- Inkscape Devel - Inkscape Mailman

GSoC 2009 project - lib2geom development
by Marco 31 Mar '09

31 Mar '09

Hi all! After some talk with Nathan Hurst I decided to apply for GSoC also this year. I would like to go on with the development of lib2geom that as I can see is now almost completely integrated into Inkscape. So, before submitting my project proposal to google I would like to know your opinions! Kind Regards, Marco =======================PROPOSAL START============================ Name: Marco Cecchetti Project: lib2geom development Abstract Lib2geom has become the core geometric library of Inkscape and it is also an open source project on its own. My project aims to expand the library and give evidence of its skills through the implementation of ad hoc "toys": little interactive applications showing off the power of lib2geom. Moreover many of these toys could be suitable directly for live path effects. Goals (1) Complete the work on bezier clipping routines making them more robust, and implement more specialized intersection routines for the various curve types; (2) Adding support for more basic geometric shapes (triangle, regular polygon, ...) so making easier the development of educational and technical geometric tools. (3) Implementation of a new curve type (clothoid) to let 2geom supporting spiro tools into Inkscape; a clothoid toy showing off the feature of the implemented curve will be provided, too. (4) In case I have some time left after fulfilling the other goals I would like to implement a parallel version of some 2geom routine using a multi-thread library, and explore the possibility of exploiting vectorialized support (e.g. SSE2) provided by modern cpus. Details (1) Towards the end of my last year project I start developing routines that compute the intersection points and the collinear normals of two Bezier curves. Both these routines are based on the Bezier clipping technique as it is explained on Sederberg's book [1]. Though at 99% these routines work correctly there are still cases where they show some spurious behaviour. As they are really useful tools I think that spending some time in order to make them more roubust it is a worthwhile task. Moreover I would like to investigate the utility of implement specialized intersection routines for various curve tipes (bezier-conics), (conics-conics) (conics-line) and so on. This would save the time needed to convert them to symmetric power basis or to bernstein basis before computing the intersection and would make also the computation more robust by removing the possible approximation errors introduced by the conversion stage. (2) In order to simplify the implementation of technical drawing tools or the development of a geometric educational version of inkscape, I think that it is useful to have direct support in lib2geom for simple geometric objects and the availability of geometric construction methods. In my last year project I implemented several tools, such as (infinite) line, ray, circle, ellipse, arcs and I provided routines for making up parallel line, perpendicular line, angle bisector, circle by 3 points, ellipse by 5 points. Now I want to go on adding new geometric objects such as triangle, regular polygon, and the left conic sections, moreover I want to implement other construction methods such as circle by 2 points and tangent to a line, inscribed and circumscribed circle of a triangle and of a regular polygon, conic section by 5 points and so on. (3) A Clothoid, also known as Euler's spiral, is characterized by the property that its curvature varies linearly with the arc length. Because of this important property they are useful in the route design of highways and railways to achieve smooth transitions between lines. For the same reason they are also appreciated in the design of font types. I'll provide the Clothoid class with a complete implementation of all methods inherited by the abstract Curve class such as bounds computation, evaluation for a given parameter value, conversion to and from a s-power basis path, and so on. The Clothoid-Toy will provide the ability of interactively play with a clothoid curve by modifying the initial and the final clothoid arc points, set its length, watch the projection of an external point on the curve, and all other abilities that the Clothoid class will own. (4) Several 2geom routines such as the bezier clipping algorithm and the curve-curve distance computation will benefit from a parallel implementation. The trick is in partitioning the work in several tasks that can be executed by different threads managed by a task scheduler. All that can be achieved using a multi-thread library like the Intel Threading Building Blocks (released under the GPL license) or through the Boost.ThreadPool library still in active developement. Likewise exploiting a vectorialized instruction set such as SSE2 will allow us to parallelize the computation of large amount of data by producing more compact assembly code, that is working on "packed" data we save a large number of clock cycles. Benefits to Inkscape My first goal is to simplify the development of technical drawing tools into Inkscape and the development of an educational version of Inkscape focused on geometry learning and here I think to programs like Cabri [2]. All the first 3 points of my project go in this direction. The geometric construction methods I want to implement will make possible to deploy new technical live path effects. Expecially the fulfil of the third point will simplify the design of a complete spiro tool into Inkscape (at present such functionality is supported by a live effect). Having a complete and robust api for curve intersections will be useful to simplify the implementation of cut path tools, and boolean shape operations. Finally, the fourth point will put the basis for exploiting the power of modern cpu architetures. About me I am a student in mathematics (University of Pisa, Italy) I am always been keen on computer graphics and I am interested in computational geometry. My main programming language is C++, but in the past I also programmed in pascal, fortran 90, assembler, java, javascript, actionscript, bash script. Moreover I am knowgeable about several Boost libraries, XML APIs and the Scalar Vector Graphics Markup Language. Contributions to open source projects My main contribution in the open source field is the Google Summer of Code Project I applied for the last year. It was about 2geom development. In the project time-frame I achieved the following goals: - Implementation of a new curve type that fulfils the traits of an SVG Elliptical Arc and of an "Elliptic Toy" for playing with this curve type. - Some basic GSL library wrapping classes such as Vector, VectorView, Matrix, MatrixView and a fitting tool. - Ellipse and Circle curves with the ability to be created or by the algebraic implicit equation, or by providing a sufficient number of points lying on the curve. - Nearest point routines for all curve types, computing the point on the curve nearest to a given external point. - A curve-curve distance routine. - A Polynomial class for symbolic computation. - Implicitization method for polynomial curve intersection, and Bezier clipping method for Bezier curve intersection and for collinear normal matching. - Several simple geometric objects and construction methods that I have already listed above. I'm developing an extension of Boost.Function [3] in order to implement a model of function concept supporting multi-signature and so overloading. [4] 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 consists in having added svg text element support to the import filter. [5] [1] Sederberg, Computer Aided Geometric Design [2] http://www.cabri.com/cabri-2-plus.html [3] http://www.boost.org/doc/html/function.html [4] source code: https://svn.boost.org/svn/boost/sandbox/overload/trunk documentation: http://docs.google.com/View?docid=dfxvjncs_1x6456m [5] source code: http://tinyurl.com/2es9cu =======================PROPOSAL END============================== -- Using Opera's revolutionary e-mail client: http://www.opera.com/mail/

