on the canvas
04:27 < pxegeek> centering's overrated :)
04:27 < pxegeek> That's what translate layers are for!
04:28 < dooglus> ok, take 0.4 off it to have it jauntily non-centered
04:28 < pxegeek> Hmm...  unexport works a little too well. 
04:28 < pxegeek> Didn't mean to delete that 
04:29 < dooglus> it doesn't break the linkage
04:29 < dooglus> and you can re-export it
04:29 < pxegeek> Or just hit undo a few times
04:31 < dooglus> yup
04:31 < pxegeek> Whee!  
04:31 < dooglus> you can nest these things too
04:31 < pxegeek> One thing at a time!
04:31 < dooglus> so your row of circles can be duplicated vertically, just by putting another dup layer above the circle
04:32 < dooglus> exporting its index as a different name, and linking that to 'y'
04:34 < pxegeek> got it
04:42 < pxegeek> Ooops - canvas renderer hiccupped
04:46 < pxegeek> looks like darco escaped lightly if he only had power cuts....  http://www6.comcast.net/news/articles/general/2007/12/17/Winter.Storm/
04:49 < dooglus> hmmm - but:  don't copy&paste a duplicate layer to make a 2nd duplicate layer if you've exported its index already, or it gets into an infinite loop
04:49 < dooglus> (2 nested loops using the same loop index)
04:50 < pxegeek> I was wondering about things like that....
04:50 < pxegeek> Here's the hiccup...  http://home.comcast.net/~pxegeek/synfig/hiccup.PNG
04:51 < pxegeek> Fixed itself when I hit refresh
04:53 < dooglus> expecting a square grid?
04:58 < pabs3> pxegeek: http://synfig.org/forums/viewtopic.php?p=66#66
04:58 < pxegeek> :)
04:59 < pabs3> pxegeek: you can use DeltaCopy to sync the files: http://www.aboutmyip.com/AboutMyXApp/DeltaCopy.jsp
05:00 < pabs3> best store them on a usb disk in case you get a virus or something I recon
05:01 < dooglus> pabs3: I can keep a copy too if that helps
05:01 < pabs3> pxegeek, darco and I should be enough
05:01 < dooglus> ok
05:01 < pxegeek> How often are they being generated on the server?
05:02 < dooglus> "daily, monthly and weekly"?
05:03 < pabs3> I'm looking into making dumps without the password hashes in them that we can autosync to sourceforge
05:03 < pabs3> dooglus: right, but the monthly/weekly ones haven't yet been run, I'll do a cp now
05:04 -!- ZanQdo [n=Daniel@201.201.2.22] has quit ["Adios"]
05:05 < pabs3> hmmm, LGM is slowly approaching
05:05 < dooglus> pxegeek: can you reproduce that glitch?  I can't
05:06 < pxegeek> No.  That session of Synfig expired.  Has ceased to be.  Gone to meet it's maker
05:06 < pxegeek> I tried to create a gif of grid expanding from a single spot.  
05:07 < pxegeek> It got confused because the path had spaces in it 
05:07 < pxegeek> I tried to kill it, and it didn't like that.
05:11 < pxegeek> pabs - it looks like deltacopy expects to be run on both server and client
05:15 < pabs3> ya, weird
05:16 < pxegeek> I'm also having trouble seeing how I can map the drive
05:17 < pxegeek> I always connect to darco's machine using putty
05:19 < pabs3> the other option would be winscp
05:19 < pabs3> but that transfers a full copy each time rather than just the changed bits
05:20 < pxegeek> I was just going to ask about that.
05:20 < pxegeek> ~100Meg a week won't kill me :)
05:20 < pabs3> there is the command-line rsync prog in cygwin btw
05:27 < pxegeek> Hmmm...  about 15 mins to copy.  I'll have to look into rsync options
05:38 < pxegeek> If you copied them to htdocs, we could grab them by http ;)
05:40 < pabs3> just noticed phpBB3 has been released
05:41 < pabs3> they contain user password hashes and other private data, so I'd rather not - will look at a way to strip those out
05:46 < pxegeek> (I wasn't serious - hence the ;) )
05:47 < pxegeek> OK - if the virtual host dies horribly today - we're covered.
05:47 < pabs3> ah :)
05:48 < pxegeek> In your copious leisure time, maybe you could put together a disaster recovery plan (aka what to do with these files)
05:48 < pxegeek> Ugh - didn't mean to sound condescending there. 
05:50 < pabs3> :) sure
05:53 < pabs3> the files tarballs contain REINSTALL.txt that shows the customisations we have
05:54 < pabs3> I'll expand them a bit
05:59 < pxegeek> Maybe one day we could build a virtual server and re-install from scratch
06:00 < pxegeek> Sometime when I'm on sabbatical maybe
06:00 < pxegeek> No substitute for actually doing it.
06:00 < pxegeek> dooglus - found another hiccup
06:02 < pxegeek> http://home.comcast.net/~pxegeek/synfig/duplicatetest.gif
06:02 < pxegeek> check the top lefthand corner
06:02 < pabs3> indeed
06:02 < pxegeek> 2nd ball down
06:24 < pxegeek> dooglus - last post for the night
06:24 < pxegeek> http://synfig.org/forums/viewtopic.php?p=67#67
06:26 < pxegeek> Thanks for the NSIS patches pabs!
06:26 < pxegeek> 'night all
06:27 < pabs3> np
06:27 < pabs3> nite pxegeek 
06:52 -!- pxegeek [n=chatzill@c-71-59-140-184.hsd1.or.comcast.net] has quit [Read error: 110 (Connection timed out)]
09:51 -!- TMM [n=hp@c5147518c.cable.wanadoo.nl] has quit ["Ex-Chat"]
10:20 -!- AkhIL [n=AkhIL@90.188.195.74] has left #synfig []
10:20 -!- AkhIL [n=AkhIL@90.188.195.74] has joined #synfig
10:39 -!- ZanQdo [n=Daniel@201.201.2.22] has joined #synfig
12:35 -!- ZanQdo [n=Daniel@201.201.2.22] has quit ["Adios"]
12:38 < dooglus> pxegeek: I'd be more worried about http://home.comcast.net/~pxegeek/synfig/duplicatetest.gif if it wasn't a gif.  gif images are prone to that kind of corruption, due to a dodgy .gif target in synfig.  if you render to .png do you see the same?
12:47 -!- ZanQdo [n=Daniel@201.201.2.22] has joined #synfig
12:48 -!- ZanQdo [n=Daniel@201.201.2.22] has quit [Read error: 104 (Connection reset by peer)]
12:51 -!- genete [n=Genete@84.122.51.41.dyn.user.ono.com] has joined #synfig
12:51 < genete> hi!
