--- Log opened Wed Mar 05 00:00:00 2008
00:00 < factor> k
00:00 -!- _Elk [n=_Elk@ti0110a340-0355.bb.online.no] has quit []
00:01 < factor> see ya
00:07 -!- Headbang [n=head@dhcp-077-251-107-023.chello.nl] has joined #synfig
00:25 -!- Headie [n=head@dhcp-077-251-107-023.chello.nl] has joined #synfig
00:31 -!- xerakko [n=Miguel@debian/developer/xerakko] has quit [Read error: 110 (Connection timed out)]
00:34 < factor> out
00:34 -!- factor [n=Factor@32.151.120.1] has quit ["mIRGGI meni puis"]
00:43 -!- Headbang [n=head@dhcp-077-251-107-023.chello.nl] has quit [Read error: 110 (Connection timed out)]
00:44 -!- enter_your_nickn [i=62ccc2ad@gateway/web/ajax/mibbit.com/x-e98e558a55d95ab0] has joined #synfig
00:45 -!- Headbang [n=head@dhcp-077-251-107-023.chello.nl] has joined #synfig
00:46 < enter_your_nickn> Hey all - trying to animate, and it seems as if lock past keyframes doesn't work - things move in them when the object is moved in future frames.
00:46 < enter_your_nickn> What could I be screwing up?
00:47 -!- enter_your_nickn [i=62ccc2ad@gateway/web/ajax/mibbit.com/x-e98e558a55d95ab0] has quit [Client Quit]
00:47 -!- enter_your_nickn [i=62ccc2ad@gateway/web/ajax/mibbit.com/x-2b30578f17b164fe] has joined #synfig
01:04 -!- Headie [n=head@dhcp-077-251-107-023.chello.nl] has quit [Read error: 110 (Connection timed out)]
01:09 -!- Headbang [n=head@dhcp-077-251-107-023.chello.nl] has quit []
01:19 < dooglus_> enter_your_nickn: there was a bug report about that recently
01:19 < dooglus_> so it's possible you're not messing anything up
01:20 < dooglus_> see http://sourceforge.net/tracker/index.php?func=detail&aid=1906375&group_id=144022&atid=757416
01:32 -!- factor [n=factor@ip68-14-160-70.ok.ok.cox.net] has joined #synfig
01:32 -!- factor [n=factor@ip68-14-160-70.ok.ok.cox.net] has quit [Remote closed the connection]
01:39 -!- factor [n=factor@ip68-14-160-70.ok.ok.cox.net] has joined #synfig
01:59 < enter_your_nickn> Thanks dooglus - it could be what I'm seeing.  The whole waypoint things is a little confusing to me.
02:02 < enter_your_nickn> Basically, a Bline I moved down to a position in frame 8s stays there at 11, 12, 14.  However, if I move the object back up at 18s, it goes up, and the object
02:02 < enter_your_nickn> stayed where it was supposed to be in 14.  However, in 8, 11, 12, the object has automagically moved slightly lower (almost like synfig is "bouncing" it).
02:03 < enter_your_nickn> This is the behavior I do not want - I had assumed that the keyframe lock would prevent any changes in locked frames.
02:03 < enter_your_nickn> Any ideas how I could keep the object from moving in 8, 11, 12?
03:01 < dooglus_> I don't know how, no.  It's a bug that will be fixed soon I guess, just as soon as we work out what should really happen.
03:01 < dooglus_> synfig puts a waypoint at the nearest keyframe each side of the time you modified, but that's not enough if you're using TCB waypoints
03:02 < dooglus_> I guess it'll end up placing more waypoints if you have the keyframe lock on
03:02 < dooglus_> I'm off to bed now - if you ask anything else, I'll try to answer tomorrow if you're still here :)
03:02 < dooglus_> 'night
03:02 -!- Netsplit card.freenode.net <-> irc.freenode.net quits: AkhIL, factor
03:02 -!- Netsplit card.freenode.net <-> irc.freenode.net quits: pabs3, timonator, CIA-35, cizra
03:04 -!- Netsplit card.freenode.net <-> irc.freenode.net quits: Dyloxin
03:09 -!- factor [n=factor@ip68-14-160-70.ok.ok.cox.net] has joined #synfig
03:09 -!- AkhIL [n=AkhIL@90.188.201.182] has joined #synfig
03:09 -!- Dyloxin [n=Dyloxin@c-71-198-192-47.hsd1.ca.comcast.net] has joined #synfig
03:10 -!- pabs3 [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has joined #synfig
03:10 -!- CIA-35 [n=CIA@208.69.182.149] has joined #synfig
03:10 -!- timonator [n=timonato@cl-2290.ham-01.de.sixxs.net] has joined #synfig
03:10 -!- cizra [n=cizra@gw.tdng.ttu.ee] has joined #synfig
03:17 < enter_your_nickn> Thanks for the help.  This program is looking pretty good - thanks to all!