2 1

Libre Graphics Meeting 2009 Call for Presentations
by Jon Phillips 30 Mar '09

30 Mar '09

Apologies for cross-posting. Please spread around :) ### Hi, FYI, the standard LGM press release: Libre Graphics Meeting (LGM), the premiere workshop and conference for developers and enthusiasts of free software graphics, will be held May 6-9, 2009, at Ecole Polytechnique in Montreal, Quebec, Canada. LGM invites you to share your work with the community. Topics of interest include reports on major open source graphics projects, technology previews, engineering talks, power-user techniques, graphics business best practices, and general issues such as open file formats and collaboration. Interested projects and individuals should submit proposals for presentations, demonstrations, and birds-of-a-feather (BOF) sessions by visiting the talk submission page at the CREATE project wiki. Proposals should be at most two paragraph abstracts and should include any special technical requirements. http://create.freedesktop.org/wiki/Conference_2009/Submit_Talk Individuals who already plan to attend can now register free of charge at the LGM Web site. http://www.libregraphicsmeeting.org/2009/signup.php The call for participation ends April 1, so don't delay! About Libre Graphics Meeting: For four years, the Libre Graphics Meeting has been the premiere conference for developers, users and supporters of free software graphics applications. Developers from projects such as GIMP, Inkscape, Blender, Krita, Scribus, Hugin, the Open Clipart Library, and the Open Font Library gather to work on interoperability, shared standards, and new ideas. Work at prior LGMs has pushed the state of the art in important areas such as color management, cross-application sharing of brushes and other assets, and common formats. The face-to-face meetings and opportunities for collaboration are important to developers, but LGM offers plenty for end users as well. Tutorials, talks, and birds-of-a-feather (BOF) meetings to help free software users get the most out of their applications fill out the LGM schedule, and demonstrations from artists showcase what is possible. For more information, visit www.libregraphicsmeeting.org To support LGM 2009's community fundraising campaign, visit pledgie.com/campaigns/2926 -- Jon Phillips http://rejon.org/ +1.415.830.3884 (sf/global) +86 132-6817-8381 (china) Sent from: Beijing 11 China.