12:52 -!- ZanQdo [n=Daniel@201.201.2.22] has joined #synfig
12:55 < dooglus> pabs3: atrus has posted some patches for the build process on the patch tracker.  do you want to take a look at them some time?
12:58 < dooglus> hi genete
12:58 < genete> hi dooglus!
12:58 < genete> Particles are running with Duplicate Layer!
12:58 < genete> :)
12:59 < dooglus> how quickly?
12:59 < dooglus> do you see any differences?
12:59 < genete> I don't feel any
12:59 < dooglus> ok
12:59 < genete> only the file size!
12:59 < dooglus> hopefully it's easier to maintain
12:59 < genete> that's so much!
12:59 < genete> oh yes!
13:00 < genete> I'm managing to make them on/off properly
13:02 < dooglus> I hadn't added and of the new stuff to the list of files to translate
13:02 < dooglus> pabs3 pointed it out to me yesterday
13:02 < CIA-37> synfig: dooglus * r1275 /synfig-core/trunk/po/POTFILES.in: Sorted.
13:02 < CIA-37> synfig: dooglus * r1276 /synfig-core/trunk/po/POTFILES.in: Added 10 new files for translation.
13:02 < dooglus> I've done it now, so you should be able to translate reciprocal, integer, and duplicate
13:02 < dooglus> (nice timing by the CIA bot there)
13:02 < genete> last night I agreed with him to translate when near to the release. 
13:03 < genete> but anyway I'll keep it up to date asap
13:03 < dooglus> I need to work on the 'clone' function for duplicate layers
13:03 < dooglus> it causes synfig to hang now
13:04 < genete> what's the "clone function"?
13:04 < dooglus> it's what's used when you right-click a layer and duplicate it
13:04 < dooglus> or when you use copy & paste to copy a layer
13:04 < dooglus> try doing that on a duplicate layer with an exported index and with a layer under it
13:06 < dooglus> cloning the duplicate layer will re-use the exported "index" parameter, so the two nested duplicate layers end up using the same index
13:06 < genete> ah yes I've readed it. 
13:06 < dooglus> the inside loop will put the index back to 'from' when it's finished
13:06 < dooglus> so the outside loop never makes any progress
13:07 < genete> but when is needed a clone of a dup layer? just create a new one...
13:07 < pabs3> dooglus: sure, I'll take a look tomorrow (just got home, bit tired)
13:08 < dooglus> I would, but I've broken the windows built enough recently
13:08 < pabs3> :)
13:08  * pabs3 has too
13:09 < dooglus> it's the only language they understand
13:11 -!- ZanQdo [n=Daniel@201.201.2.22] has quit ["Adios"]
13:12 < genete> dooglus: I'm managing to set off the particles just making a switch X offset to a far place. What's the big Real I can introduce? INF gives me 2000000000.0
13:13 < genete> is that?
13:16 < dooglus> I don't remember how that works
13:16 < dooglus> I think if the real is a distance, the limit is that number of whatever units you use
13:16 < dooglus> 2000mm or 2000cm or 2000in
13:17 < dooglus> it doesn't seem like good behaviour - people using inches get to use bigger distances than people using centimetres
13:18 < genete> What only worries me is that I'm placing the particles so far away to make them dissapear. So I guess for "anybody" they would be outside the Canvas wide-height
13:19 < genete> s/anybosy/everybody
13:20 < dooglus> yes
13:20 < dooglus> it's quite a big number
13:20 < genete> It is the only way I've found to set particles on/off properly. 
13:20 < dooglus> would conditionals help?
13:21 < genete> he he
13:21 < genete> I would say yes althogh they don't help :)
13:21 < dooglus> so you can write things like (if index > radius) amount = 0
13:21 < genete> it is not so easy
13:22 < genete> the problem is that each particle is internally time shifted randomly
13:22 < genete> and when you change a global parameter it changes for all particles unless it is time shifted randomly too
13:23 < genete> but it complicates too much the convert structure. Now is so complitaced!
13:23 < genete> complicated*
13:23 < genete> anway a conditional convert is more than needed ;)
13:24 < genete> for example you can control things when moving a shape around, like widgets
13:24 < genete> lots of uses
13:26 < genete> the most cool thing from particles using Dup layer is that you can control the amount of particles during animation...
13:26 < dooglus> yes
13:26 < dooglus> right
13:26 < dooglus> I guess you could animate the 'amount' before to effectively turn some off
13:27 < genete> the amount of particles?
13:35 -!- heime [n=heime@200-126-92-194.bk6-dsl.surnet.cl] has joined #synfig
13:37 < genete> dooglus: I would like to publish the new particle template using Dup layer
13:37 < heime> hola que tal! Hay alguien aqu? que hable espa?ol?
13:37 < genete> do you think that I should maintain the old one system?
13:37 < genete> hola heime, yo por ejemplo ;)
13:38 < heime> que bien! hola
13:38  * pabs3 would just say use dup
13:38 < genete> pero normalmente es mejor hablar en ingl?s porque as? todo elmundo lo entiende y adem?s podr?s obtener mejores respuestas de los expertos
13:39 < genete> heime: en todo caso pregunta en espa?ol si te cuesta el ingles. Anyway ask in spanish if english is a trouble for you...
13:40 < heime> bien
13:41 < heime> bueno por esta vez solo queria saber que el funcionamiento de este canal, pero estoy m,uy entusiasmado en aprender synfig
13:42 < genete> te recomiendo la lectura del wiki. Esta muy completo. ?Que sistema operativo usas?
13:42 < heime> ubuntu 7.10
13:43 < genete> weno weno, lo tienes facil. Puedes conseguirte la versi?n m?s reciente simlemente siguiendo las instrucciones de compilaci?n en el wiki
13:44 < heime> si lo estoy usando el tutorial, pero es mucho mas enriquecedor hablar con personas que tambien lo usen 
13:44 < genete> ah! y no te olvides del foro!
13:44 < genete> que tal tu ingl?s?
13:46 < heime> b?sico, me ha costado leer el tutorial por lo mismo (el que esta en la pagina synfig.org) 
13:46 < heime> no se si existe otro
13:47 < genete> traducido no. Yo estoy traduciendo la interfaz a espa?ol. pero los articulos del wiki (he escrito unos cuantos) prefiero escribirlos en ingl?s.
