--- Log opened Sun Jan 13 00:00:30 2008
00:03 -!- TMM [n=hp@ip565b35da.direct-adsl.nl] has joined #synfig
00:21 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has joined #synfig
00:25 < timonator> i like how synfig development seems to have accellerated humongously from when i last looked
00:39 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has quit ["Not here"]
00:57 -!- pabs3 [n=pabs@121.220.225.161] has joined #synfig
01:19 < dooglus> pxegeek: does it still happen?  if so, can I have a copy of the .sif file please?
01:27 < pabs3> timonator: it is still mostly dooglus, I've done a bit too since I'm not working full-time
01:27 < pabs3> timonator: we still need more people though
01:28 < pabs3> dooglus: I notice some 2007 logs here: http://dooglus.rincevent.net/synfig/logs/2008/
01:29 < timonator> don't look at me like that, i couldn't do anything
01:30 < timonator> you saw the incredible speed at which my interest in things drops
01:30 < pabs3> :)
01:37 < pabs3> dooglus: would it be much work to change the html code widget in the colour dialog to be editable?
01:37 < pabs3> currently changing it and pressing enter just resets the colour
01:41 < pabs3> hmmm, setting the CPU affinity on Linux seems to help synfigstudio stability here
01:47 < dooglus> pabs3: thanks.  it's a crontab that runs daily and links the logs into the year directory, along with the last week or so's logs (in case the script didn't run yesterday)
01:47 < dooglus> pabs3: I always use the html box to set color
01:47 < dooglus> pabs3: what are you typing exactly?
01:48 < pabs3> 000000, and it worked when I tried it just then, hmm
01:49 < dooglus> you have to type 1, 3 or 6 digits
01:49 < dooglus> try a single (hex) digit
01:52 < pabs3> yeah, it is working now, not sure what I did wrong
01:52 < dooglus> maybe the color was already what you wanted?  the sliders aren't in hex...
01:53 < dooglus> leading or trailing spaces stop it working
01:59 < pabs3> nah, it was orange or similar, I wanted black
02:04 < pabs3> dooglus: the TRANSLATORS comment needs to go on the line before the string to be translated
02:04 < pabs3> r1332 needs fixing
02:09 < pxegeek> I'm guessing pabs forgot to hit enter - that's what I keep forgetting to do
02:09 < pxegeek> dooglus - http://synfig.org/forums/viewtopic.php?t=28&sid=b4ee754e27b904cd58247bc564b50d4a
02:09 < pxegeek> It's Zelgadis' work in progress
02:24 < pabs3> hmm, say I have two blines that intersect, and I want to convert that intersection into a region, what should I do?
03:03 < pxegeek> Ummmm....  Create a region and link control points to teh blines?
03:04 < pxegeek> I don't think there's an easy way
03:04  * pxegeek thinks pabs is actually drawing something!
03:04 < pabs3> correct :)
03:05 < pxegeek> Something cool?
03:05 < pxegeek> Did you find a better way of creating your region?
03:08 < pabs3> I'm thinking create 2 regions, link them to the blines, unlink them, delete the unneeded ducks, then look at the sewing blines tuturial
03:10 < pabs3> unfortunately this isn't a drawing/animation I can make public, but I'm thinking about doing a second one (about the synfig community) that uses the same concept
03:11 < dooglus> pabs3: the drawing tool has a feature to do that.  it's not perfect though
03:11 < dooglus> pabs3: draw a region in the drawing tool, roughly following the outline you want, and hold the control key as you release the mouse at the end
03:15 < dooglus> pabs3: that comment is on the line before _("default")
03:19 < dooglus> pabs3: and it seems to appear in the es.po just fine:
03:19 < dooglus> #. TRANSLATORS: this is the string used in the history dialog when renaming a layer to/from its default name
03:19 < dooglus> #: src/synfigapp/actions/layersetdesc.cpp:74
03:19 < dooglus> #: src/synfigapp/actions/layersetdesc.cpp:75
03:19 < dooglus> msgid "[default]"
03:19 < pabs3> ah, so it is, sorry
03:20 < dooglus> find the region filling thing ok?
03:20 < pabs3> all the stability probs I was having the other day were fixed by setting the CPU affinity, so I updated the FAQ with how to do that on Linux
03:21 < pabs3> yep, worked a treat, thanks
03:21 < pabs3> that should be on the tips page if it isn't already
03:21 < pabs3> I like how middle-mouse drag moves the canvas around :)
03:22 < dooglus> I was working on it quite a lot, trying to get it to work better - it's really quite hit and miss
03:22 < dooglus> it doesn't add new points either I don't think - so if your blines don't already have suitable verticies for the new region to use, you're stuck
03:22 < dooglus> I didn't get anything I'm happy with working yet though
03:24 < dooglus> pabs3: the last point in http://synfig.org/How_do_I#Fill_an_outline.3F is it.  is that the page you meant?
03:25 < pxegeek> dooglus - did I miss the pen.cpp commit?
03:25 < pxegeek> what's it do?
