## inkscape-devel

 [Inkscape-devel] [inkscpae-devel] Applying for GSoc 2008 - 2geom project From: Marco - 2008-03-21 14:24:29 ```Hi, I'm new to the mailing list. I'm interested in contributing in the development and integration of 2geom as a student for GSoC. At present I dived into the most of 2geom codebase. I checked out the library from here: https://lib2geom.svn.sourceforge.net/svnroot/lib2geom/lib2geom/trunk I noticed that the library is already mostly implemented, at a first glance it seems that there are more functionalities there that in libnr. I've always to find out what is the integration level. The idea description about 2geom is enough syntetic, I would like to get some more info and hints so that I'll be able to write down a sensible and detailed proposal/plan for gsoc the next week. In the case it isn't possible to apply for 2geom I could be interested in Live Path Effects too but I've aways to look at the codebase. Follows some rumination on 2geom library. I was wondering if the goal of 2geom is to become an independent library or the developed features have always to be thought with inkscape in mind. I got that path crossings are implemented by a uniform partioning of the interval where the parameter takes its value in small subintervals (2^6 for each path) these generate subpaths whose intersection points are computed as if they were line segment. I'm wondering if it was taken into account that a uniform partitioning of the parameter interval (usually [0,1] for single curve) doesn't mean a uniform partitioning of the path, that is, it could happen that [0, 1/2] is mapped to 9/10 of the path and [1/2, 1] to only 1/10 of the path. To summarize the issue is: making the parameter subintervals 'small enough' doesn't mean to get subpaths enough short that it is possible to manipulate them as they were straight line segments. What's the exact role of s-power basis ? Maybe to have a common form for all the types of curve supported by the svg specification (bezier, elliptical arc, segment) ? Are there specific computational advantages using s-power basis ? I would like to read Sánchez-Reyes's articles (as Applications of the s-Power Basis in Geometry Processing) but I haven't access to them. Follows a brief description of myself and of my previous programming experiences. I'm a student in mathematics (University of Pisa, Italy). I'm interested in computational geometry so I think that applying to gsoc for inkscape could be a good start point. I'm developing an extension of Boost.Function (http://www.boost.org/doc/html/function.html) in order to implement a model of function concept supporting multi-signature and so overloading. You can check out the source code from here: https://svn.boost.org/svn/boost/sandbox/overload/trunk or browsing the documentation: http://docs.google.com/View?docid=dfxvjncs_1x6456m I contributed in the development of a proof of concept svg import filter for Open Office. This import filter has been developed mainly by Fridrich Strba at Novell, my contribute was to add svg text element support to the import filter. If you want gives a glance at the code: http://svn.gnome.org/viewvc/ooo-build?view=revision&revision=9662 Sorry for my poor english. Kind Regards, Marco Cecchetti -- Using Opera's revolutionary e-mail client: http://www.opera.com/mail/ ```
 Re: [Inkscape-devel] [inkscpae-devel] Applying for GSoc 2008 - 2geom project From: Jasper van de Gronde - 2008-03-21 14:45:57 ```Marco wrote: > Hi, I'm new to the mailing list. > I'm interested in contributing in the development and integration > of 2geom as a student for GSoC. > At present I dived into the most of 2geom codebase. > I checked out the library from here: > https://lib2geom.svn.sourceforge.net/svnroot/lib2geom/lib2geom/trunk > I noticed that the library is already mostly implemented, > at a first glance it seems that there are more functionalities there > that in libnr. ... Well, libnr also has some non-geometry related functionality (see nr-blit, nr-compose*, nr-gradient and nr-pix*). Obviously 2geom is not a replacement for that functionality. BTW, I recently put a patch in the bug tracker (https://bugs.launchpad.net/inkscape/+bug/203318) which eliminates two old NR structures in favour of the newer NR classes (NR::Point and NR::Matrix). As a start you might consider creating a follow-up patch to eliminate these NR classes in favour of the 2geom versions (they are very similar so it should be relatively easy, especially the Point class). ```
 Re: [Inkscape-devel] [inkscpae-devel] Applying for GSoc 2008 - 2geomproject From: - 2008-03-22 19:39:10 ```> -----Original Message----- > From: inkscape-devel-bounces@... > [mailto:inkscape-devel-bounces@...] On > Behalf Of Marco > > Hi, I'm new to the mailing list. > I'm interested in contributing in the development and > integration of 2geom as a student for GSoC. > At present I dived into the most of 2geom codebase. > I checked out the library from here: > https://lib2geom.svn.sourceforge.net/svnroot/lib2geom/lib2geom/trunk > I noticed that the library is already mostly implemented, at > a first glance it seems that there are more functionalities > there that in libnr. I've always to find out what is the > integration level. > The idea description about 2geom is enough syntetic, I would > like to get some more info and hints so that I'll be able to > write down a sensible and detailed proposal/plan for gsoc the > next week. > In the case it isn't possible to apply for 2geom I could be > interested in Live Path Effects too but I've aways to look at > the codebase. I actually would very much like to work on integration of 2geom, since I've introduced 2geom into Inkscape during GSoC last year. With your apparent good math knowledge, perhaps I can interest you in doing mesh distortion LPE ? The math is not so easy, and to make it computationally speedy perhaps 2geom needs some changes. Or perhaps you have your own LPE ideas. I think developing LPEs might have you working a lot on 2geom too. It does require quite some Inkscape coding as well though to create on-canvas editing of the mesh. Regards, Johan ```