03:18 -!- enter_your_nickn [i=62ccc2ad@gateway/web/ajax/mibbit.com/x-2b30578f17b164fe] has left #synfig []
03:29 -!- Netsplit card.freenode.net <-> irc.freenode.net quits: AkhIL, factor
03:30 -!- pixelgeek [n=chatzill@c-71-59-140-184.hsd1.or.comcast.net] has joined #synfig
03:32 -!- Netsplit over, joins: factor, AkhIL
03:48 -!- Netsplit card.freenode.net <-> irc.freenode.net quits: rore
03:53 -!- Netsplit over, joins: rore
04:09 < pixelgeek> http://www.linuxjournal.com/article/9951
04:24 < pabs3> thanks pixelgeek 
04:59 -!- ZanQdo [n=Daniel@201.201.2.22] has joined #synfig
05:03 < AkhIL> maya rocks... 
05:04  * AkhIL dreams blender with full functional plug-in api
05:05 < AkhIL> just like synfig has 8)
05:32 -!- Netsplit card.freenode.net <-> irc.freenode.net quits: pabs3, timonator
05:33 -!- Netsplit over, joins: pabs3, timonator
05:45 -!- Netsplit card.freenode.net <-> irc.freenode.net quits: @ChanServ
--- Log closed Wed Mar 05 05:47:23 2008
--- Log opened Wed Mar 05 05:47:28 2008
05:47 -!- dooglus [n=dooglus@rincevent.net] has joined #synfig
05:47 -!- Irssi: #synfig: Total of 13 nicks [1 ops, 0 halfops, 0 voices, 12 normal]
05:47 -!- Irssi: Join to #synfig was synced in 22 secs
05:49 -!- dooglus_ [n=dooglus@rincevent.net] has joined #synfig
05:49 -!- dooglus_ [n=dooglus@rincevent.net] has quit [Connection reset by peer]
06:14 -!- ZanQdo [n=Daniel@201.201.2.22] has quit ["Adios"]
06:56 -!- pixelgeek [n=chatzill@c-71-59-140-184.hsd1.or.comcast.net] has quit [Read error: 113 (No route to host)]
07:11 -!- Zelgadis [n=zelgadis@87.103.170.30] has joined #synfig
07:11 -!- xerakko [n=Miguel@debian/developer/xerakko] has joined #synfig
07:13 < pabs3> hi xerakko 
07:28 -!- ZanQdo [n=Daniel@201.201.2.22] has joined #synfig
08:03 -!- xerakko [n=Miguel@debian/developer/xerakko] has quit ["Ex-Chat"]
08:09 -!- ZanQdo [n=Daniel@201.201.2.22] has quit ["Adios"]
08:33 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: rore, Zelgadis
08:36 -!- Netsplit over, joins: Zelgadis, rore
08:37 < pabs3> so, who is going to post the march challenge?
08:41 < pabs3> hmm, perhaps a cartoon strip (possibly animated) about synfig or animation or something
09:39 < rore> pick a word at random in the dictionary to find the topic of the challenge :D
09:39 < rore> (no, forget it, the "random" word always ends being a horrible word like peritonitis and stuff)
09:46 < pabs3> heh
09:49 -!- Zelgadis [n=zelgadis@87.103.170.30] has quit ["Schastlivo!"]
10:05 -!- factor [n=factor@ip68-14-160-70.ok.ok.cox.net] has quit [Read error: 110 (Connection timed out)]
11:12 -!- MangoFusion [n=jamesu@host86-147-212-148.range86-147.btcentralplus.com] has joined #synfig
11:25 -!- Headie [n=os@deossekopshop.xs4all.nl] has joined #synfig
13:30 -!- Headbang [n=head@dhcp-077-251-107-023.chello.nl] has joined #synfig
13:34 -!- Zelgadis [n=dmitryev@87.103.170.14] has joined #synfig
14:05 -!- Headbang [n=head@dhcp-077-251-107-023.chello.nl] has quit [Read error: 110 (Connection timed out)]
14:05 -!- Headbang [n=head@dhcp-077-251-107-023.chello.nl] has joined #synfig
14:05 -!- Headbang [n=head@dhcp-077-251-107-023.chello.nl] has quit [Client Quit]
14:06 -!- Zelgadis [n=dmitryev@87.103.170.14] has left #synfig []
14:27 < dooglus> http://www.youtube.com/watch?v=ccYYVh9Cbjc
14:33  * pabs3 liked that one
14:53 -!- Zelgadis [n=zelgadis@87.103.170.30] has joined #synfig
15:03 < pabs3> the person who made it has a trippy website
15:03 < Zelgadis> dooglus: yesterday I not properly understand your question "what does it involve?". You asked what duties and responsibilities have mentor?
15:05 < Zelgadis> pabs3: I think we shouldn't try participate in SoC if dooglus wouldn't agree to be mentor.
15:08 < pabs3> I agree
15:46 -!- MangoFusion [n=jamesu@host86-147-212-148.range86-147.btcentralplus.com] has quit []
15:48 -!- MangoFusion [n=jamesu@host86-147-212-148.range86-147.btcentralplus.com] has joined #synfig
16:00 < Zelgadis> night all
16:00 -!- Zelgadis [n=zelgadis@87.103.170.30] has quit ["Schastlivo!"]
16:04 < dooglus> Zelgadis: yes, I meant what would I have to do.