2 1

Re: [Inkscape-devel] [FWD: [Inkscape-docs] Inkscape File formats]
by John & Melonie Curwood 30 Mar '09

30 Mar '09

Hi, Sorry It's taken so long to get back to you. You asked about a format list after my last emial. well here they are. Now I think I have the lists correct, but it seemed everytime I looked through the Open and Save dialogs I saw something new, so I've been through the dialogs several times refining the lists. To save space I will only add the discrepncy lists to the email. If you want the whole lot just let me know. FORMATS THAT CAN ONLY BE OPENED IN LINUX .outline .ps .eps .cdr .txt .icns FORMATS THAT CAN ONLY BE SAVED IN LINUX .ai .epsi .dxf (AutoCAD DXF) .xcf FORMATS MENTIONED IN MANUAL THAT AREN'T IN EITHER wIN OR LINUX .dia FORMATS NOT IN MANUAL THAT CAN BE OPENED/SAVED BY WIN OR LINUX .emf .wmf .wpg .gif .cdr .txt .icns .gpl .zip Now as I said in my original email the final list I am happy to add to the manual, it's just here for general info. Cheers, John Curwood Alexandre Prokoudine wrote: > On Fri, Mar 13, 2009 at 10:51 PM, <marketing@...2141...> wrote: > >> Hi everyone, >> >> I'm Currently working on the Floss Manuals Inkscape manual. At the moment I >> am editing the working with files chapter and have a few questions about >> file formats. I originally posted my question to the Inkscape docs list and >> was advised to ask my question of the Inkscape-dev list. My original >> posting is below. >> >> I look forward to any info you may have. >> > > You were also asked to provide list of file formats of interest :) > > Alexandre > > >

3 2

Re: [Inkscape-devel] Rendering problem for an object with blur / 0.46 Official vs 209225 vs 20967
by worms invasion 30 Mar '09

30 Mar '09

I don't know, really. I *am* on not-so-fast machine myself, and seeing so many "legacy" drawings being degraded this way is... awful. I just can't imagine all the people who will encounter the same experience, and what they will think about Inkscape. What worries me, is that it will force everyone to use "better" or "best", when obviously they don't need it *really*, in fact. Not everyone uses advanced filters. Where is the benefit? I can't, I really can't work on awful renderings like what 0.47 shows, and using a better filter quality puts my machine down. Inkscape has become unusable for me. The dramatic change in the "average quality" is a total turn off :-( Maybe the blur and filter qualities could be stored on a per-document basis, and not for Inkscape itself? Maybe Inkscape could recognize that the file was created with This or That version, and issue a warning of some sort in that case when opening the file, saying "this file was created with an earlier version of Inkscape, the average blur and filter quality will be set to better for this file". These "old" files do not contain advanced filters as 0.47 allows them, anyway, so the performance degradation should not be noticeable and the rendering will be correct on canvas, allowing proper editing? I don't know what would then happen if the user would want to add the (awesome) new filters to the "old" file. I don't know :-( --- On Sun, 3/29/09, Jasper van de Gronde <th.v.d.gronde@...528...> wrote: > From: Jasper van de Gronde <th.v.d.gronde@...528...> > Subject: Re: [Inkscape-devel] Rendering problem for an object with blur / 0.46 Official vs 209225 vs 20967 > To: "Inkscape Devel List" <inkscape-devel(a)lists.sourceforge.net> > Date: Sunday, March 29, 2009, 6:41 AM > worms invasion wrote: > > ... > > For all the reasons explained above, I think that, > although done with > > the best intention, the fact that the blur/filter > display "average > > quality" has changed results in the opposite effect > as the one > > intended. It seems to me that, at the moment, there > are more SVG "done > > with 0.46" than "done with O.47', and that the case I > am describing > > will arrive for a long, long time... > > > > If I missed an obvious point, I'd very much like to > hear it. > > I don't think you missed anything, and perhaps we should > rethink the > default setting of the filter quality, but the fact remains > that for > people with not-so-fast machines the best filter quality > simply isn't an > option when interacting with the graphics. (And it is in > fact not > uncommon for drawing programs to have low-quality rendering > modes.) > > So, assuming that simply removing the lower quality > settings is > (currently) not an option, what do you think is an > acceptable way of > decreasing the quality in exchange for speed? > > ------------------------------------------------------------------------------ > _______________________________________________ > Inkscape-devel mailing list > Inkscape-devel(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/inkscape-devel >

3 2

GSOC: Cross platform WMF and EMF IO and other ideas
by Krzysztof Kosiński 30 Mar '09