 Re: [Inkscape-devel] [inkscpae-devel] Applying for GSoc 2008 - 2geomproject From: Marco - 2008-03-24 18:27:27 ```On Sat, 22 Mar 2008 20:39:32 +0100, wrote: >> -----Original Message----- >> From: inkscape-devel-bounces@... >> [mailto:inkscape-devel-bounces@...] On >> Behalf Of Marco >> >> Hi, I'm new to the mailing list. >> I'm interested in contributing in the development and >> integration of 2geom as a student for GSoC. >> At present I dived into the most of 2geom codebase. >> I checked out the library from here: >> https://lib2geom.svn.sourceforge.net/svnroot/lib2geom/lib2geom/trunk >> I noticed that the library is already mostly implemented, at >> a first glance it seems that there are more functionalities >> there that in libnr. I've always to find out what is the >> integration level. >> The idea description about 2geom is enough syntetic, I would >> like to get some more info and hints so that I'll be able to >> write down a sensible and detailed proposal/plan for gsoc the >> next week. >> In the case it isn't possible to apply for 2geom I could be >> interested in Live Path Effects too but I've aways to look at >> the codebase. > > I actually would very much like to work on integration of 2geom, since > I've introduced 2geom into Inkscape during GSoC last year. > With your apparent good math knowledge, perhaps I can interest you in > doing mesh distortion LPE ? The math is not so easy, and to make it > computationally speedy perhaps 2geom needs some changes. Or perhaps you > have your own LPE ideas. I think developing LPEs might have you working > a lot on 2geom too. > It does require quite some Inkscape coding as well though to create > on-canvas editing of the mesh. > > Regards, > Johan Well, my goal is to work on 2geom or lpe (that btw exploits 2geom a lot), personally I'll prefer to implement some algorithm or write test programs, rather than doing integration, on the other side I'm a newbye and so an integration task could be a soft way to learn 2geom library. Will it be ok if I spend a first part on an integration task and a second part implementing some specific needed algorithm for 2geom or lpe (and here suggestions are welcome) ? Btw, I'm working on elliptical arc support in 2geom, just to take more confidence with the library. I think to be succeeded in implement a to_sbasis routine but I have always to integrate large_arc and sweep flags support. Could you point me to some paper on mesh distortion, so I can evaluate the required effort to implement such a feature. Regards, Marco -- Using Opera's revolutionary e-mail client: http://www.opera.com/mail/ ```
 Re: [Inkscape-devel] [inkscpae-devel] Applying for GSoc 2008 - 2geomproject From: - 2008-03-24 18:32:57 ```> -----Original Message----- > From: Marco [mailto:mrcekets@...] > Sent: maandag 24 maart 2008 19:27 > To: Engelen, J.B.C. (Johan) > Cc: inkscape-devel@... > Subject: Re: [Inkscape-devel] [inkscpae-devel] Applying for > GSoc 2008 - 2geomproject > > On Sat, 22 Mar 2008 20:39:32 +0100, > wrote: > > >> -----Original Message----- > >> From: inkscape-devel-bounces@... > >> [mailto:inkscape-devel-bounces@...] On Behalf Of > >> Marco > >> > >> Hi, I'm new to the mailing list. > >> I'm interested in contributing in the development and > integration of > >> 2geom as a student for GSoC. > >> At present I dived into the most of 2geom codebase. > >> I checked out the library from here: > >> > https://lib2geom.svn.sourceforge.net/svnroot/lib2geom/lib2geom/trunk > >> I noticed that the library is already mostly implemented, > at a first > >> glance it seems that there are more functionalities there that in > >> libnr. I've always to find out what is the integration level. > >> The idea description about 2geom is enough syntetic, I > would like to > >> get some more info and hints so that I'll be able to write down a > >> sensible and detailed proposal/plan for gsoc the next week. > >> In the case it isn't possible to apply for 2geom I could be > >> interested in Live Path Effects too but I've aways to look at the > >> codebase. > > > > I actually would very much like to work on integration of > 2geom, since > > I've introduced 2geom into Inkscape during GSoC last year. > > With your apparent good math knowledge, perhaps I can > interest you in > > doing mesh distortion LPE ? The math is not so easy, and to make it > > computationally speedy perhaps 2geom needs some changes. Or perhaps > > you have your own LPE ideas. I think developing LPEs might have you > > working a lot on 2geom too. > > It does require quite some Inkscape coding as well though to create > > on-canvas editing of the mesh. > > > > Regards, > > Johan > > > Well, my goal is to work on 2geom or lpe (that btw exploits > 2geom a lot), personally I'll prefer to implement some > algorithm or write test programs, rather than doing > integration, on the other side I'm a newbye and so an > integration task could be a soft way to learn 2geom library. > > Will it be ok if I spend a first part on an integration task > and a second part implementing some specific needed algorithm > for 2geom or lpe (and here suggestions are welcome) ? Ah that's not too bad an idea. What about the tweak tool Bulia is nagging us with ;) ? > Btw, I'm working on elliptical arc support in 2geom, just to > take more confidence with the library. > I think to be succeeded in implement a to_sbasis routine but > I have always to integrate large_arc and sweep flags support. Great! > > Could you point me to some paper on mesh distortion, so I can > evaluate the required effort to implement such a feature. Sorry I cannot. Perhaps 2geom already has a toy that does it, 2dsb2d. Please ask the 2geom guys if they know papers or have other info. Regards, Johan ```