16:55 -!- genete [i=d90c1032@gateway/web/ajax/mibbit.com/x-3838e3fe27dcb065] has joined #synfig
16:55 < genete> hi
17:04 -!- pabs3 [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has quit [Read error: 110 (Connection timed out)]
17:05 < genete> dooglus: I'm trying to understand how work the transformation layers.
17:11 -!- Yoyobuae [n=Yoyobuae@201.224.135.156] has joined #synfig
17:13 < Yoyobuae> genete: transformation layers sample pixels from the context, and paint them elsewhere, according to some transformation function
17:13 < Yoyobuae> it mostly all happens in accelerated_render() functions
17:14 < genete> hi Yoyobuae.
17:14 < Yoyobuae> hi =)
17:15 < genete> My main doubt is if ducks are transformed too and how.
17:16 < genete> I understand that a pixel is transformed by the get_color function but don't know what fucntion transforms the ducks.
17:16 < Yoyobuae> they are transformed
17:16 < genete> (I'm looking to rotate.cpp now)
17:19 < Yoyobuae> ducks store a transform_stack_ which they use to transform their points when get/set_trans_point() is called
17:23 < genete> I wonder: if a transformation layer only makes 2D operations that doesn't involve proyections (for example homogeneous transformations like scale, translate and rotate) why do they do a full pixel transformation if only transforming the ducks is enough?
17:24 < Yoyobuae> transforming ducks doesn't equal transforming parameters
17:24 < Yoyobuae> the layers never know where the ducks are, they render themselves using their parameters
17:25 < Yoyobuae> which do not change, no matter what transform is on top
17:26 -!- MangoFusion [n=jamesu@host86-147-212-148.range86-147.btcentralplus.com] has quit []
17:27 < Yoyobuae> in a sense ducks are lying to you =)
17:28 < genete> I know that it is a big change but why don't the homogeneous transformation layers only calculate the transformed parameters and ask the layer to draw itself if the real parameters were the transformed? I think it will improve render a lot
17:28 < genete> s/if/like if
17:30 < Yoyobuae> there's a diference, for example: scaling the pixels scales EVERYTHING, the shape and the bline widths
17:30 < Yoyobuae> but scaling the vertexes/tangents changes the shape, but the bline widths remain the same
17:30 < Yoyobuae> unless you propose to change the bline width too xD
17:32 < genete> you're right. some parameters are not suitable for homogenous transformations
17:33 < genete> I ask those questions because I'm figuring a way to implement bones using layers (a bone, a transformation layer)
17:33 < Yoyobuae> yeah i though so =)
17:33 < genete> but usually bones only affects to vertices and tangents not to widths
17:34 < genete> I have all the math involved solved but I haven't the way to implement it on synfig solved yet.
17:35 < genete> Another thing I wonder is why the 2D transformations are done using direct code and not templates in ETL. For example matrix calculations are done directly. it is probably faster but not expandable
17:36 < genete> For the implementation of bones I need homogenous 2D vectors in this form: (x,y,1) 
17:36 < genete> in that way scale, rotate and translate oiperations are done all using multiplicaitons with matrixes. =)
17:37 < Yoyobuae> i think that the layers shouldn't be too involved in calculating things
17:37 < Yoyobuae> for example using hardware rendering, all those transforms will be made hyperfast
17:38 < genete> look at this: 
17:38 < Yoyobuae> but the code that does it is spread all over the place, and no one would like to implement code for every single layer
17:38 < genete> http://svn.voria.com/code/synfig-core/trunk/src/modules/lyr_std/rotate.cpp
17:39 < genete> and search the Rotate::get_color function =)
17:40 < genete> I think we should start to do it as soon as possible. 
17:40 < genete> if not that infamous 2000% spped up will never come up :)
17:41 < Yoyobuae> how about this: make a class that takes care of all that rendering stuff: transforming, sampling, rendering, etc
17:41 < Yoyobuae> have layers use that class instead
17:41 < genete> sounds fine.
17:42 < Yoyobuae> have you ever seen OpenGL code
17:42 < Yoyobuae> ?
17:42 < genete> probably some render artifacts would be erased if the standard operations were unified.
17:42 < Yoyobuae> true
17:42 < genete> no, never
17:43 < Yoyobuae> oh, nvm
17:43 < Yoyobuae> it has functions that set the transformation using matrixes
17:43  * genete want to rememeber Yoyobuae that I don't program in C++ since 10 years ago... 
17:44 < Yoyobuae> once those are set you would call a glVertex() function that creates a new vertex using the set transform
17:44 < Yoyobuae> the engine takes care of doing the transformation
17:44 < Yoyobuae> then those vertexes could be used in some rendering operation, like drawing a polygon
17:44 < genete> (in those days I only have Borland C++ 1.0 and a text interface) =)
17:45 -!- MangoFusion [n=jamesu@host86-147-212-148.range86-147.btcentralplus.com] has joined #synfig
17:45 < Yoyobuae> yeah i used the text Borland too
17:45 < Yoyobuae> i learn C/C++ with it =)
17:46 < genete> I know the matrix operations theory. Where I feel lost is inside a C++ template =)
17:46 < Yoyobuae> doing graphics with graphics.h =D
17:46 < genete> yes graphics.h... oh! what days!