03:25 < dooglus> pxegeek: nothing
03:25 < pxegeek> :(
03:25 < dooglus> pxegeek: removes some debugging print statements
03:25 < pabs3> dooglus: it is better than nothing, and you can always add new ducks manually 
03:26 < dooglus> pabs3: sure
03:26 < dooglus> pabs3: I think I did commit quite large changes, but the bigger changes I was working on ended up on a dead-end local git branch
03:27  * pxegeek checks to see if my build scripts are working again now we've figured out what's been happening with configure
03:29 < pxegeek> Awwwww... C****
03:30 < pxegeek> dooglus - you are correct
03:30 < pxegeek> ignore
03:30 < pxegeek> me
03:30 < dooglus> pxegeek: re. <switch>?
03:30 < pxegeek> Yes - I found a bug last night with prologue kid
03:31 < pxegeek> If you do view play, then it would get to the end and crash
03:31 < pxegeek> I was regressing back through versions, and left 849 installed
03:31 < pxegeek> hence no switch
03:31 < dooglus> pxegeek: ok
03:31 < pxegeek> Did you try prologue kid?
03:32 < dooglus> not much, no
03:32 < pxegeek> Try opening it, then View, play
03:32 < pxegeek> (It'll probably work just fine for you)
03:32 < dooglus> which file?
03:32 < dooglus> there are lots, aren't there
03:33 < dooglus> you mean the one that comes with core?
03:33 < pxegeek> The one in examples?
03:33 < pxegeek> yes
03:33 < dooglus> ok
03:33 < pxegeek> It worked fairly reliably in 849, and crashed fairly reliably in 900+
03:34  * pxegeek revs back to 1320 to look at Zelgadis' file
03:35 < dooglus> pxegeek: it plays ok here
03:36 < dooglus> pxegeek: prologue kid, that is
03:36 < pxegeek> Maybe try it under Windows next time you boot into Windows?  ;)
03:38 < dooglus> ok
03:38 < dooglus> I just tried it in my optimized (and non-debug) build, and it's ok there too
03:38 < dooglus> pxegeek: do you have a build with debugging enabled for windows?
03:39 < pxegeek> Ah - Zelgadis' file loads fine now.  Can you handle unexpected xml tags better?
03:39 < dooglus> 'cos without debugging symbols, I won't know where it's crashing
03:39 < pxegeek> Or have you already done that?
03:39 < dooglus> pxegeek: I do, now...
03:39 < pxegeek> "One Step Beyond"
03:39 < dooglus> pxegeek: the file will refuse to load now I think, with an error message on the console, but it won't crash
03:39 < pxegeek> :) :) :)
03:40 < dooglus> pxegeek: try editing the file by hand and changing switch to witch or something
03:40 < dooglus> (and /switch too)
03:40 < pxegeek> Hmmm... I don't think Windows Synfig likes View/Play
03:41 < pxegeek> It just crashed on Zelgadis' file too.
03:43 < dooglus> pxegeek: view/play is a bit of a hack I think
03:43 < dooglus> I did enough to get it working, but I don't think it's particularly nice
03:43 < pxegeek> Holy mackerel!  Zelgadis' animation is superb.
03:43 < pxegeek> I can't wait to see that one finished
03:44 < pxegeek> I love the shadows on the floor
03:59 < dooglus> pabs3: did you see http://synfig.org/forums/viewtopic.php?t=29&highlight= ?
03:59 < dooglus> would the 'die die die' changes help?  is it worth suggesting he try svn?
04:04 < dooglus> 'night all
04:06 < pabs3> I saw it, I think he should try svn, and also try linking against the system ltdl
04:06 < pabs3> posted a comment
04:06 < pabs3> gnite dooglus 
04:11 < pxegeek> 'night dooglus
04:56 -!- akagogo [n=carlos@201.230.82.1] has joined #synfig
04:58 -!- akagogo [n=carlos@201.230.82.1] has left #synfig []
04:58 -!- akagogo [n=carlos@201.230.82.1] has joined #synfig
05:16 -!- akagogo [n=carlos@201.230.82.1] has quit ["Leaving."]
05:23 -!- akagogo [n=carlos@201.230.82.1] has joined #synfig
05:27 < pabs3> hi akagogo
05:27 < akagogo> hi
05:31 < pabs3> how are your adventures with synfig going?
05:34 < akagogo> Well, i'm a little bit confused
05:34 < akagogo> i don't know where to start xD
05:37 < akagogo> i want to learn a lot about synfig :D
05:40 < pabs3> best place to start is http://synfig.org/Documentation
05:40 < akagogo> i and my friend want to make a animation for a competition
05:41 < pabs3> the getting started tutorial is a good intro http://synfig.org/Getting_Started
05:42 < pabs3> and the animation basics one for animation http://synfig.org/Animation_Basics
05:45 < akagogo> Had already read it
05:45 < akagogo> I have to get used, need time xD
05:46 < pabs3> fair enough :)
05:53 -!- akagogo [n=carlos@201.230.82.1] has left #synfig []
10:22 < dooglus> pabs3: the mandrake user with the build problem replied at http://www.synfig.org/forums/viewtopic.php?t=29
10:26 < pabs3> thanks, replied
10:46 -!- genete [n=Genete@84.122.44.154.dyn.user.ono.com] has joined #synfig
10:50 < genete> hi and bye!
10:51 -!- genete [n=Genete@84.122.44.154.dyn.user.ono.com] has quit [Client Quit]
11:42 < dooglus> pabs3: http://synfig.org/forums/search.php?search_id=newposts shows 'posts since you last logged in'
11:47 < CIA-37> synfig: dooglus * r1339 /synfig-studio/trunk/src/synfigapp/actions/valuedescconvert.cpp: Improve the entry in the History dialog made when a ValueNode is converted.
