- Inkscape Devel - Inkscape Mailman

User account for Inkscape Wiki
by Eduard Braun 26 Nov '15

26 Nov '15

Hi all, I'd like to request a user account for http://wiki.inkscape.org/ I noticed documentation on INX parameters <http://wiki.inkscape.org/wiki/index.php/INX_Parameters> is quite outdated and I want to update it. Best Regards, Eduard

2 3

Inkscape's software design
by Jorge Lopez 26 Nov '15

26 Nov '15

Hello, I'm studying a Master Degree in Computer Engineering and we have to make an article about an Open Source program (Inkscape in my case). In this article we are suposed to interview some developers in order to know more about software design in the project they are working on, so I'm looking for some people willing to answer some simple questions about software design on Inkscape's development. If you are interested, please contact me ( jorgeludc@...400... ). Thanks in advance!

1 0

Fwd: Save in JPG files
by Bruno de Campos 23 Nov '15

23 Nov '15

Hi, I'm using Inkscape and I realize that the program does not save in .jpg format. This is kinda disappointing because I have to open another program to save my file in a proper format. Do you have any tips or extensions or any updates that I could use to save in jpg format? PS.: I'm using Version 0.91 Thanks in advance for your attention, -- Bruno de Campos

4 8

Offset LPE
by Jabiertxo Arraiza Cenoz 21 Nov '15

21 Nov '15

Hi to all, working on offset LPE by "accident" y discover the way to remove ovelaps, so the code on evenodd mode is ok. Now I need to handle correcly nonzero cases and make some tweaks to the knot handler. Krzysztof, Liam sorry for the noise and thanks for your works! Cheers, Jabier.

1 0

LPE: Use sigc++ from PathParam to modify another Param
by Papoj Thamjaroenporn 21 Nov '15

21 Nov '15

Dear Inkscape-devel, I would like to find a way to trigger a change from PathParam and change another Param in the same LPE effect. What is the correct way of doing this? What I have attempted: I noticed that PathParam already provides a bunch of useful signals, like signal_path_pasted (when a new path has been pasted in placed of the existing path in PathParam). So I just set up in my LPE effect like: myPathParam.signal_path_pasted.connect( sigc::mem_fun(*this, &LPEWrinkleStretch::triggerAChangeInPeakPointArrayParam) ); And in my own method, I just modify another WrDmPeakPointArrayParam with set_and_write_new_value(). Nothing fancy here. Although I have made sure that the change happened and the resulting change was what I expected, this notification-callback approach seems to break the LPE entirely: now interacting with anything in the effect afterward (moving other unrelated handles, changing values, etc) will not update the LPE at all, although the doEffect() method is still called constantly. So obviously, I am doing something wrong here. I have noticed that there is another effect, lpe-bspline, that makes use of the callback mechanism to update the effect. I understood the basic mechanism: you process path effects normally, create a result in SVG format in text, save that text onto the internal XML “Repr” directly, and then you provide some sort of save state for undoing. However, I am not sure how to do this for another Param (not the resulting path from the LPE effect). Do I write my own version of saving such param into the internal XML tree directly? I don’t really have a new widget, but an update should be triggered after clicking on “Paste path” button in PathParam, so do I have to override newWidget() method? Can someone also explain to me how DocumentUndo::done() works (what the variables are)? Thank you for always being supportive and helpful. Regards, _______________________ Papoj "Hua" Thamjaroenporn papoj@...3117...

2 1

TSL errors for LP
by Brynn 20 Nov '15

20 Nov '15

Hi Friends, I'm having trouble connecting to Launchpad today. (1 success out of 8 tries :-( ) The error says "Secure Connection Failed" with option to report it to Mozilla. Info: https://support.mozilla.org/en-US/kb/tls-error-reports "If you experience this problem, contact the owners of the website and ask them to update their TLS version to a version that is still current and still secure." Has this already been reported to LP? If not, how can I do that? There doesn't appear to be any way to allow the connection to go through anyway, but I'm still investigating. Maybe have to use IE? Thanks for your help :-) brynn

2 2

Union PathVector
by Jabier Arraiza 20 Nov '15

20 Nov '15

Hi to all. I want to make a pathvector joined with itself like Inkscape do whith only a path selected, with subpats and in non-zero fill rule. I look to Krzysztof code with boolops but it handle it with diferent pathvectors, not with the same. If anybody can point me on a direction, his help are welcome. All the best, Jabier.