13:48 < genete> si tienes la ultima vesi?n (svn) comprobar?s que muchas cosas ya est?n en espa?ol (suponiendo que tengas el espa?ol como lengua por defecto en tu ubuntu.
13:52 < heime> tengo la versi?n que biene con ubuntustudio en mi trabajo, y en mi casa uso la que se puede instalar desde los repositorios creo que es la 0.61.06
13:53 < heime> efectivamente est?n algunas cosas traducidas
13:53 < genete> hay una versi?n publicada m?s reciente. Mira en Downloads del wiki. 
13:54 < heime> ok
13:56 < genete> me tengo que ir. Hasta otra.
13:56 < genete> see you dooglus (and sory for speaking spanish but heime has english troubles). Bye
13:56 -!- genete [n=Genete@84.122.51.41.dyn.user.ono.com] has quit ["Abandonando"]
13:57 < heime> adios
13:57 < heime> asta luego
13:57 < heime> see you
13:57 -!- heime [n=heime@200-126-92-194.bk6-dsl.surnet.cl] has left #synfig ["Ex-Chat"]
14:02 < CIA-37> synfig: dooglus * r1277 /synfig-core/trunk/src/synfig/ (layer_duplicate.cpp layer_duplicate.h): When cloning a duplicate layer, always make a new 'Index' parameter, even if the cloned layer has it exported. Otherwise we can get an infinite loop.
14:22 -!- crazy_bus [n=philip@CPE-121-218-135-226.nsw.bigpond.net.au] has joined #synfig
14:29 -!- crazy_bus is now known as lounge_lizard
14:35 -!- lounge_lizard is now known as crazy_bus
14:41 -!- genete [n=Genete@84.122.51.41.dyn.user.ono.com] has joined #synfig
14:50 -!- Zelgadis [n=zelgadis@87.103.170.176] has joined #synfig
14:50 -!- Zelgadis [n=zelgadis@87.103.170.176] has left #synfig []
14:50 -!- Zelgadis [n=zelgadis@87.103.170.176] has joined #synfig
14:52 -!- Zelgadis [n=zelgadis@87.103.170.176] has left #synfig []
14:52 -!- Zelgadis [n=zelgadis@87.103.170.176] has joined #synfig
14:52 -!- Zelgadis [n=zelgadis@87.103.170.176] has left #synfig []
14:52 -!- Zelgadis [n=zelgadis@87.103.170.176] has joined #synfig
14:56 -!- factor [n=factor@ip68-14-160-70.ok.ok.cox.net] has quit ["Ex-Chat"]
15:00 < Zelgadis> Hi
15:00 < genete> hi Zelgadis :)
15:00 < Zelgadis> Is pabs3 here?
15:03 < Zelgadis> probably sleeping... :)
15:17 < CIA-37> synfig: dooglus * r1278 /synfig-studio/trunk/src/synfigapp/actions/valuedescdisconnect.cpp: Don't allow the Index parameter of the Duplicate layer to be disconnected.
15:20 -!- ZanQdo [n=Daniel@201.201.2.22] has joined #synfig
15:39 -!- crazy_bus [n=philip@CPE-121-218-135-226.nsw.bigpond.net.au] has quit ["sleep"]
16:45 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has joined #synfig
17:10 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has quit []
17:11 < dooglus> genete: I've made a few changes to the Dup layer, and documented them at the end of this page: http://synfig.org/Duplicate_Layer
17:11 < genete> nice!
17:11 < dooglus> genete: since there seem to be quite a few things still needing attention with the dup layer, let's use that page to make a list of them, rather than having separate sf.net bugs for each
17:12 < genete> ok, it makes sense
17:14 < dooglus> I can't see any other way the user can mess up the convert-to-duplicate for the dup layer now
17:14 < dooglus> I've stopped him doing a 'disconnect' and a 'convert' on it
17:15 < genete> I've followed your changes reading them here. 
17:15 < genete> The Duplicate convert type has been disabled then?
17:16 < dooglus> well
17:16 < dooglus> no
17:16 < dooglus> it's still needed for the duplicate layer
17:16 < dooglus> but the user should never need to select it manually
17:16 < genete> ok
17:16 < genete> fone
17:16 < genete> fine
17:16 < dooglus> it is only useful in duplicate layers, and they have it automatically
17:17 < dooglus> it's useful in any layer that needs to reference the index too, of course, but that's achieved by linking to an existing dup layer's index
17:17 < CIA-37> synfig: dooglus * r1279 /synfig-studio/trunk/src/gtkmm/instance.cpp: Don't allow the Index parameter of the Duplicate layer to be converted.
17:17 < dooglus> there shouldn't be any way to create a 'duplicate' valuenode except for by making a new duplicate layer
17:18 < dooglus> oh, I guess I need to disable 'connect' there too
17:18 < dooglus> that's another way of breaking the conversion
17:18 < dooglus> yup.  if I 'connect' one index to another, synfig hangs
17:19 < genete> but only for the index not the Front To Step ones
17:19 < genete> thsy should be normal params
17:20 < dooglus> sure
17:20 < dooglus> all these changes have only been for the Index param of the Dup layer
17:23 < genete> I'm working on a particles demo reel
17:24 < dooglus> I fixed 3 problems, and found 3 new ones: http://synfig.org/Duplicate_Layer#Known_Problems
17:37 -!- pxegeek [n=chatzill@c-71-59-140-184.hsd1.or.comcast.net] has joined #synfig
17:43 < Zelgadis> bye
17:44 < dooglus> bye
17:44 < genete> bye
17:44 -!- Zelgadis [n=zelgadis@87.103.170.176] has left #synfig []
17:46 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has joined #synfig
17:47 < CIA-37> synfig: dooglus * r1280 /synfig-studio/trunk/src/synfigapp/actions/valuedescdisconnect.cpp: Don't allow Duplicate ValueNodes in the Children dialog to be disconnected.
18:32 < CIA-37> synfig: dooglus * r1281 /synfig-studio/trunk/src/gtkmm/instance.cpp: Don't show the 'Convert' sub-menu for Duplicate ValueNodes in the Children dialog.
18:32 < CIA-37> synfig: dooglus * r1282 /synfig-studio/trunk/src/synfigapp/actions/valuedescconnect.cpp: Don't show the option of connecting to an existing Index parameter of the Duplicate layer.
18:49 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has quit []
18:57 < dooglus> genete: I'm getting the Duplicate layer to have its Index automatically exported.
18:57 < dooglus> what should its exported name be?
19:00 < dooglus> how about just "Index 1", or "Index 2" if that is already used, etc?
19:00 < genete> Index001 and so on?