17:49 < genete> Yoyobuae: what does hit_check fucntion do?
17:51 -!- Yoyobuae_ [n=Yoyobuae@201.224.135.156] has joined #synfig
17:51 -!- Yoyobuae [n=Yoyobuae@201.224.135.156] has quit [Nick collision from services.]
17:51 -!- Yoyobuae_ is now known as Yoyobuae
17:51 < Yoyobuae> genete: when you click on the canvas, hit_check() finds which layer you clicked on
17:52 -!- Headie [n=os@deossekopshop.xs4all.nl] has quit ["Lost terminal"]
17:52 < genete> so you need it for transformated layers
17:54 < genete> becasue you need to check if you clicked in a transformed layer.
17:54 < genete> I see
17:54 < Yoyobuae> actually rotate layers cannot be clicked on =)
17:55 < Yoyobuae> hit_check() returns a handle to the clicked layer
17:56 < genete> rotated cannot be clicked on?
17:57 < genete> I mean transformed layer not transformer layer
17:57 < Yoyobuae> you can try click on it, but you only can select layers above or beneath it
17:58 < Yoyobuae> yeah you can select the transformed layers underneath, but which layer get select depends on the transform
17:59 < genete> Yoyobuae: so my proposal for a bone layer involve a mixture of rotate/translate/scale layer which effect decreases to 0 from the "origin" in a exponential way...
17:59 < genete> and inside a determinded region (the bone influence area) the effect is constant. 
18:00 < genete> overlapping effects from different bone layers have proportional effects
18:00 < genete> overlapping areas ^
18:00 < Yoyobuae> ok
18:01 < genete> also bone layers can affect to other bone layers (translate and rotate only becasue scale is like a translate fom origin for a bone)
18:02 -!- MangoFusion [n=jamesu@host86-147-212-148.range86-147.btcentralplus.com] has quit []
18:02 < genete> I'm in the process of write that on a document.
18:04 < Yoyobuae> so to use it, one would create a bone layer over a drawing, create the bones accordingly, then animate those
18:04 < Yoyobuae> the drawing will get transformed into the resulting pose
18:06 < genete> I hope so.
18:06 < genete> a bone layer would have only a bone inside
18:07 < genete> in fact my concept is: one bone <=> one layer
18:07 < genete> if not it would implicate lists that I don't know if they can be possible.
18:08 < genete> (also a create bone tool and so on...)
18:09 < genete> (in fact it doesn't matter if the bone layer can have one or more bones)
18:10 < Yoyobuae> the thing with having one bone for each layer is that all those layers won't have the same context
18:11 < Yoyobuae> if you put them one on top the other, it might be possible for one bone to get in the way of the others
18:11 < Yoyobuae> for example when an arm gets in front or behind the body
18:11 < genete> but how can bones (inside the same layers) be affected by other bones? they can be rotated by other bones (or not) depending on the skeleton hierarchy ... =(
18:12 < Yoyobuae> linking =D
18:12 < genete> thay should have a way to stablish parentship between bones
18:13 < genete> can you link two bones to the same parent?
18:13 < genete> it should be a parameter not a linking
18:13 < Yoyobuae> i think so
18:14 < genete> oh yes
18:14 < genete> introducing the child parameter
18:15 < genete> and linking the parent parameter to the child parameter
18:15 < genete> cool
18:16 < Yoyobuae> actually it should be possible to make a simple stick figure on synfig that simlulates bones
18:16 < genete> I was thinking on it to proof the concept
18:18 < Yoyobuae> my finger animation also kinda did that also, except the "transform" was an animation inside the child canvas =D
18:18 < genete> I'll finish writing the math theory and show you and dooglus to see if it is possible (conceptually) or not. :)
18:18 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: rore
18:18 < Yoyobuae> anything is possible =)
18:19 < genete> =)
18:19 < genete> (except the 2000% speed up) 
18:19 < genete> ;)
18:19 < Yoyobuae> it is also possible, it will just be quite some work XD
18:20 < Yoyobuae> but using the class i told you about is a good step.  That class could be switched to use different rendering strategies
18:21 < Yoyobuae> allowing easier testing
18:21 -!- Netsplit over, joins: rore
18:22 < genete> in fact it could be tested for only a selected layer and see if there is any speed up using duplicate as a increasing render time factor
18:22 < genete> for example rotate layer
18:27 -!- cizra [n=cizra@gw.tdng.ttu.ee] has quit [Read error: 104 (Connection reset by peer)]
18:27 -!- MangoFusion [n=jamesu@host86-147-212-148.range86-147.btcentralplus.com] has joined #synfig
18:28 < genete> hi rore =) what are you going to do with the price you won? ;)
18:32 < rore> buy freenode a server that won't netsplit? :] 
18:35 -!- pabs3 [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has joined #synfig
18:40 -!- pabspabspabs [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has joined #synfig
18:46 -!- genete [i=d90c1032@gateway/web/ajax/mibbit.com/x-3838e3fe27dcb065] has quit ["http://www.mibbit.com ajax IRC Client"]