11:47 < CIA-37> synfig: dooglus * r1340 /synfig-studio/trunk/src/synfigapp/actions/valuedescset.cpp: When editing an exported valuenode, show the name it is exported as in the History dialog.
13:41 -!- Zelgadis [n=zelgadis@87.103.170.86] has joined #synfig
13:44 -!- AkhIL [n=AkhIL@90.188.192.229] has quit [Read error: 104 (Connection reset by peer)]
14:09 -!- AkhIL [n=AkhIL@90.188.192.229] has joined #synfig
15:08 < pabs3> dooglus: yeah, I use that all the time :)
15:36 < dooglus> pabs3: I just improved the hex color code parsing a little.  it won't fail on whitespace any more
15:36 < pabs3> excellent, thanks
15:47 < CIA-37> synfig: dooglus * r1341 /synfig-studio/trunk/src/synfigapp/Makefile.am: Sorted and formatted. There should be no real changes.
15:47 < CIA-37> synfig: dooglus * r1342 /synfig-core/trunk/src/synfig/color.cpp: In the hex HTML code text entry box of the color dialog, ignore everything other than hex characters. Previously even stray whitespace would cause it to fail.
15:52  * pabs3 zzz
15:53 -!- Zelgadis [n=zelgadis@87.103.170.86] has quit ["Leaving."]
16:04 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has joined #synfig
16:26 < timonator> dooglus: you are on fucking fire!
16:30 < MangoFusion> anyone got a synfig extinguisher?
17:12 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has quit []
17:15 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has joined #synfig
17:40 < dooglus> timonator: I'm making lots of easy changes that don't really affect much :)
17:40 < timonator> it's still impressive
17:40 < dooglus> timonator: most of my recent changes have been to make the History dialog more informative
17:41 < timonator> are there any "junior tasks" for not so proficient programmers or ones that have not yet dived into the codebase?
17:41 < dooglus> timonator: there are lots of actions that need to have their text improved
17:41 < dooglus> timonator: it's not exactly a highly skilled job
17:42 < dooglus> timonator: you've seen how synfig has multiple levels of 'undo' and 'redo'?  you can step through the history dialog
17:43 < dooglus> timonator: each thing you do that makes an entry in the history dialog is an "Action" - defined in its own class in synfig-studio/src/synfigapp/actions/
17:44 < dooglus> when you link two parameters together, the history dialog shows "Link"
17:44 < dooglus> just "Link"
17:44 < timonator> ah
17:44 < timonator> and you want me to make it go
17:44 < dooglus> it would be better to see "Link color of circle1 and circle2" or something
17:44 < timonator> "linked MyCanvas.Origin to MyShape.Origin"
17:44 < timonator> that seems like the right task for me
17:44 < dooglus> seeing just "Link" in the history dialog isn't very informative if you're trying to work out where to undo to
17:44 < timonator> is there anonymous commit permission?
17:45 < dooglus> timonator: there isn't, but you can do an anonymous svn checkout, make a bunch of changes, then do an "svn diff" and save the output into a file
17:45 < timonator> i can begin with link, i suppose
17:45 < dooglus> upload the file to the patch tracker and I'll be able to check it in
17:46 < timonator> this would be an excellent opportunity to try using mq!
17:46 < dooglus> what's mq?
17:46 < timonator> mercurial queues
17:46 < timonator> do you know quile?
17:46 < dooglus> I used to work with a package called mq, but it was some kind of message-queue library
17:46 < dooglus> commercial software from IBM I think
17:46 < timonator> do you know mercurial?
17:46 < dooglus> no
17:46 < timonator> it's a distributed version control system
17:46 < dooglus> sounds poisonous
17:47 < CIA-37> synfig: dooglus * r1343 /synfig-studio/trunk/src/synfigapp/value_desc.h: Add new prototype ValueDesc::get_description() for formatting a string description of what a particular ValueDesc represents.
17:47 < CIA-37> synfig: dooglus * r1344 /synfig-studio/trunk/src/synfigapp/ (Makefile.am value_desc.cpp): Add new file value_desc.cpp and add it to the Makefile.
17:47 < CIA-37> synfig: dooglus * r1345 /synfig-studio/trunk/src/synfigapp/actions/valuedescset.cpp: Moved the code that formats the ValueDesc description into value_desc.cpp, and call it there rather than having it local to this action. Other actions will want to use it as well.
17:47 < CIA-37> synfig: dooglus * r1346 /synfig-studio/trunk/src/synfigapp/actions/valuedescconvert.cpp: Be more explicit in the History dialog about what ValueNode is being converted.
17:47 < dooglus> impressive eh?
17:47 < dooglus> made 4 chances and chatted to you at the same time???
17:47 < dooglus> changes
17:48 < timonator> holyfuck
17:48 < timonator> you *are* on fire
17:48 < dooglus> I made the changes a while ago.  CIA is just slow to notice them
17:48 < timonator> ;)
17:48 < dooglus> I use 'git' locally for change control
17:48 < dooglus> do you know git?
17:48 < timonator> yes
17:49 < pxegeek> pass the extinguisher someone - we don't want dooglus burning out!