19:00 < dooglus> leading zeros?
19:00 < dooglus> spaces are allowed
19:01 < genete> like the regions and outlines...
19:01 < genete> what does the region layer do when check on the Autoexport checkbox?
19:03 < dooglus> it exports a bline with the name that's in the name field of the tool options
19:03 < dooglus> that can have leading zeros and optional spaces, as you like
19:03 < dooglus> ok, I'll use "Index %d" and you can translate that how you like :)
19:04 < genete> When you create a new Dup layer what's its default name?
19:04 < dooglus> "Duplicate"
19:04 < genete> but they are unique aren't they?
19:04 < dooglus> the layer's name?  no.
19:04 < dooglus> exported value names are uniquue
19:04 < dooglus> that will be "Index 1", "Index 2", etc.
19:05 < genete> and what if it already exists
19:05 < genete> exported I mean
19:05 < dooglus> then it uses the next number
19:06 < dooglus> see what you think - it's in r1283
19:06 < genete> ok, it is fine
19:06 < genete> I'm rendering now a 2 minutes animation
19:06 < dooglus> if you make 3 duplicate layers, it will use Index 1, .. Index 3
19:06 < dooglus> if you rename "Index 2" to something else,
19:07 < dooglus> then the next dup layer you make will use "Index 2", because that's now the first available unused name
19:07 < dooglus> (for the exported Index)
19:07 < genete> so you can rename exported indexes
19:08 < dooglus> you can rename any exported valuenode, yes
19:08 < dooglus> there's nothing special about indexes really
19:08 < genete> when a dup layer find a dup layer what does it do? make several copies of the dup (and also its childs) ?
19:08 < dooglus> yes
19:09 < dooglus> it's like multiplication
19:09 < dooglus> did you see my field of flowers?
19:09 < dooglus> 6 petals per flower, 32 flowers in the scene
19:09 < dooglus> 6 * 32 petals
19:09 < genete> yes, but le me be more concise
19:10 < genete> there are two Dup layers
19:10 < genete> one is the top one
19:10 < dooglus> did you see http://synfig.org/forums/viewtopic.php?t=21 ?  a simple example of 2 dups
19:11 < dooglus> the top one will render everything under it "x" times
19:11 < dooglus> if there is another dup under it, that dup will be rendered "x" times
19:12 < genete> hey I lost that post! dang!
19:12 < dooglus> and each of those times, that 2nd dup will render what's under *it* "y" times
19:12 < genete> but must be the children encapsulated?
19:12 < dooglus> so the stuff that is under the 2nd dup will be rendered "x*y" times, since all possible combinations of the two indexes
19:12 < dooglus> no
19:12 < genete> but ...
19:13 < genete> how is it computed from bottom to top or from top to bottom?
19:13 < genete> I mean who duplicate first?
19:13 -!- KiBi [n=kibi@kibi.dyndns.org] has quit [Remote closed the connection]
19:13 < genete> the inner one?
19:15 < dooglus> rendering is done recursively from the top down
19:15 < genete> what happen if you change the z depth dinamically?
19:15 < dooglus> each layer calls the renderer recursively, saying "render everything under me" and then adds its own stuff to the result
19:15 < dooglus> genete: I think changing the z depth does what you would expect.
19:16 < dooglus> I didn't try changing it dynamically
19:16 < dooglus> but I think it will work
19:16 < genete> Duplicate haven't a Z depth param....
19:16 < dooglus> you can effectively move the dup layer down under a circle
19:16 < dooglus> aah...
19:16 < dooglus> it used to when I first made it...
19:16 < genete> yes
19:16 < dooglus> "amount", "blend method" and "z-depth" kind of come as a unit
19:17 < genete> now they are lost...
19:17 < dooglus> it would make sense to have z-depth back
19:17 < dooglus> and maybe "blend method", but I don't know
19:17 < CIA-37> synfig: dooglus * r1283 /synfig-studio/trunk/src/synfigapp/canvasinterface.cpp: Automatically export the Index parameter of new Duplicate layers.
19:17 < genete> and amount too!
19:17 < genete> and blend method too!
19:17 < dooglus> it's very easy to add them all back...  I'll do it so you can see how it works.
19:18 -!- Nahztu [n=nahztu@c-71-231-51-19.hsd1.or.comcast.net] has joined #synfig
19:19 < Nahztu> Hey
19:19 < dooglus> hi!
19:19 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has joined #synfig
19:19 < dooglus> genete: they're added, in svn r1284
19:19 < genete> hi Nahztu!
19:19 < genete> ok dooglus, cool!
19:19 < Nahztu> Hey! :D
19:21 -!- ZanQdo [n=Daniel@201.201.2.22] has quit ["Adios"]
19:21 -!- KiBi [n=kibi@kibi.dyndns.org] has joined #synfig
19:22 < Nahztu> I was wondering... I haven't used a Forum like ever, and I have posted some finished work, and some work in progress. I don't know what the rules are with quantity, or anything. I feel weird, like I'm just filling the forums, any thoughts?
19:22 < dooglus> feel free to keep posting
19:22 < dooglus> you've posed what, 5 images?
19:22 < Nahztu> Ok, that's a load off.
19:24 < dooglus> the forum is very new, and really not very busy
19:24 < Nahztu> I got 5 finished, and 1 in progress. I revised the one in progress, and I want to post that too, heh. I guess I just want to show anything "good" that I have made to anyone that is interested, I won't post anything I don't like.
19:24 < dooglus> it can surely stand a lot more traffic than it is currently getting
19:25 < dooglus> and your images aren't even really on the forum - you're just linking to them, so you won't be stressing it in the slightest
19:25 < dooglus> did you see the tutorials on the wiki?
19:25 < dooglus> they show you how to start using animation
19:26 < dooglus> it's really quite easy once you know how
19:26 < Nahztu> Ya, about that. I am using Windows Vista, and it sucks, ha! For some reason when I do try to animate things I get weird errors, and the whole program just crashes.
19:27 < genete> I only can see one way on annoy a forum. Post links to very big animated GIFs or Huge image dimensions. Also include the same at the signature. It would slow down some users bandwitdh.
19:27 < dooglus> Nahztu: there's a workaround for that, involving disabling hyperthreading for the synfig process.  let me find it.
19:28 < dooglus> http://synfig.org/FAQ#Can_I_do_anything_to_improve_the_stability_of_the_Windows_version_of_Synfig.3F
19:28 < Nahztu> I know, it's interesting, I use Synfig like it's a drawing program, but I will eventually get the hang of the animation function.