18:57 -!- pabs3 [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has quit [Read error: 113 (No route to host)]
19:02 -!- pabs3 [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has joined #synfig
19:15 -!- pabspabspabs [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has quit [Read error: 110 (Connection timed out)]
19:27 -!- genete [n=Genete@84.122.32.253.dyn.user.ono.com] has joined #synfig
19:30 -!- pabspabspabs [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has joined #synfig
19:46 -!- pabs3 [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has quit [Read error: 110 (Connection timed out)]
19:46 -!- pabs3 [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has joined #synfig
19:50 -!- genete [n=Genete@84.122.32.253.dyn.user.ono.com] has quit ["Abandonando"]
19:55 -!- pabspabspabs [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has quit [Read error: 110 (Connection timed out)]
20:18 -!- _Elk [n=_Elk@ti0110a340-0355.bb.online.no] has joined #synfig
20:36 -!- MangoFusion [n=jamesu@host86-147-212-148.range86-147.btcentralplus.com] has quit []
20:45 -!- genete [n=Genete@84.122.32.253.dyn.user.ono.com] has joined #synfig
20:57 -!- MangoFusion [n=jamesu@host86-147-212-148.range86-147.btcentralplus.com] has joined #synfig
21:09 -!- factor [n=Factor@32.149.195.152] has joined #synfig
21:20 -!- Yoyobuae [n=Yoyobuae@201.224.135.156] has quit ["ChatZilla 0.9.80 [Firefox 2.0/2006101023]"]
21:25 -!- factor- [n=Factor@32.149.239.24] has joined #synfig
21:27 -!- xerakko [n=Miguel@debian/developer/xerakko] has joined #synfig
21:34 < genete> factor: are you factor or factor- or both? DNS are different...
21:40 < factor-> cell borked up
21:41 < factor-> me both
21:41 < factor-> one dead connection
21:41 < factor-> dead one would  be the other one 
21:41 < factor-> :)
21:41 < genete> dynamic DNS?
21:42 < factor-> dunno
21:42 < factor-> i am on cell phone
21:42 < genete> cool
21:42 < factor-> what are the dns s
21:43 < genete> domain name system
21:43 < factor-> cant see the host names via cell phone
21:43 < factor-> just user names
21:44 < genete> right click on the user name
21:44 < factor-> should be from cingular or at and t
21:44 < factor-> heh
21:44 < factor-> i am on cell phone
21:44 < factor-> no right clicking or even a mouse
21:44 < genete> but your chat app can have contextual menu
21:45 < genete> (I have one in my palm)
21:45 < genete> or some sort of user info
21:52 -!- factor [n=Factor@32.149.195.152] has quit [Read error: 110 (Connection timed out)]
21:53 < genete> factor-: are you in Seattle now?
21:54 < genete> near the Greenwood Memorial Park?
21:54 < factor-> listening to linux world podcast, talking about git
21:55 < factor-> was synfig going to eventualy move from svn to git
21:55 < factor-> nope
21:55 < factor-> tulsa oklahoma
21:55 < genete> a global DNS locator told me you are in Seattle ... :)
21:56 < genete> but that's your DNS server location.
21:56 < factor-> cell maybe entering the internet from there
21:58 -!- pabspabspabs [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has joined #synfig
21:59 < factor-> Lo pabs3
22:01 < genete> My God! Oklahoma streets are a checkboard!! 
22:04 < factor-> checkerboard?
22:04 < factor-> oh one mile squares
22:04 < genete> it looks like a fractal ...
22:08 < dooglus> genete: URL?
22:09 < genete> http://maps.google.es/maps?f=q&hl=es&geocode=&q=oklahoma&sll=39.118341,-94.58293&sspn=0.045948,0.090637&layer=c&ie=UTF8&ll=35.439414,-97.46933&spn=0.772011,1.450195&z=10
22:10 < genete> hi dooglus :)
22:10 < dooglus> hi
22:11 < genete> are you OK? you were so quiet yesterday ...
22:12 < dooglus> I'm OK, yes.  Have been taking a bit of a break from Synfig the last few days
22:12 < dooglus> I've had to sort some other things out - tax, holidays, etc :)
22:12 < dooglus> I'm going snowboarding the week after next
22:12 < genete> good. Internal order is important too. :)
22:13 < factor-> Heh
22:13 < genete> where?
22:13 < dooglus> slovakia
22:13 < genete> cool
22:13 < factor-> no tax holiday here ,except having to pay them
22:13 < dooglus> I put a comma between them
22:13 < dooglus> I don't think holidays from tax are possible
22:14 < dooglus> unless you also have a holiday from income and expenditure
22:14 < genete> dooglus: factor is in a mobile ;)
22:14 < genete> maybe the comma was so tiny ;)
22:14 < dooglus> genete: I read, yes.  I can IRC from mine, too, using an ssh client on it - I connect to the same IRC host as I'm using now
22:14 < factor-> Catching up on linux world podcast
22:15 < dooglus> so I'm effectively using irssi via the mobile
22:15 -!- pabs3 [n=pabs@d122-105-74-150.per9.wa.optusnet.com.au] has quit [Read error: 110 (Connection timed out)]
22:15 < factor-> heh
22:15 < genete> dooglus: regarding to the question you ask me in the bug traker for the feather...
22:15 < dooglus> yes?
22:15 < genete> I really don't know if negative is useful or not.