17:49 < timonator> it's similar to hg
17:49 < dooglus> I see http://hgbook.red-bean.com/hgbookch12.html - isn't that the svn site, red-bean?
17:49 < timonator> "dude, i converted the file into wave by using mv MyFile.mp3 MyFile.wav" - ahaha, funniest thing i heard today
17:50 < dooglus> I'm thinking of http://svnbook.red-bean.com/ - I guess red-bean.com have a few different books :)
17:50 < timonator> possibly
17:51 < dooglus> that's cool.  usually when I convert mp3 to wav the file gets huge.  but this 'mv' converter keeps the file small.
17:51 < dooglus> does mercurial interface with svn?
17:54 < dooglus> timonator: "link" is maybe a tricky one to start with.
17:54 < dooglus> but "disconnect" should be easier
17:54 < timonator> mercurial has a svn "importer"
17:55 < timonator> it can checkout a repository, invoking it again will just update the stuff
17:56 < timonator> probably similar to gitsvn or something
17:57 < pxegeek> I could be wrong, but I think that mv recompresses the wav file dooglus, so it ends up being the same size as the original mp3.  
17:57 < pxegeek> uh-huh.
18:00 < dooglus> pxegeek: I guess it's short for "music vocoder" or something
18:01 < dooglus> pxegeek: on Windows I always just used the "Realtime Electronic Note And Music Encoder" command
18:05 < timonator> dooglus: what action should i start with if not link?
18:06 < dooglus> timonator: how about "Disconnect"?
18:06 < timonator> oke
18:06 < dooglus> in synfig-studio/trunk/src/synfigapp/actions/valuedescdisconnect.cpp
18:06 < dooglus> valuedescdisconnect.h in the same place shows:
18:06 < dooglus> private:
18:06 < dooglus> ValueDesc value_desc; synfig::Time time;
18:07 < dooglus> ie. the disconnect action has a "ValueDesc"
18:07 < dooglus> that's what the gui uses as a way of refering to the thing that is being disconnected
18:07 < timonator> OK
18:07 < timonator> let me try figuring out the rest by myself :)
18:07 < dooglus> if you look at my recent changes to valuedescconvert.cpp you'll see what needs doing
18:08 < timonator> oke
18:08 < dooglus> I recently wrote a ValueDesc::get_description() that gives you a string from a ValueDesc, so you don't need to re-write it for each action
18:08 < timonator> oh, need to do some laundry stuff, brb
18:08 < dooglus> ok
18:08 < dooglus> I'm going to eat; I'll bbl too
18:14 < timonator> so basically to get the description into the gui i'll make a synfig::String Action::MyActionThingie::get_local_name()const function
18:15 < dooglus> yeah
18:16 < dooglus> and edit the ACTION_INIT to be ACTION_INIT_(without_local_name?) or something - like in the convert action
18:17 < timonator> oh my? in the convert action it just says ACTION_SET_LOCAL_NAME(Action::ValueDescConvert,N_("Convert"));
18:17 < timonator> that can't be right.
18:19 < dooglus> I don't think it matters
18:19 < dooglus> it's not used if you define your own get_local_name()
18:19 < dooglus> I tried deleting that line and found studio wouldn't run
18:20 < timonator> oh ok
18:20 < timonator> i thought i'd have to insert the function into it somehow
18:20 < dooglus> it builds, but when I run it, I see:
18:20 < dooglus> symbol lookup error: /home/chris/programs/synfig/git/install/lib/libsynfigapp.so.0: undefined symbol: _ZN9synfigapp6Action16ValueDescConvert12local_name__E
18:21 < timonator> i don't really understand where i can get the names of the two objects the action works with
18:21 < dooglus> what's that command called that does C++ name demangling?
18:21 < timonator> don't know :/
18:21 < dooglus> disconnect works on just one thing
18:22 < dooglus> and that thing is in the private member 'value_desc'
18:22 < timonator> how can that work?
18:22 < timonator> don't you usually disconnect one thing from another thing?
18:22 < dooglus> "disconnect" breaks the link between the thing you click on and (whatever it's linked to)
18:22 < timonator> ok, so i won't get any information on what it was linked to before disconnecting?
18:23 < dooglus> maybe a layer parameter has been converted to a valuenode - like an "add" or "composite" or whatever.  "disconnect" will convert it back to being a constant
18:23 < dooglus> ie. it wasn't really linked to anything, other than its own sub-parameters
18:23 < dooglus> I don't think that's necessary, no.  just say what the thing itself is.  that's enough for the user to recognise the line in the history dialog
18:24 < timonator> OK
18:24 < dooglus> value_desc.get_description() gives a String describing the valuenode in question
18:24  * timonator makes
18:24 < dooglus> I use it in the convert action
18:24 < dooglus> with strprintf() to join it into the surrounding text
18:25 < timonator> i gathered that much
18:25 < timonator> strprintf is a creation of the synfiggers, right? because it creates a Synfig::String
18:26 < timonator> woops, re-definition?