19:28 < dooglus> also, I think if you click on the time bar rather than dragging, it will be a little more stable maybe?
19:29 < Nahztu> Hey, I guess I should ask this... what's the goal of The Synfig Forums? Do forums have goals?
19:31 < genete> Nahztu: Forums are good to make long posts without interruptions. Also allows visual feedback by the insertion of images.
19:32 < dooglus> it's just somewhere to have discussions I guess
19:32 < dooglus> this place works too, but it's different - more interactive, less permanent, shorter lines
19:32 < CIA-37> synfig: dooglus * r1284 /synfig-core/trunk/src/synfig/layer_duplicate.cpp: Show the "Amount", "Blend Method" and "Z-Depth" parameters in the Duplicate layer. Are they useful?
19:32 < Nahztu> I was just curious, didn't mean to seem I was questioning the purpose... I just a really curious person.
19:32 < MangoFusion> one doesn't have to be glued to the screen for hours on end just to follow a conversation, that's the advantage of a forum ;)
19:33 < MangoFusion> similar kind of thing with mailing lists
19:33 < dooglus> another advantage is that subjects are separated out into threads, rather than being all mixed together like they are here
19:33 < MangoFusion> though you tend not to bump into the old decrepid posts
19:34 < genete> dooglus: don't you see the Amount, Blend M, and Z Depth useful?
19:34 < genete> for the Dup layer?
19:34 < dooglus> genete: I found the blend method didn't work much
19:35 < dooglus> 'behind' works - it reverses the order of the rendering, apparently
19:35 < dooglus> genete: but also, I think before you suggested that we don't need these parameters
19:35  * genete wonders if a Dup layer should not be a specialized Paste Canvas 
19:37 < dooglus> genete: I don't think so - you can use a pastecanvas to encapsulate a dup and a few layers if you like
19:37 < dooglus> to stop the dup working all the way down
19:38 < genete> but Duplicate doesn't seem to affect to the Paste Canvas parameters...
19:38 < dooglus> genete: that's true, and it's something I'll look into fixing next
19:38 < dooglus> genete: it's the last outstanding problem on my list :)
19:39 < genete> I'm sorry if I suggested to remove the Amount, Z depth etc. I don't remember... :(
19:39 < dooglus> genete: I asked what it would use for amount, and you said 'the amount from the layer it duplicates'
19:39 < dooglus> or something?
19:40 < genete> don't remember :(
19:40 < dooglus> genete: I'll find the log :)
19:42 < dooglus> genete: actually, I won't.  it doesn't matter does it :)
19:42 < dooglus> back in a bit
19:43 < genete> it doesn't matter.
20:05 -!- Nahztu [n=nahztu@c-71-231-51-19.hsd1.or.comcast.net] has quit [Read error: 113 (No route to host)]
20:06 -!- Nahztu [n=nahztu@c-71-231-51-19.hsd1.or.comcast.net] has joined #Synfig
20:07 < Nahztu> I can animate, but I have to do it perfect the first time because if I try to change the seconds on the timetrack it will crash, it says this program isn't responding, and the crash.
20:08 < dooglus> did you try the thing suggested in the FAQ?
20:08 < Nahztu> I did, it doesn't list set affinity on Vista.
20:08 < dooglus> oh
20:09 < dooglus> pxegeek might be able to help with that.  I don't know if he used vista much or not though
20:09 < dooglus> but he's the windows expert around here :)
20:10 < Nahztu> I wish I never picked it up, but it came with Laptop so I have it now... it sucks, maybe I just don't know it's full potential, whatever the problems I have had with it aren't so bad, but it's frustrating sometimes.
20:11 < dooglus> did you try using linux?  is there something in windows that you need?
20:11 < Nahztu> Linux?
20:11 < dooglus> it's a free operating system
20:11 < dooglus> you can use it instead of Windows
20:11 < Nahztu> I don't have Linux, I have always used Windows.
20:12 -!- xerakko [n=Miguel@debian/developer/xerakko] has joined #synfig
20:12 < Nahztu> I know the "alwaqys use", and "it's tradition" is dangerous in a lot of sense, but I don't know the first thing about running other operating systems... I think my dad would lynch me if he knew I was anything but Windows.
20:13 < dooglus> is your surname Gates?
20:15 < pxegeek> I just got a PC that came with vista, but I haven't tried it with Synfig yet
20:15 < pxegeek> I'll go poke and see what it looks like
20:15 < pxegeek> I know task manager is still there
20:15 < MangoFusion> dooglus: quick question, am i right in figuring that synfig-core doesn't really render using multithreading on win32?
20:15 < dooglus> I guess you don't need to put synfig on it - just check what happened to the 'affinity' thing
20:16 < Nahztu> What does the word "affinity" mean in this instance?
20:17 < dooglus> MangoFusion: I don't know whether synfig every renders multi-threaded.  studio uses a thread for the gui and another for each render, but I don't think it splits the tiles up into multiple threads
20:17 < dooglus> Nahztu: it's to do with which CPU core the job runs on - modern CPUs have multiple CPU cores
20:17 < MangoFusion> well i was just looking in render.cpp, and there is a big ifndef that just calls render if compiling for win32. weird stuff
20:17 < dooglus> MangoFusion: I've never seen the CPU go about 50% on this dual core machine, anyway.
20:18 < MangoFusion> (in render_threaded)
20:18 < dooglus> s/about/above/
20:18 < MangoFusion> maybe i am not looking in the right place, haha
20:19 < dooglus> I'm not very familiar with the threading code
20:20 < MangoFusion> appears to use fork. interesting
20:20 < pxegeek> nahztu - I think if you go into the task manager, select the process tab and find synfig
20:21 < pxegeek> right click on it and select Virtualization.  
20:21 < pxegeek> It's a check box on my kids' PC, but I don't think it's got multiple cores anyway
20:22 < genete> brb
20:22  * genete is away: Estoy ocupado
20:23 < pxegeek> Vista Home running on a Via processor
20:24 < pxegeek> Hmmm....  I have Vista whatever on the media center - I suppose I could try running Synfig on that.  
20:32 -!- xerakko [n=Miguel@debian/developer/xerakko] has quit ["Ex-Chat"]
20:34 -!- xerakko [n=Miguel@debian/developer/xerakko] has joined #synfig
20:34 < Nahztu> I'll check
20:34 < Nahztu> It's checked automatically.
20:35 < dooglus> I don't know what it does - but maybe unchecking it will help?
20:35 < Nahztu> Ah, ok
20:36 < Nahztu> It says if I uncheck it I could lose information... to what extent I wonder?