30 Mar '09

I have several ideas for GSOC 2009. I'll present all of them here. Idea 1 is about improving Inkscape's data interoperability. This means being able to paste parts of drawings from Inkscape directly into OpenOffice.org and MS Word and the other way around. This would involve writing a cross platform WMF and EMF output extension (our current EMF extension is Windows only). This could be accomplished by using libwmf, but this project seems to be abandoned since 2002. Another option is to use some code from OpenOffice.org. Recently (2006), MS published a full specification of those formats, so writing a new library to read and write them isn't out of the question. Idea 2 is has a similar goal to Idea 1 and it is to revamp the I/O extension system. This would involve unifying the code paths for open/save, export/import, clipboard, drag and drop, and create bitmap copy; unifying command line parameters for export (e.g. all extensions would accept --export-id); converting all extensions to load/save using GIO streams; and redesigning the Export dialog to allow exporting in any supported format. Idea 3 is to refactor the tools code, so that each tool is a separate object, and the number of files that list all tools is reduced to a minimum. This would get rid of toolbox.cpp, as well as greatly simplify the writing of new tools. Idea 4 is to replace Verbs with subclasses of Gtk::Action, with only one class per action (instead of one Verb class per several verbs). There would be a separate action group per document, and an additional one for global actions. Actions could be registered dynamically at startup. Which idea, when implemented, would be the most beneficial to Inkscape? Regards, Krzysztof Kosiński -- View this message in context: http://www.nabble.com/GSOC%3A-Cross-platform-WMF-and-EMF-IO-and-other-ideas… Sent from the Inkscape - Dev mailing list archive at Nabble.com.

5 18

SWIG bindings
by Krzysztof Kosiński 30 Mar '09

30 Mar '09

I stumbled upon this interesting project: http://www.swig.org/ It is a binding generator with support for a lot of languages. We could use this to kickstart Inkscape scripting (as a trivial example, it would be possible to change the license of all files in a directory programatically) and allow extensions to call internal Inkscape functions (e.g. path boolean operations, convert stroke to path, etc.). There was also a GSOC project idea about this from another student, I think he should definitely use this. Regards, Krzysztof Kosiński -- View this message in context: http://www.nabble.com/SWIG-bindings-tp22774855p22774855.html Sent from the Inkscape - Dev mailing list archive at Nabble.com.

1 0

Re: [Inkscape-devel] Inkscape-devel Digest, Vol 34, Issue 40
by Guillermo Espertino 29 Mar '09

29 Mar '09

> It's not exactly like this. Before my hanges we used scalable icons > rendered > by Inkscape's own renderer. Now we still use scalable icons through a > standard theming mechanism, but it isn't playing well with the > existing code. Krzysztof: If I remove the hicolor directory I get no icons at all (running ubuntu linux). The previous code worked and let the user change the theme. In the current state, all I got is an incomplete set that I cannot customize. Wouldn't be possible to roll back to the working version and move these changes to a branch? Gez

1 0

Re: [Inkscape-devel] GSoC Idea: Inkscape API reg. tests
by Glimmer Labs 29 Mar '09

29 Mar '09

On 3/29/09, Jasper van de Gronde <th.v.d.gronde@...528...> wrote: > > This sounds absolutely wonderful. And if you do this, be sure to > consider opportunities for testing inkscape as well. Some time back I > tried testing Inkscape just based on verbs and this already allowed me > to do some cool stuff (but not enough to be standardized into a > framework), having more control over Inkscape's actions could allow > semi-automatic testing of quite a few Inkscape features in a way similar > to the current rendering tests. > That's a very interesting idea, and underscores the importance of a generalized API, you never know what people are going to use it for. Here a quick proof of concept I wrote to try it out: Using my old API I wrote a script that made various shapes and did some selections moved things around etc. Then I ran it twice and saved the file each time (as svg) and ran a diff. No difference but the filename, just as I had hoped. Worried that lots of floating point math would be it's downfall, I used the same technique with scripts that generate these images: http://www.cs.grinnell.edu/~bergsore/dimond-fractal.jpg http://www.cs.grinnell.edu/~bergsore/morph.jpg (They are jpegs here but you can imagine how messy the paths are.) It worked perfectly. Obviously this won't catch errors in the way Inkscape displays things, like bitmap comparisons, but if automated it would be very quick. A fast and easy alternative regression test could definitely be one of my deliverables. What do you guys think? Soren Berg

1 0

Re: [Inkscape-devel] Rendering problem for an object with blur / 0.46 Official vs 209225 vs 20967
by worms invasion 29 Mar '09