18:27 < dooglus> timonator: it's in the ETL, yes
18:27 < dooglus> timonator: although a Synfig::String is just another name for a std::string
18:27 < dooglus> synfig/string.h says: "typedef std::string String;"
18:27 < timonator> OK
18:28 < timonator> weird. it seems like the macro ACTION_INIT already defined the function for me, i must have missed something
18:29 < dooglus> ACTION_INIT is there - that defines its own get_local_name() which returns the constant string "Disconnect"
18:29 < timonator> aaah
18:29 < timonator> yes, i see it now
18:29 < dooglus> you need to use the longer name - that stops it defining get_local_name() and lets you define your own
18:30 < timonator> yes
18:30 < timonator> seems to compile fine now :)
18:32 < timonator> sometimes i get the feeling that make makes way too much
18:33 < dooglus> the command to demangle is:
18:33 < dooglus> $ c++filt _ZN9synfigapp6Action16ValueDescConvert12local_name__E
18:33 < dooglus> synfigapp::Action::ValueDescConvert::local_name__
18:33 < dooglus> make does make way too much
18:33 < timonator> still looks half-mangled to me, does the __ at the end come from it being private or something?
18:34 < dooglus> because 'make install' in synfig re-installs all the synfig headers, and so 'make' in studio thinks everything needs rebuilding (since it depends on the synfig headers)
18:34 < dooglus> the __ at the end is in the source
18:34 < timonator> i didn't change synfig, though
18:34 < dooglus> no, but you did a 'make install' in synfig?
18:34 < timonator> no
18:34 < dooglus> oh, ok
18:34 < timonator> would there be a reason to do that?
18:34 < dooglus> you only need to "make install" in studio/src/synfigapp
18:34 < dooglus> no
18:36 < timonator> ok, what action's next?
18:36 < timonator> grepping for ACTION_INIT( should be enough?
18:38 < dooglus> yes
18:38 < timonator> what are you currently working on?
18:38 < dooglus> all the actions are in that folder, too
18:38 < dooglus> I'm not at the moment
18:38 < dooglus> so take your pick
18:39 < dooglus> stuff that works on a single layer, with a single value_desc in its private: section are good candidates I guess
18:39 < dooglus> http://dooglus.rincevent.net/random/make-in-synfigapp.txt shows the rebuild when I change a single .cpp action file and 'make install'.  123 lines of output.  do you see more than that?
18:40 < timonator> i'll check later
18:40 < dooglus> ok
18:40 < dooglus> so long as it's fast enough to rebuild, that's the main thing
18:40 < dooglus> a full rebuild of all the actions takes a while
18:40 < timonator> i'm not :P
18:41 < dooglus> want to send me an 'svn diff' output for the first change?
18:41 < timonator> i think i made a mistake, used Synfig::String rather than synfig::String
18:41 < dooglus> the actions are all "using namespace synfig;
18:41 < dooglus> " already, so just "String" is enough
18:42 < timonator> but there's synfig::String all over the place ;)
18:42 < dooglus> right
18:42 < timonator> also: CanvasRemove doesn't have value_desc, instead it has String canvas_id, i should use that one instead, yes?
18:42 < dooglus> I tend to go with whatever is all over the place myself, rather than doing what's "right"
18:43 < dooglus> I don't even know if it's possible to remove a canvas
18:43 < timonator> uh
18:43 < timonator> well, the action exists
18:43 < timonator> i guess it's "remove canvas" or something
18:43 < dooglus> yeah, but is it hooked up to anything?
18:43 < timonator> how do i know? :/
18:43 < dooglus> if it's not on a menu somewhere, then the action won't be called.
18:44 < dooglus> if you run studio and do something common until you see a less-than-helpful action appear in the History dialog, that's a good one to fix
18:44 < timonator> OK.
18:45 < dooglus> if you grep the sources for both the class name ("CanvasRemove") and the ACTION_SET_NAME ("canvas_remove") and don't find either, it's a safe bet that the action isn't used anywhere
18:45 < timonator> all right
18:45 < dooglus> (seeing it referenced in action.cpp doesn't count either - that's just adding the action to the Book of actions, so the other code can look it up by name
18:45 < dooglus> )
18:46 < timonator> yeah
18:46 < dooglus> I don't see it referenced anywhere at all, other than in the two canvasremove.* files, and in action.cpp
18:46 < timonator> OK
18:46 < timonator> heh. it won't remove the root canvas or an inline canvas
18:47 < timonator> are there any other kind of canvas?
18:47 < timonator> say
18:48 < timonator> why isn't there a simple way to make a waypoint for a value at its curent position?
18:48 < timonator> i always have to change it first
18:48 < dooglus> I don't know much about how canvases work, but I once you export an inline canvas, doesn't that stop it being inline?
18:49 < timonator> uh, dunno
18:49 < dooglus> I think there is - let me look
18:49 < timonator> oh my, what have i started?
18:51 < dooglus> if the parameter is already animated (ie. has a waypoint at any point in time) then right-clicking it gives "add waypoint" as a menu option, which does exactly what you want
18:51 < dooglus> but if it's not yet animated, that's not an option
18:51 < dooglus> it would be useful though, I agree
18:51 < dooglus> having to move it and move it back is annoying
18:52 < timonator> it is!
18:52 < timonator> especially if dealing with regions
18:53 < dooglus> look at synfigapp/actions/waypointadd.cpp
18:53 < dooglus> the "Action::WaypointAdd::is_candidate(const ParamList &x)" function is what determines whether the action should show up in the menu or not
18:53 < timonator> i could make a nicer text for set edit mode
18:53 < dooglus> and it says:
18:53 < dooglus>         if(!ValueNode_Animated::Handle::cast_dynamic(x.find("value_node")->second.get_value_node()))
18:53 < dooglus>             return false;
18:54 < dooglus> ie. if the valuenode isn't of type "ValueNode_Animated" then don't show "add waypoint"
18:54 -!- xerakko [n=Miguel@debian/developer/xerakko] has joined #synfig
18:55 < timonator> do i also have to add get_local_name to the class definition?