20:37 < MangoFusion> this threading code is quite weird
20:38 < Nahztu> It's not a big deal, I just have to make sure that's the timeframe I want exactly... so I gues it's a problem because I can't go back,a nd edit, heh.
20:40 < pxegeek> Yes, I think it'll be automatically checked by default to allow synfig to run on multiple threads
20:41 < pxegeek> However, Synfig isn't quite thread safe, so taking the check mark off will probably help
20:42  * pxegeek wonders if the THG task manager works under Vista
20:42 < MangoFusion> looking at this duplicated spaghetti code, i am not surprised it isn't quite thread safe :)
20:42  * pxegeek was being diplomatic
20:43 < pxegeek> It falls over faster than a one-legged man at an arse kicking contest when you try and run it on multiple threads
20:46 < Nahztu> Ok, I'll uncheck it thanks.
20:47 < MangoFusion> strange thing is i haven't found anything related to threading that doesn't use fork() - yet
20:47 < MangoFusion> maybe i need to use an IDE
20:47 < MangoFusion> hehe
20:48 < dooglus> MangoFusion: gtkmm/asyncrenderer.cpp has 			render_thread=Glib::Thread::create(
20:48 < MangoFusion> really? so its in studio?
20:48 < Nahztu> Hmm... that's what it meant by saved things may be lost... after I unchecked it Synfig disable its save function, ha!
20:49 < dooglus> Nahztu: I don't think so...  that warning was from Vista, not Synfig
20:49 < dooglus> Nahztu: didn't synfig disable 'save' because you didn't do anything yet, so there's nothing to save?
20:49 < Nahztu> I know, but coincidence, I can't save.
20:49 < dooglus> Nahztu: what version are you using?  (of synfig)
20:50 < Nahztu> No, I've done a lot, but luckily I saved before it told me I couldn't, let me try again... just thought it might have been a strange coincidence.
20:50 < dooglus> Nahztu: it sounds like an old bug - but I thought it was fixed in 0.61.07
20:50 < dooglus> Nahztu: that's why I asked what version you have.
20:51 < Nahztu> Well, I did some things and tried to save, but I think that Virtualization thing has something to do with it.
20:51 < Nahztu> I got this version not long ago.
20:52 < Nahztu> What dates the latest version?
20:52 < dooglus> from the toolbox, Help > About
20:53 < Nahztu> March 3rd 2006, too old?
20:53 < dooglus> yeah.  got a version number?
20:53 < dooglus> 0.61.?
20:54 < Nahztu> Says using SYNFIG 0.61.05.
20:54 < dooglus> oh, that's very old
20:54 < Nahztu> Weird, just got it like two, or three months ago.
20:54 < Nahztu> Whatever I guess I need newest.
20:55 < dooglus> http://synfig.org/Download#Download has the newest
20:55 < MangoFusion> things are worse than i thought
20:55 < Nahztu> Will I still have all my old works?
20:55  * MangoFusion runs
20:55 < MangoFusion> :)
20:55 < dooglus> Nahztu: yes
20:55 < Nahztu> Great!
20:55 < Nahztu> Thanks
20:55 < dooglus> MangoFusion: that bad, eh?
20:55 < MangoFusion> i was thinking it was going to be more in the core
20:56 < MangoFusion> but now i see it seems to be all over the place
20:56 < Nahztu> So, just download over old, or uninstall?
20:56 < dooglus> Nahztu: I don't know if it matters.
20:56 < Nahztu> Ok, I'll go with instinct.
20:56 < dooglus> Nahztu: did you see the newest unofficial packages?
20:56 < Nahztu> And laziness, heh.
20:56 < dooglus> "SVN 1239"
20:57 < Nahztu> Haven't seen newest anything yet.
20:57 < Nahztu> Which am I supposed to Download?
20:58 < dooglus> on http://synfig.org/Download#Download
20:58 < dooglus> see the line that says: Windows: gtk gtkmm synfig synfigstudio SVN 1239 (gtk & gtkmm required)
20:58 < Nahztu> Yes
20:58 < Nahztu> So do I download all of those?
20:59 < dooglus> I don't know if you already have the newest version of gtk and gtkmm - you might
20:59 < dooglus> but you can download and install them to make sure
20:59 < dooglus> do them in the order they are there
20:59 < Nahztu> Ok, in any certain order?
20:59 < Nahztu> Oh, ok
21:05 < MangoFusion> ah i see what is happening now.
21:05 < MangoFusion> in studio there is "AsyncRenderer" which creates a new thread to do the rendering, as opposed to doing it in the gui handling code (making the app appear to lock).
21:06 < MangoFusion> but there is also threading support in the core, which only really works on unix platforms
21:07 < MangoFusion> thus i would have to theorise that crashes on win32 are more specific to studio than core
21:07 < MangoFusion> very interesting
21:18 < Nahztu> Hahaha! "X glibmm-ERROR**: unhandled exception (type std::exception) in signal handler what: do_write_to_file() failed.  aborting..." Hurray!
21:29 < MangoFusion> if i disable the code in core which threads using fork(), rendering works much faster on mac os x (same effect as if i disabled threading in the options)
21:30 < MangoFusion> anyone know where the code for the  "Use Only a Single Thread" option is?
21:33 < MangoFusion> n/m i found it
21:35 < MangoFusion> actually the blabber with disabling the fork code in core had no effect whatsoever. forgot to press apple when changing the option
21:35 < MangoFusion> must be some deadlock
21:35 < MangoFusion> that is causing the slowdown
21:38 < dooglus> Nahztu: make sure you close synfig down before installing the new stuff
21:39 < pxegeek> nahztu - if you want to be really cutting edge, I can upload 1272 for you ;)
21:39 < pxegeek> I think 1239 is a couple of weeks old ;)
21:41 < pxegeek> Stupid Vista Ultimate.  I can't log in because "a device attached to the system is not functioning"
21:41 < MangoFusion> hahaha
21:42 < KiBi> Ul-ti-ma-te.
21:42 < pxegeek> Mango, I'd agree with your assessment - it seemed to be when using studio that windows was unstable
21:42 < pxegeek> Once I restricted its use of threads, things got a whole lot more useable
21:43 < pxegeek> I think the only bugs I see now are the ones that the Linux folks see
21:43 < MangoFusion> would be nice if the core code was re-written so it used something a bit more platform independent for threading (e.g. pthreads)
21:43 < MangoFusion> as for studio, aieeeeee!
21:43  * pxegeek adds it to dooglus' todo list
21:44 < dooglus> Ultimate as in "the last version of windows you'll ever try"?