2 1

Offset LPE
by Jabiertxo Arraiza Cenoz 18 Nov '15

18 Nov '15

Hi to all. I just finish a LPE called "Offset" based on Liam and Krzysztof work. The branch give better results than actual offset LPE, more similar to Liam JoinType LPE than a Offset one. Anyway, the LPE could be improved working on Geom/PathStroke and Krzysztof boolops code. As a fast improvements things to do: Boolops: Commented in previous mail, I have problems with fill-rule options on AminusB function. Also if we can handle overlaping curves, I could simplify the LPE. PathStroke: Handle strange results specialy on very big width of stroke Tavmjong, you remember our conversation about extrapolated arcs in Toronto? ¿Need a refresh? ¿Could you explain to Liam in a math point of view, to know if he want to make the proposal for the SVG working group? lp:~inkscape.dev/inkscape/offset All the best, Jabier.

1 0

Vertical text & 'text-orientation'
by Tavmjong Bah 18 Nov '15

18 Nov '15

Hi, I've implemented the CSS 3 'text-orientation' property which controls how glyphs are rendered in vertical text. This property replaces the 'glyph-orientation-vertical' property which Inkscape never implemented and is deprecated in SVG 2. While doing this work I also corrected some other layout issues with vertical text and with multi-line text. As this work is rather invasive I've pushed my changes to a new branch which can be checked out via: bzr branch lp:~inkscape.dev/inkscape/writing_modes Please test this branch! Here is a list of the most important changes: * Latin characters in vertical text will be drawn by default sideways. This is the default direction in both SVG 1.1 and in CSS 3 Writing Modes. Currently Inkscape draws the glyphs upright. * Punctuation characters in CJK vertical text will be orientated correctly. * The Text Toolbar can be used to set 'text-orientation' which controls how characters are orientated in vertical text (it has no effect on horizontal text). The three choices are 'mixed' where the glyphs are drawn in their Unicode defined natural orientation, 'upright' where all glyphs are drawn upright, and 'sideways' where all glyphs are drawn sideways. * The new CSS 3 values for 'writing-mode' will be recognized. The values written out are still the SVG 1.1 values except for the new 'vertical-lr' value which has no equivalent SVG 1.1 value (and is required for Mongolian). A new button appears on the Text Toolbar to select this value. (The SVG 1.1 values are deprecated but must still be recognized by SVG renderers.) * Fonts without vertical metrics will have a vertical advance of 1em. The CSS specification explicitly does not define what advance to use but the SVG 1.1 spec suggests 1em and the W3C Japanese text document specified an advance of 1em. * In laying out multi-line text, any 'leading' value from the font is ignored. CSS dictates that the font-size corresponds to the em height of a font which is defined as 'ascent' + 'descent'. The change to ignore the "internal" font 'leading' may make small visual changes in how multi-line text appears. * In flowed text, Inkscape has been using a "scaled" value of 'ascent' to position the first line. This has been changed to follow the correct CSS prescription of using the value of 'half-leading' + normal 'ascent' for horizontal text (to find the alphabetic baseline) and 'half -leading' + half em width for vertical text (to find the center baseline). (The 'leading' in this case is the extra space needed between lines as dictated by the 'line-height' property. It is not the 'internal' font 'leading'.) Tav

3 6

Offset LPE
by Jabier Arraiza 16 Nov '15

16 Nov '15

Hi to all. Im on Offset LPE using Liam and Krzysztof work. I have two questions, First: Using "PathIntersectionGraph" to make boolean operation "getAminusB", givea unwanted result, to me, maybe because fillrule work as evenodd. I remember read in list the fill rule work is not ended on new boolops, I'm wrong Krzysztof? Second: If I'm brong with first, I need to try "manual" erase the unwanted data, for example in the middle of a infinite sign path. To do this I run the function: "self_crossings" but this function give me points not self crossed in the path, also, give me, for each real crossing, one of the times not the two that have a self crossing. Maybe Im wrong in th concept of selfcrosing... Thanks in advance for the help, all the best, Jabier.

1 0