18:55 < dooglus> we wouldn't want to be showing "add waypoint" if the parameter was already converted to a linkable valuenode (add, cos, atan2, etc)
18:55 < dooglus> but if it is a constant, why not allow it to have a waypoint added?
18:55 < dooglus> timonator: no; it's already there (in Super I think?)
18:55 < timonator> hmm
18:55 < timonator> ../synfigapp/.libs/libsynfigapp.so: undefined reference to `synfigapp::ValueDesc::get_description() const'
18:56 < dooglus> no, in Base
18:56 < timonator> i just realised the error is completely unrelated :/
18:57 < dooglus> I'm wrong again.  it turns out that the "ACTION_MODULE_EXT" macro you see in the header file is what declares it
18:58 -!- MangoFusion [n=jamesu@host217-43-243-35.range217-43.btcentralplus.com] has quit [Read error: 110 (Connection timed out)]
19:01 < timonator> i don't understand why it won't build
19:02 < dooglus> you didn't rerun the autobuild magic
19:02 < dooglus> so it's not linking in value_desc.o
19:02 < dooglus> because value_desc.cpp is new
19:02 < timonator> autobuild magic!
19:02 < dooglus> the autoreconf or whatever
19:02 < timonator> yes
19:02 < dooglus> whatever turns Makefile.am -> Makefile.in -> Makefile
19:03 < dooglus> Makefile.am mentions value_desc.cpp - do the others?
19:03 < timonator> i'll just run autorecong
19:03 < timonator> autoreconf*
19:03 < dooglus> chris@trpaslik:~/programs/synfig/git/synfigapp$ grep -l value_desc.cpp Makefile*
19:03 < dooglus> Makefile
19:03 < dooglus> Makefile.am
19:03 < dooglus> Makefile.in
19:03 < dooglus> all 3 should mention it, you see
19:04 < dooglus> autoreconf will get the Makefile.am up to date, but the Makefile itself won't be re-written until you run 'configure' again I think
19:04 < dooglus> s/.am/.in/
19:05  * timonator remakes
19:05 < timonator> why can keyframes have a length?
19:05 < dooglus> I've hardly used them
19:05 < dooglus> but isn't the length readonly?
19:06 < dooglus> it just tells you the distance between keyframes
19:06 < timonator> oh, ok
19:06 < dooglus> I'm guessing
19:08 < timonator> oh, the rotate tool still doesn't have a way to specify the center of rotation
19:09 -!- omry [n=omry@bzq-84-108-20-56.cablep.bezeqint.net] has quit [Remote closed the connection]
19:09 < dooglus> the centre is the duck you click on?
19:10 < timonator> no
19:10 < dooglus> no
19:10 < timonator> the center is the center of all the ducks marked when rotating
19:10 < dooglus> the center is the center of mass?
19:10 < timonator> if all ducks weigh the same, yes
19:13 < timonator> ok, um
19:13 < timonator> why can't i duplicate my keyframe? :/
19:14 < timonator> aaah
19:14 < timonator> i have to move to the destination in time first
19:18 < timonator> aaw, first synfig makes me happy that i can select multiple waypoints at a time
19:19 < timonator> but then it lets me down by letting me only change the interpolation style of the last one
19:19 < dooglus> and then makes you happy again by being open source and allowing you to fix it - yay :)
19:20 < timonator> i wish, hah
19:22 < pxegeek> I just wish it was a little more graceful when you realized that you thought you were rendering a single frame, only to realize that you've done 600+ identical frames....
19:22 < pxegeek> and no way to interrupt them
19:23 < dooglus> you're enjoying the change that makes 'render current frame' OFF by default?
19:23 < dooglus> good
19:23 < timonator> ok, uh...
19:23 < timonator> changing the types of 70 waypoints by hand
19:23 < timonator> ......
19:23 < timonator> woah
19:24 < timonator> for some nodes there is a "change out interpolation to" submenu
19:24 < timonator> but only for those that have nodes inside themselves that also posess waypoints
19:29 < timonator> can i somehow un-animate a complete node that has lots of subnodes?
19:29 < timonator> oh
19:29 < timonator> disconnect wasn't the right thing
19:41 < dooglus> disconnect will make it back into a constant valuenode
19:41 < dooglus> not animated
19:41 < dooglus> but also not converted
19:42 < dooglus> you can drag the top-level waypoint to a different time, and all the sub-parameters' waypoints will move with it
19:43 < timonator> but it won't delete them?
19:43 -!- MangoFusion [n=jamesu@host81-156-80-146.range81-156.btcentralplus.com] has joined #synfig
19:51 -!- factor [n=factor@ip68-14-160-70.ok.ok.cox.net] has joined #synfig
19:52 < pxegeek> Hi Factor
19:55 < factor> lo 
20:06 < dooglus> timonator: genete mentioned a trick - if you delete a keyframe, it deletes all the waypoints at that time
20:06 < dooglus> so you can make a keyframe at a time you don't use, drag the parent waypoint there, then delete the keyframe
20:31 < timonator> woah, that's awesome
20:31 < timonator> it's a damn hack, but it sure does the trick
20:32 < timonator> you're my hero!