21:44 < dooglus> pxegeek: thanks a bunch :)
21:45 < MangoFusion> the ultimate deception i think
21:45 < pxegeek> It's almost like Bill Gates wants everyone to use Linux
21:45 < MangoFusion> or a mac
21:46 < MangoFusion> or maybe just google apps
21:46 < MangoFusion> hehe
21:48 < Nahztu> It's funny that anytime I see a program is working fine Vista always sees a problem, and wants to shut everything down.
21:49 < Nahztu> It's also annoying.
21:59 < Nahztu> Is there anyway to delete Creations?
22:02 < rore> dooglus: for studio : http://rore.dyndns.org/rore/synfig/fr.po -- only 8 untranslated strings, either blank strings, or "licence key" thing, or things I really don't know how to translate at the moment. 
22:02 < dooglus> rore: thanks
22:04 < dooglus> rore: is that how you want to be credited in the AUTHORS file?
22:06 < rore> hm, yes
22:07 < dooglus> rore: and what's French for French?  does it have accents?
22:07 < pxegeek> nazhtu - delete creations?
22:07 < pxegeek> Francais
22:07 < pxegeek> with a squiggle under the c
22:09 < pxegeek> Ha- Windows Vista Ultimate - no match for a hardware reboot!
22:09  * genete is back (gone 01:47:35)
22:10 < genete> back again and logs caught...
22:12 < rore> Fran?ais - I love the "squiggle under the c" description :) 
22:14 < dooglus> I've checked it in, but I don't know if I used the right squiggle
22:14 < dooglus> gedit tells me I didn't use UTF-8
22:15 < genete> dooglus: did you know that if you are rendering a movie using synfigstudio and you create layers during the render they are shown in the render?
22:16 < dooglus> genete: I didn't, no
22:16 < dooglus> genete: I'm not surprised though
22:17 < dooglus> genete: I think the renderer renders the current animation - like the previewer does
22:17 < dooglus> if you change it while it's rendering, it will render whatever you change it to
22:17 < genete> but it is very un-useful!
22:17 < dooglus> I guess if you want to carry on editing it while it renders, use the command line renderer
22:17 < dooglus> sure...
22:18 < genete> also the ValueBase Nodes and the params are updated dinamically, what is a time waste.
22:18 < dooglus> I don't understand
22:19 < genete> just render a big animation and move the mouse over the params, they get updated for each value it is getting during the render
22:19 < genete> almost I see that in my particles thing
22:20 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has quit []
22:21 < dooglus> genete: oh, I see.
22:22 < genete> sometimes I obtain strange things just dragging the time cursor when rendering. It introduces flicks in the animation
22:22 < dooglus> genete: it seems that the renderer in studio is quite broken and useless then
22:22 < dooglus> genete: unless you leave it completely alone to do its work
22:23 < genete> eeerr, it seems not so intelligent...
22:23 < genete> maybe just launching a command line thread with the current params would be enough and let studio continue working with the file. 
22:24 < genete> So it would ask to save your work before render. (What is always a safe behavior)
22:24 < pxegeek> nahztu - I just installed synfig on vista ultimate.
22:24 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has joined #synfig
22:24 < dooglus> genete: that would need to you save the animation I guess
22:24 < dooglus> but otherwise, yes
22:25 < pxegeek> When I look in the task manager, there is a 'set affinity' option when I rightclick synfigstudio on the prcesses tab
22:25 < pxegeek> What sort of CPU do you have?  
22:25 < pxegeek> And which vista?
22:25 < Nahztu> I don't have "Ultimate", could that be the problem?
22:26 < pxegeek> Maybe.  Or if you only have a single threaded CPU, it may not show automatically
22:26 < dooglus> rore: does the commit I made look OK?  I copied the squiggle-c from a web page I found
22:30 < rore> that seems ok (althought "more" seems to have a problem with the "?", but I guess that's because my term is in UTF8 and the file is in ISO-8859)
22:30 < pxegeek> Running it with affinity set to one CPU works fine.  Checking both 'CPUs' and moving to a different time on the canvas causes an immediate crash.
22:31 < genete> http://www.darthfurby.com/genete/synfig/particle-show.avi
22:32 < genete> it is a little long (24 MB) but due to high quality
22:32 < genete> also a little boring and crappy. 
22:32 < genete> See the two black regions inserted dinamically while it was rendering.
22:42  * pxegeek looking forward to the Youtube version.
22:42 < genete> pxegeek: yes it is so long :(
22:42 < pxegeek> My VLC seems to have problems with your avi 's genete...
22:43 < genete> pxegeek: have ffmpeg installed?
22:43 < pxegeek> yes
22:43 < genete> use ffplay that comes with ffmpeg package.
22:43 < genete> it would show it
22:44 < genete> and don't play it directly into your browser.
22:44 < genete> download and then play
22:44 < genete> anyway I'll upload a reduced one.
22:48 < pxegeek> yes - that seems to be better - vlc kept complaining that the avi was broken
22:48 < pxegeek> Very impressive!
22:48 < genete> it is crappy :( very bad demo reel. I'll not publish on youtube...
22:48 < genete> I'll do a better one. smaller and smarter
22:49 < genete> anyway you can see lots of cool features there
22:49 < dooglus> my git-svn repository seems to have broken.  it won't commit any more.
22:52 < MangoFusion> hmm
22:52 < dooglus> $ git-svn fetch
22:52 < dooglus> W: Ignoring error from SVN, path probably does not exist: (175002): RA layer request failed: REPORT request failed on '/code/!svn/bc/1286': REPORT of '/code/!svn/bc/1286': 200 OK (http://svn.voria.com)
22:52 < dooglus> Path '' was probably deleted:
22:54 < MangoFusion> http://rafb.net/p/1HzcIb22.html
22:54 < MangoFusion> any ideas?
22:54 < dooglus> no
22:55 < dooglus> I struggled for quite a while to understand the random crashes slowdowns and deadlocks in studio
22:55 < dooglus> I eventually had to give up
22:55 < MangoFusion> i can't seem to find anywhere in code where cond_frame_queue_empty is actually hooked up
22:55 < MangoFusion> unless i am mis-understanding glib signals
22:56 < MangoFusion> in any case that code doesn't seem to have any point to it
22:57 < MangoFusion> oh wait i see, it is signal'd in frame_ready()
22:57 < dooglus> did you see http://www.gtkmm.org/docs/glibmm-2.4/docs/reference/html/classGlib_1_1Cond.html#cb37b09bf04ae36e30ec32c2f3d4af69 ?