29 Mar '09

Hi, I'll try to explain better what I mean with a new screenshot: http://www.corpsmoderne.net/~eve97/vertex/rocket_bug/display_inkscape_046_v… First, I'd like to state a few things: - I'm not an artist, but I draw things in Inkscape, and I draw them in SVG to have them displayed as SVG. So my "visual perception" of SVG is what Inkscape shows to me, not a PNG (which I often not even generate). It means I design the object *as I see it rendered on Inkscape canvas*. - I have several drawings made with 0.46 or previous versions of Inkscape, and there are many places where I could download SVG designed with that (openclipart, etc.) So, let's take the following hypothesis: - After seeing a nice SVG in my browser, I download or open in Inkscape 0.47 a file which was drawn with a previous 0.46 version, in order to use it as a clipart, or to rework on it. On the top of the screenshot, it is a shape drawn in Inkscape 0.46, with filter quality set to average (default). It is what I see in my browser, or on the PNG if I generate it. My 0.47 Inkscape is set to default for both Blur display and Filter display, as I have a small machine (512Mb memory, Intel Pentium III mobile, 1133Mhz). I open the file, and I see what shows at the bottom of the screenshot. I hope you'll admit that it is awful and doesn't look like at all what shows in my browser. BUT it is what I work on, I have no other way to work on SVG. I'm not switching back and forth to PNG or to the browser to check what it looks like there. What am I supposed to do at this point? Remember, I'm an average user, not knowning the Inkscape internals. A) I say: this Inkscape program is mad, I'm not going to use it. I close it, end of story. B) I could try to change the blur quality and/or the filter quality to better or best. OK, I do that. But the problem is, the canvas is not updated when I do that, and it's a general preference, not a per-document preference. So, first, I'll have to: - with my document still opened, change the Blur / Filter effect and see that what is displayed on canvas has not changed. - close the document and Inkscape, reopen Inkscape which now has Blur and Filter set to better or best, and see that the file looks like what it should. (or something similar, because I have no idea of the exact real value of the quality). SO: setting the displays to a "lower" average quality in Inkscape 0.47 than in Inkscape 0.46, which was supposed to be intentional to help the designer (the designer in Inkscape 0.47) conserve CPU, results in fact in obliging me (who opened on older file) to set Inkscape to high quality, for all my documents. And I have a small machine - I can't afford to do that. C) I could just do what I was adviced to in the channel: remove the filters on the object. This means I need to find all objects with blur/filter, one by one, and remove that. After such a tedious task, then I lose the original information (here, the blur was 0.3), meaning I am absolutely unable to re-create or re-see the original file, as the artist designed it. What am I supposed to do now? For all the reasons explained above, I think that, although done with the best intention, the fact that the blur/filter display "average quality" has changed results in the opposite effect as the one intended. It seems to me that, at the moment, there are more SVG "done with 0.46" than "done with O.47', and that the case I am describing will arrive for a long, long time... If I missed an obvious point, I'd very much like to hear it. Regards. --- On Fri, 3/27/09, Niko Kiirala <niko@...1267...> wrote: > From: Niko Kiirala <niko@...1267...> > Subject: Re: [Inkscape-devel] Rendering problem for an object with blur / 0.46 Official vs 209225 vs 20967 > To: inkscape-devel(a)lists.sourceforge.net > Date: Friday, March 27, 2009, 1:03 PM > Fri, 27 Mar 2009 09:31:44 -0700 > (PDT) > worms invasion <wormsxulla@...36...> > kirjoitti: > > > In short: the exported PNG appears correct in the > three versions, but > > the screenshots of the rendering in Inkscape look very > (or very) > > different across versions. The file looks absolutely > normal, as I > > draw it, in 0.46 official. > > > > In r20225, the object is already badly rendered in > Inkscape. > > > > In r20967, it's awful... > > The default values for filtering quality in r20967 are not > too good. > Neverteheless, as you noticed, the lower quality is used > only while > editing the image and the exported image looks the same in > all > those versions. > > Lowering the quality while editing is intentional: > rendering blur can bog down even the mightiest > workstations. If it > doesn't bother you, you can turn filter quality settings > all the way up > and the quality will be just as good as while exporting. > > -- > Niko Kiirala > niko@...1267... > > ------------------------------------------------------------------------------ > _______________________________________________ > Inkscape-devel mailing list > Inkscape-devel(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/inkscape-devel >

2 1

GSOC: Interested in joining Inkscape
by sonali mendis 29 Mar '09

29 Mar '09

Hi, I'm Sonali Mendis from University of Moratuwa, Sri Lanka. I went through the project ideas listed at Inkscape page and I think it is a great project of my interest. I'm a graphic designer plus a developer and I think and I want to add my ideas and effort to improve this open source project. While using Inkscape, I thought *Guide Improvement *is essentially needed. In addition to that I have another my own idea of introducing different brush types with ability to increase and decrease the size by click of a key while drawing to the pencil tool and introduce an eraser tool with the same options. I need your kind feed back on the ideas mentioned above to select one which will benefit the Inkscape most. Thanking You, Regards, Sonali Mendis

3 2