20:32 < dooglus> I'm just passing on genete's discovery - he's the real synfig explorer :)
20:33 < factor> yes he is big time
20:33 < factor> he does some really cool stuff
20:33 < factor> and synfig is the better for it
20:34 < pxegeek> The demoreel certainly is...
20:34 < dooglus> most of the stuff I've done on synfig has been his idea, too
21:12 < dooglus> timonator: did you finish changes for synfig?  I don't want to repeat what you've done
21:12 < timonator> no, didn't do anything ever since
21:12 < timonator> food was involved
21:23 < dooglus> will you be doing more?
21:24 < timonator> not today
21:24 < timonator> i'm to hold a presentation tomorrow
21:24 < timonator> on product piracy in china
21:24 < timonator> and stuff
21:30 < pxegeek> I understand they're quite good at it
21:34 < dooglus> piracy?  isn't that where you steal things?
21:34 < dooglus> I thought the chinese mostly just made copies of things
21:36 < pxegeek> no, no, that's exactly what piracy is - making copies of stuff - especially if you want to listen to it in the privacy of your own home.  
21:37 < pxegeek> Don't you listen to anything the RIAA told you in school?
21:39 < timonator> yeah, right!
21:39 < dooglus> I went to school a while ago, back when the schools used to educate, rather than indoctrinate
21:40 < pxegeek> Ah - back in the days before intelligent design....
21:42 < pxegeek> when dinosaurs still roamed the earth
21:55 < KiBi> GrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrRRRrrrrrRRRrrrrrRRrrrrRRRRrrr
21:57 < factor> brb
21:57 -!- factor [n=factor@ip68-14-160-70.ok.ok.cox.net] has quit ["Ex-Chat"]
22:11 < timonator> oh, actually i have code to implement a meaningful message for ValueDescDisconnect
22:33 < dooglus> if you can upload the 'svn diff' somewhere I'll apply it
22:37 < timonator> all right
22:37 < timonator> do you like netcat?
22:37 < dooglus> pxegeek: they did teach the creation story, but it was in the stories class, not in the science class
22:39 < dooglus> timonator: that's fine.   89.176.28.156 55505
22:39 < timonator> all right
22:39 < timonator> should be done now
22:39 < dooglus> thanks
22:39 < timonator> no problem
22:40 < dooglus> how shall I credit you?  want a 'real name' used?  or your IRC nick?
22:40 < dooglus> for the commit log
22:41 < timonator> Timo Paulssen is fine, but credit is kind of overkill for what i did, heh
22:41 < timonator> i guess this makes me have to continue work
22:41 < timonator> ing
22:59 < dooglus> heh, if you like :)
23:04  * pxegeek chuckles over 'stories' class
23:11 < dooglus> 9 actions done, 56 still to do...
23:15 -!- genete [n=Genete@84.122.44.154.dyn.user.ono.com] has joined #synfig
23:15 < genete> hi
23:17 < CIA-37> synfig: dooglus * r1347 /synfig-studio/trunk/src/synfigapp/actions/valuedescdisconnect.cpp: Improve the 'Disconnect' message in the History dialog. Patch from Timo Paulssen (timonator on IRC).
23:18 < timonator> yeah!
23:18 < timonator> i feel so honored and awesome
23:21 < dooglus> :)
23:21 < dooglus> hi genete
23:21 < genete> hi dooglus 
23:22 < dooglus> did you meet the newest synfig developer?
23:22 < dooglus> :)
23:22 < genete> yeah!
23:22 < genete> I'm catching the logs
23:23  * timonator giggles
23:24 < timonator> oh, in "svn up" C means "merged", not "conflicting"
23:26 < rore> timonator: G is for merged, C is for conflict
23:26 -!- xerakko [n=Miguel@debian/developer/xerakko] has quit ["Ex-Chat"]
23:27 < rore> timonator: grep '<<<' your/file
23:27 < rore> check "svn status" as well
23:30 -!- jdc18 [n=jdc@200.107.46.12] has joined #synfig
23:30 < jdc18> hi
23:31 < genete> hi jdc18 :)
23:31 < jdc18> i am having trouble installing synfig-core
23:31 < jdc18> i installed etl
23:32 < jdc18> but when run ./configure i get this error checking for ETL >= 0.04.10... configure: error:  ** You need to install the ETL (version 0.04.10 or greater).
23:33 < jdc18> ./configure --includedir=/usr/local/include/ 
23:33 < jdc18> the same error
23:34 < genete> jdc18: are you installing form the latest release? What OD are you using?
23:34 < genete> s/OD/OS
23:35 < jdc18> gentoo
23:35 < timonator> there are no conflict markers in there
23:35 < timonator> and there is no ".mine", ".theirs" or anything
23:35 < timonator> nor any files wich have a revision number in their filename
23:35 < jdc18> i tried downloading the source, then i tried via svn
23:35 < timonator> oh, actually
23:35 < timonator> there are .mine and two revision number files
23:36 < timonator> how could i overlook the conflict markers?!
23:36 < genete> jdc18: did you succesfully installed something from the last release source?