22:15 < genete> the fact is
22:16 < genete> that negative vlaues for some 
22:16 < genete> methods
22:16 < dooglus> you say it works for some layers - what is 'working' for negative amounts of fluffiness?
22:16 < genete> searching fluffiness
22:17 < genete> for some methods/layers it gives same render results
22:17 < genete> except the bounding box
22:17 < genete> it is smaller for negative values
22:17 < genete> so then i discovered the eyedrop problem
22:18 < genete> for fast gaussian blur it doesn't work. 
22:18 < genete> for gaussian blur it hangs on
22:18 -!- factor-- [n=Factor@32.146.9.61] has joined #synfig
22:19 < genete> and the rest of feather methods works the same
22:19 < factor--> phone rebooted
22:22 < dooglus> what's one of the rest?
22:22 < dooglus> I mean, give me a layer/method combination where it 'works'
22:24 < genete> Bline Region/Box Blur
22:27 < dooglus> the bounding box defines the area that synfig things the layer takes up
22:27 -!- factor-- is now known as factor
22:27 < dooglus> if it's too small, the tile renderer won't render tiles that the bounding box doesn't overlap
22:27 < dooglus> so zoom in on a layer with negative feather (and a too-small bounding box)
22:28 < dooglus> and you'll see it's not rendering everything it should
22:28 < dooglus> http://dooglus.rincevent.net/random/negative-fevver.png
22:29 < dooglus> (it's being cut off on the right hand edge, due to the too-small bounding box
22:29 < genete> so it renders bad also
22:29 < dooglus> other than that, box-blur feathering of -x seems to be identical to feathering of +x
22:29 < dooglus> I think feathering should be forced to be positive
22:30 < dooglus> I can't think what a negative value should mean
22:30 < genete> I agree
22:30 < dooglus> ok
22:30 < dooglus> I will make it so
22:30 < genete> but what would happen
22:31 < dooglus> only 11 source files mention "feather".  ugh.
22:31 < genete> if two waypoints makes some portion of the interpolation 
22:31 < genete> negative?
22:31 < dooglus> for each frame, set_time() gets called
22:31 < genete> I mean, are you going to avoid negative values in the valuenodes or the interpolation?
22:32 < genete> or both?
22:32 < dooglus> that uses the waypoints to work out the parameter values, and Layer::set_param() gets called for those values
22:32 < dooglus> set_param() will round x<0 to 0 for feather params
22:32 < genete> so interpolated values also are cut 
22:32 < dooglus> the curve can go negative all it likes, but when the curve's value is given to set_param, the layer will round up to 0
22:33 -!- MangoFusion [n=jamesu@host86-147-212-148.range86-147.btcentralplus.com] has quit []
22:33 < genete> is there other parameter non allowed negative by now? 
22:33 < dooglus> the valuenodes will still have negative values
22:33 < genete> I want to test
22:33 < dooglus> yes - plant layer has some I think
22:33 < genete> yup
22:34 < dooglus> yes:
22:34 < dooglus> 	IMPORT_PLUS(splits,{
22:34 < dooglus> 			needs_sync_=true;
22:34 < dooglus> 			if (splits < 1)
22:34 < dooglus> 				splits = 1;
22:34 < dooglus> 		});
22:34 < dooglus> it's an integer though
22:34 < dooglus> step is a Real, but its rounding is weird
22:35 < genete> divisions is rounded to be mimimun 1
22:35 < dooglus> 0 or less -> 0.01; else < 0.00001 -> 0.00001; else > 1 -> 1
22:35 < genete> step was dangerous because produced millions of steps when reach 0
22:35 < dooglus> yes
22:36 < dooglus> so if you put '0', you get 0.01, producing only 100 steps
22:36 < dooglus> because if you put '0', you don't know what you're doing :)
22:36 < dooglus> but if you do know what you're doing, you can ask for smaller steps
22:36 < dooglus> what divisions do you mean?
22:36 < dooglus> oh, you're talking about 1/step
22:37 < dooglus> ie. step is rounded to be maximum 1
22:37 < genete> yup
22:37 < dooglus> is that enough to experiment with?
22:38 < genete> I though there were a real param that negative values were not allowed but seems that the only one is step
22:39 < dooglus> when you hit control-s to save, does it annoy you that the filename isn't ready to type over?  it does me.
22:40 < genete> I have it ready to type over :/
22:40 < genete> gnome here :)
22:40 < genete> dooglus: what about the eyedrop bug?
22:41 < dooglus> genete: it's the same as the distorted text bug
22:42 < dooglus> genete: I have the filename selected, but not focussed
22:42 < dooglus> if I type, it searches in the list of files
22:42 < dooglus> genete: I just checked in a change to make feather non-negative for blines - try it & see what you think
22:42 < genete> distorted text bug, really?
22:42 < dooglus> if it's ok, I'll do it for the other layers too
22:43 < genete> to me is selected and focused.
22:43 < dooglus> genete: there are 2 things layers are supposed to be able to do - render a big rectangular area, and render a single pixel
22:43 < dooglus> mostly rendering a big rectangle is enough - but for distorts, and for the eyedrop, they also need to define get_color() properly
22:43 < dooglus> some layers (like the text layer) don't
22:44 < genete> ok, rebuilding now...