22:57 -!- TMM [n=hp@ip565b35da.direct-adsl.nl] has joined #synfig
23:01 < MangoFusion> yes but it is needlessly cryptic
23:03 < MangoFusion> i can kindof see what is happening though - it is locking the mutex, then waiting for frame_ready to be called.
23:03 < MangoFusion> though anything wanting to use the mutex in between is a bit stuck
23:04 < dooglus> aah
23:05 < dooglus> I seem to have broken svn
23:05 < dooglus> with my squiggly c
23:05 < dooglus> :(
23:05 < dooglus> $ svn log -r1285
23:05 < dooglus> svn: REPORT request failed on '/code/!svn/bc/1285'
23:05 < dooglus> svn: REPORT of '/code/!svn/bc/1285': 200 OK (http://svn.voria.com)
23:05 -!- AkhIL [n=AkhIL@90.188.195.74] has quit [Read error: 104 (Connection reset by peer)]
23:06 < MangoFusion> will have to see if i can determine at which point the mutex is being re-used
23:06 < MangoFusion> aieeee
23:07 -!- AkhIL [n=AkhIL@90.188.195.74] has joined #synfig
23:20 < genete> dooglus: does it have solution?
23:22 < dooglus> genete: yes, darko just helped me fix it
23:22 < MangoFusion> i have a feeling there is some important logical error when locking this mutex. it is only locked inside the while loop, not outside it where certain class variables are checked which are usually only checked when the mutex is locked
23:25 < genete> dooglus: phew!
23:32 < CIA-37> synfig: dooglus * r1285 /synfig-studio/trunk/ (po/fr.po configure.ac AUTHORS): Added French translation by Aurore D (rore).
23:32 < CIA-37> synfig: dooglus * r1286 /synfig-studio/trunk/src/ (2 files in 2 dirs): Automatically export the Index parameter of new Duplicate layers when copy/pasting and when duplicating an existing Duplicate layer.
23:32 < CIA-37> synfig: dooglus * r1287 /synfig-studio/trunk/AUTHORS: Use UTF-8 encoding.
23:34 < Nahztu> When I go and explore programs/Synfig, I can't find the "Creations" folder?
23:35 < genete> Nahztu: you should create your own folders for "Creations".  Synfig doesn't obligate you to store your work in a particular place ;)
23:36 < MangoFusion> this is odd.... start_scanline is being called after end_frame.... does that even make sense? 0_0
23:36 < Nahztu> Oh, ok... so I can't delete anything I made though, just wondering.
23:36 < dooglus> MangoFusion: are you aware that there are generally 2 renders going on at once?
23:37 < MangoFusion> ah yes of course
23:37 < MangoFusion> silly me
23:37 < dooglus> MangoFusion: the 'navigator' dialog has a miniture copy of the canvas
23:37 < dooglus> MangoFusion: it's always rendered using the scanline renderer (the one-shot one)
23:37 < dooglus> the main canvas window is rendered as 64x64 tiles if you zoom in, using the tile renderer
23:38 < Nahztu> So, are most of the people that are members programers for Synfig?
23:38 < dooglus> mini-a-ture?
23:38 < dooglus> Nahztu: no
23:38 < genete> Nahztu: no
23:39 < Nahztu> Ah, I see. Just wondering.
23:39 < dooglus> Nahztu: if you have problems with using synfig, you can ask here for help - it'll probably be a lot quicker than using the forums
23:39 < dooglus> do you still have problems closing a bline?
23:40 < MangoFusion> hmm interesting
23:40 < MangoFusion> will have to dig into this more, it is not quite making sense at the moment
23:40 < MangoFusion> :)
23:41 < Nahztu> Oh, sorry... I don't have any real problem except for when I want to edit the Timetrack it says the program isn't responding, and it crashes. Also, just curious if there was a trash can, or delete function for creations in the "Creation" folder made for Synfig?
23:41 < genete> Nahztu: I did not explain you how to loop a BLine to not embarrase you with many explanations...
23:43 < genete> Nahztu: just delete them as normal files. Be sure not to be in edit mode in synfig because probably the file will be blocked for deletion by the Operating System.
23:43 < Nahztu> Oh, no... I'm ok. I didn't mean for my response to seem like I was trying to save face. I was just stating the problem before-hand. It takes a lot to embarrase me, heh.
23:49 < genete> dooglus: If I convert an angle to Add I cannot link a Real to the LHS or RHS subparams...
23:49 < dooglus> genete: that's the 3rd point at http://synfig.org/Duplicate_Layer#Known_Problems right?
23:49 < MangoFusion> i officially give up trying to sort this threading deadlock out for the moment
23:50 < MangoFusion> :9
23:50 < MangoFusion> :9
23:50 < MangoFusion> :(
23:50 < genete> Mmm but nothing related to Dup layer
23:50 < dooglus> genete: sure
23:50 < dooglus> genete: I've been trying to find a bug I fixed related to this a long time ago
23:50 < genete> but, can I Connect an angle to a Real? 
23:50 < genete> It seems I cannot
23:51 < genete> it works for some sort of type and for other not
23:52 < genete> for example converting an angle to a Range allows Connect its subparams (angles) to a Real
23:52 < genete> But scale desn't. It seems that there are some old convert types that doesn't allow "modern" compatible connections...
23:53 < dooglus> genete: that's right, but I'm trying to remember why 'Subtract' does it differently
23:54 < dooglus> there was something about having to make sure that the links between valuenodes and their sub-parameters were maintained properly
23:54 < dooglus> but I can't find the bug or the fix
23:54 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has quit ["ttg"]
23:55 < dooglus> genete: I could make add and subtract act like the other convert types, but maybe add and subtract are actually the only ones doing part of this quite right
23:55 < genete> "but maybe add and subtract are actually the only ones doing part of this quite right"?
23:56 < Nahztu> I was wondering if there was anyway to "un-hide" the creations folder. I try to find it in my computer, and it doesn't seem to exist outside of Synfig, no big hurry.
23:57 < dooglus> genete: http://sourceforge.net/tracker/index.php?func=detail&aid=1796068&group_id=144022&atid=757416 I think I'm thinking of
23:57 < dooglus> Nahztu: can you still open the creations inside synfig?
23:59 < Nahztu> I can open the creations folder in Synfig, but it's not in the Synfig Folder in Programs. I know creations doesn't have to be my disignated file folder, that's why I'm trying to transer works, and delete others, but as to my knowledge I can't yet.
--- Log closed Tue Jan 08 00:00:30 2008