23:37 < genete> timonator: you're reinstalling without removing your own modified files. That would produce a merged files there with some <<<<<.mine and other similar stuff not readable by the compiler.
23:38 < timonator> yeah
23:38 < genete> timonator: remove the files involved and rebuild.
23:38 < timonator> i sent my patch to dooglus, he applied it and commited, then i updated
23:38 < timonator> why did that make a conflict?
23:38 < genete> dooglus: wow lots of improvements with the history dialog!
23:38 < jdc18> genete: via svn is the last source isn it?
23:39 < genete> and timonator: congratulations for your commit!!! :) 
23:39 < timonator> thanks :)
23:40 < genete> jdc18: yes it is. For gentoo follow those build instructions: http://synfig.org/Build_instructions
23:40 < genete> and be sure to have all the requested develompent libraries installed.
23:41 < jdc18> genete: i did try that
23:41 < jdc18> i downloaded via svn, etl
23:41 < jdc18> i tried the same with  synfig
23:42 < jdc18> when i do configure i get the same error
23:43 < genete> jdc18: it is weird that it complains about ETL if it is already installed. Maybe you did not install in the proper place?
23:45 < genete> have you tried to "export PKG_CONFIG_PATH?urs/local/lib/pkgconfig"
23:45 < genete> before building or installing?
23:46 < timonator> yes, you need that export in gentoo
23:47 < CIA-37> synfig: dooglus * r1348 /synfig-studio/trunk/src/synfigapp/ (value_desc.cpp value_desc.h): Add an optional parameter to ValueDesc::get_description() to suppress reporting the exported name.
23:47 < CIA-37> synfig: dooglus * r1349 /synfig-studio/trunk/src/synfigapp/actions/valuedescexport.cpp: Import the History dialog's display when ValueNodes are exported.
23:47 < dooglus> timonator: I edited the patch a little
23:48 < genete> dooglus: one beer? ;) you'll be dry...
23:48 < dooglus> timonator: you wrote "when a ValueNodes is disconnected" - I removed the 's'
23:48 < timonator> woops, right
23:49 < timonator> you know, i tried having something like "Disconnected a from b"
23:49 < dooglus> timonator: I also incorrectly corrected how you wrote 'history' with a little h, but then noticed you had just copy/pasted from the other actions.  oops :)
23:49 < timonator> hehehe :)
23:49 < dooglus> timonator: oh, and I used the current tense for the action: "Disconnect", not "Disconnected"
23:49 < dooglus> all the actions are current tense I think
23:49 < timonator> i didn't actually copypaste, i copytyped
23:49 < dooglus> so yeah, "conflicts" :)
23:49 < timonator> i see then
23:50 < dooglus> "svn revert" says "get from svn; throw my local copy away"
23:50 < dooglus> jdc18: I suspect that pkgconfig isn't looking in the right place
23:51 < dooglus> jdc18: installing ETL will write a new ETC.pc somewhere.  you need pkgconfig to find it.
23:51 < dooglus> jdc18: see in the build instructions where it says; "Some distros have a pkg-config that doesn't look in /usr/local/lib/pkgconfig by default. So if you are installing in anywhere other than the system pkg-config path, please run "export PKG_CONFIG_PATH=/usr/local/lib/pkgconfig" or similar before building or installing anything. "
23:52 < genete> dooglus: late
23:52 < dooglus> genete: he fixed it already?
23:52 < genete> I told it a few lines ago... when you were commiting :)
23:52 < dooglus> ok
23:52 < dooglus> oh yeah
23:54 < genete> dooglus: and how does work the last Dup improvement?
23:54 < genete> I'm out of date
23:54 < dooglus> genete: parse error, line 1
23:54 < dooglus> genete: what are you asking?  about the Index interfering between navigator and workarea?
23:54 < genete> no
23:55 < genete> about the render the first time as composite and the rest as its blend mode
23:55 < genete> any test to show?
23:55 < dooglus> genete: I tried for hours yesterday to reproduce the bug
23:55 < dooglus> I couldn't at all
23:55 < dooglus> genete: I tried everything
23:56 < dooglus> I just loaded up my test file now, and straight away, it did it!
23:56 < genete> which bug?
23:56 < dooglus> genete: the Index interfering
23:56 < genete> ah
23:57 < genete> so now it is repeteable?
23:57 < dooglus> genete: the 'composite' for 1st layer didn't really help.  I checked it in, but it's not great
23:57 < genete> oooh
23:57 < dooglus> genete: I don't know.  I saw it once out of 1 today so far...
23:58 < genete> ok I'll test it tomorrow... when you stop a little of making commits... 
23:58 < dooglus> genete: screenshot: http://dooglus.rincevent.net/random/duptest.png
23:58 < genete> I don't want to make full rebuilds three times per day!!!
23:58 < dooglus> genete: sifz file: http://dooglus.rincevent.net/random/duptest.sifz
23:58 < dooglus> genete: you don't have to build after every commit :)
23:59 < dooglus> genete: you can build once a day if you like, or whatever
23:59 < genete> I know but your changes are global...
23:59 < dooglus> most of my recent changes have been very local
23:59 < genete> nice flowers!
23:59 < dooglus> 4 layers of dup
23:59 < dooglus> 1: petals; 2: layers of petals; 3: x duplication; 4: y duplication
--- Log closed Mon Jan 14 00:00:24 2008