22:44 < dooglus> just core
22:44 -!- xerakko [n=Miguel@debian/developer/xerakko] has quit ["Ex-Chat"]
22:44 < dooglus> just synfig-core/src/studio
22:44 < dooglus> um...
22:44 < dooglus> just synfig-core/src/synfig/
22:45 < genete> I'm eight commits late so it would involve etl too...
22:45 < genete> and probably studio :(
22:45 < dooglus> my gf was saying just now she doesn't like Internet 'chat rooms' because the nonsense you talk will probably be logged forever
22:45 < dooglus> I showed her the synfig logs - back to the first one I have: http://dooglus.rincevent.net/synfig/logs/2006/%23synfig-2006-09-29.log
22:45 < dooglus> it was pretty quiet then - just a noob dooglus, talking to himself...
22:46 < dooglus> genete: aah.  but you'll get perfect tangents now :)
22:47 < genete> :)
22:47 -!- factor- [n=Factor@32.149.239.24] has quit [Read error: 110 (Connection timed out)]
22:47 < genete> interesting logs! nobody reply!!! :)
22:47 < dooglus> and I don't think _calculus.h is used much - it shouldn't be a big rebuild if you're still using the pkgconfig hacks I set up 
22:47 < dooglus> yeah - pabs3 was off travelling the world, and nobody much was using synfig I think
22:48 < CIA-35> synfig: dooglus * r1858 /synfig-core/trunk/src/synfig/layer_shape.cpp: Don't let feather be negative for any Shape layers.
22:48 < genete> I always use the pkconfig trick :)
22:50 < genete> dooglus: did you read about the crazy ideas of stick a vertex into a bline and the inverse duck manipulation?
22:50 < dooglus> genete: I had a local modification to another ETL file here which I hadn't checked in
22:50 < dooglus> every time I checked something else in, git was marking it as modified
22:50 < dooglus> so I've just checked it in to stop that
22:50 < dooglus> best stop your rebuild and re-update I think
22:51 < dooglus> genete: I did, and considered taking a look at the code today
22:51 < genete> http://synfig.org/Linking_to_Blines
22:51 < dooglus> genete: but I thought I'd leave it for now, what with the taxes and holiday to sort out, and Yoyobuae saying he'd look at it
22:51 < genete> rebuild from starting from etl?
22:52 < genete> dooglus: right, I recommended a small step each time. I think you deserve some holidays after the last release :)
22:53 < genete> _curve_func.h?
22:56 < dooglus> yes
22:57 < dooglus> just a couple of comments from when I was checking the affine_function code
22:57 < genete> ok, in the middle of core now
22:57 < genete> brb
23:02 < CIA-35> synfig: dooglus * r1859 /ETL/trunk/ETL/_curve_func.h: Added a couple of comments.
23:04 < dooglus> "If a vertex is deleted and there are other vertices linked to that vertex" - not sure about that
23:05 < genete> it was a new idea from Zelgadis
23:05 < dooglus> sometimes there is only one vertex
23:06 < dooglus> a line can end in the middle of another line
23:06 < dooglus> (making a 'T' shape)
23:06 < dooglus> and the point where the 2 lines join is a single, shared vertex
23:06 < dooglus> there's no other vertex linked to that vertex - it is a single vertex
23:07 < dooglus> like if you draw a line with the draw tool, and then start a new line from an inside vertex of the first line
23:07 < genete> but it has two ways of access it. Ine from the | ans other from the top of the T
23:08 < genete> both valuenodes are the same, yes but have two different ways of access
23:08 < dooglus> http://dooglus.rincevent.net/random/t-junction.png
23:08 < genete> but you'r right
23:09 < genete> when you delete the intersection vertex you're deleting both 
23:09 < dooglus> it can be the other way
23:09 < dooglus> if you draw two separate lines and 'link' 2 vertices
23:09 < dooglus> then deleting the intersection vertex leaves one vertex
23:09 < dooglus> doesn't it?
23:10 < genete> but that left vertex belongs to what bline? 
23:10 < genete> the first or the second bline?
23:12 < genete> I think the best way is the first proposal: select a vertex, right click a bline (not a vertex on the bline, just the bline) and a context menu would allow attach the vertex to the bline
23:12 < dooglus> "Link"
23:12 < dooglus> right
23:16 < genete> but the most interesting (and maybe difficult for you) is the inverse duck manipulation
23:16 < dooglus> it seems I'm wrong about the single vertex thing
23:16 < genete> allow move the linked duck by the mouse and it snaps to the bline
23:17 < dooglus> in both cases (create using the draw tool, or link manually) it's the same - there are 2 vertices, and they're linked
23:17 < genete> yup
23:17 < dooglus> this Link operation - it will make both "BLine Vertex" and "BLine Tangent" ValueNodes?
23:18 < dooglus> because I can see that I might not want to get the line's tangent - just its position
23:18 < genete> as per proposal from the wiki you decide what to link just selecting it before
23:18 < genete> but I see a problem
23:19 < genete> if you link the vertex first how can you link the tangent later? 
23:19 < genete> the vertex is there and you obtain a different context menu :/
23:20 < genete> and what means the tangent linked to the bline? if the vertex is not linked where is the visual feedback and manipulation?
23:21 < genete> I think you can link a vertex alone or a vertex AND a tangent at the same time
23:21 < genete> only the tangent doesn't make sense to me 
23:21 < genete> (as far as I can imagine)
23:22 < genete> whoops... studio takes ages to compile :(
23:22 < dooglus> if I'm linking the tangent to the bline, which way does it point?  there are 2 directions along the bline...
23:23 < genete> +
23:23 < genete> good observation
23:23 < dooglus> maybe I make a "Reverse Tangent" > "BLine Tangent" convert?
23:23 < genete> maybe you always link to the same direction and the user decides what tangent selects first
23:23 < dooglus> so the direction can be changed with a checkbox easily
23:24 < dooglus> the user is selecting the (single) tangent of the endpoint of a line, right?  there's only one tangent
23:24 < genete> yes right again
23:24 < genete> reverse tangent would be enough I think
23:26 < genete> it would be interesting too to link tangents perpendicular (with a constant difference angle) 
23:26 < genete> then you can decide if it is 0, 90, 180 ...
23:26 < genete> just changing the internals of the link
23:28 < dooglus> and have a duck for the angle
23:28 < dooglus> and have a kind of 'snap to grid' but for angle ducks
23:28 < genete> because maybe you want to change the relative tangent orientation of the linked one just by the modificaiton of a duck
23:28 < dooglus> so it's easy to get 90, 45, 30, etc degrees
23:30 < genete> when you manipiulate the tangent oof the bline you modify both. the one from the bline and the one from the vertex. But the can be offset by a value animatable too.
23:30 < genete> very interesting =)
23:31 < genete> oops
23:31 < dooglus> yes, rather than 'Reverse BLine' (which is effectively an 'add 0 or add 180') we have a generic 'add x' instead
23:31 < genete> there is not tangent duck on the bline
23:32 < genete> it is the result of the tangent given by the hermite spline :)
23:32 < dooglus> no, the tangent is a result of the 2 neighbouring vertices on the bline
23:32 < genete> between two vertices ;)
23:32 < dooglus> right
23:33 < genete> anyway it is cool to have the angle offset :)
23:33 < dooglus> one of the TODO entries for studio (I think) was to make valuenodes and layer parameters more the same
23:33 < dooglus> currently valuenodes don't have ducks, only layer parameters do
23:34 < genete> (studio is finished)
23:34 < dooglus> valuenodes can't be enumerations (menus)
23:34 < dooglus> etc.
23:34 < dooglus> valuenodes don't have ducks
23:34 < genete> what feather did you modify?
23:35 < genete> which^
23:41 -!- AkhIL [n=AkhIL@90.188.201.182] has quit [Read error: 104 (Connection reset by peer)]
23:41 < genete> I've tested regions with negative values and seems to work fine... 
23:42 -!- AkhIL [n=AkhIL@90.188.201.182] has joined #synfig
23:45 -!- MangoFusion [n=jamesu@host86-147-212-148.range86-147.btcentralplus.com] has joined #synfig
23:46 < dooglus> I modified "Shape" only
23:46 < genete> region you mean
23:46 < dooglus> that's polygon, region, outline, star
23:46 < genete> ok
23:46 < genete> I've tested on bline and seems to work on
23:46 < genete> you can enter negative values but it simply ignores it.
23:47 < genete> and consider it 0
23:47 < dooglus> make a new region, don't enter animate edit mode and enter a negative value
23:47 < dooglus> it'll round to 0
23:47 < dooglus> if you're in animate edit mode, it doesn't set the layer parameters directly, it sets the valuenodes
23:47 < genete> on non animated mode yes I've noticed that
23:48 < dooglus> the valuenodes can be negative
23:48 < dooglus> if you make a waypoint and then delete the waypoint, the feather is still a (constant) valuenode
23:48 < dooglus> you can tell by the thin blue line in that row - see it?
23:49 < genete> under time track from the parameter panel?
23:49 < dooglus> between the timetrack and the type in the param panel
23:49 < dooglus> it's there for valuenodes, and not there for constants
23:50 < genete> Mmm
23:50 < dooglus> the 'vector list' in a polygon, for instance, is a valuenode even when not animated
23:50 < genete> I see it for constant too
23:50 < genete> The offset of a bline is valuenode too
23:50 < genete> and it is not animated
23:51 < dooglus> http://dooglus.rincevent.net/random/thin-blue-line.png
23:52 < dooglus> the offset is a valuenode if you drew it using the bline tool and use 'link offsets'
23:52 < genete> but there is not outline here
23:52 < genete> only region
23:52 < genete> how can I link a paramter to nothing?
23:52 < dooglus> heh
23:53 < genete> http://i85.photobucket.com/albums/k74/Genete/Pantallazo-CapasFotogramasClaveDesc.png
23:54 < genete> bug?
23:56 < genete> If I animate it and then disconnec the blue line dissapears
23:59 < dooglus> disconnect means to disconnect the parameter from its valuenode
23:59 < dooglus> so its now just a value
--- Log closed Thu Mar 06 00:00:00 2008