- Inkscape Devel - Inkscape Mailman

glib 2.36.0 deprecates g_type_init
by Stefan Husmann 18 Mar '13

18 Mar '13

Hello, in glib 2.36.0 it is no longer allowed to declare g_type_init, the following patch solves this for me in inkscape. diff -Naur orig/src/extension/dbus/wrapper/inkscape-dbus-wrapper.c new/src/extension/dbus/wrapper/inksca pe-dbus-wrapper.c --- orig/src/extension/dbus/wrapper/inkscape-dbus-wrapper.c 2013-03-18 00:08:01.000000000 +0100 +++ new/src/extension/dbus/wrapper/inkscape-dbus-wrapper.c 2013-03-18 00:09:27.000000000 +0100 @@ -95,9 +95,9 @@ DBusGConnection *connection; GError *error; DBusGProxy *proxy; - +#if !GLIB_CHECK_VERSION(2,35,0) g_type_init (); - +#endif error = NULL; connection = dbus_g_bus_get (DBUS_BUS_SESSION, &error);

1 0

cairo cms @ gsoc 2013
by Alexandre Prokoudine 18 Mar '13

18 Mar '13

FYI https://www.oyranos.org/wiki/index.php?title=OpenICC/GoogleSoC2013#Get_the_… Alexandre Prokoudine http://libregraphicsworld.org

1 0

Trunk merge into branch request
by Johan Engelen 15 Mar '13

15 Mar '13

Hi all, If someone remembers how, please merge trunk into this branch: https://code.launchpad.net/~inkscape.dev/inkscape/guides-improvements (the last merge was done Feb 11). Thank you very very much, Johan

1 0

Unused header: dom/svg2.h
by Alex Valavanis 15 Mar '13

15 Mar '13

Seems like nothing includes the dom/svg2.h header. Can I kill it? AV

2 1

the website
by Alexandre Prokoudine 14 Mar '13

14 Mar '13

Hello, I've just committed news on GSoC, and here's the output I'm rather concerned about: Warning: post-commit hook failed (exit code 1) with output: ********************************************************************** NOTE: This is an informational message only; your commit has been processed normally. This repository has moved to a new location. This project was upgraded to the new SourceForge platform, so a one-time move of this repo was performed. There are instructions for switching to the new location, including any commits you may have done on this repository after the upgrade, at: https://sourceforge.net/p/forge/community-docs/Repository%20Upgrade%20FAQ/ The new repository URL can be found by visiting: https://sourceforge.net/p/<your-project> This repository will soon become read-only. ********************************************************************** Alexandre Prokoudine http://libregraphicsworld.org

1 0

Ask for an account under wiki.inkscape.org
by Xiangfu Liu 14 Mar '13

14 Mar '13

Hi I would like to update this page: Compiling_Inkscape_on_Windows I meet some error when install the TDM's GCC and compile btool. TDM's GCC needs 'Uncheck' the [Check for updated files on the TDM-GCC server] or it will always install the latest gcc Compile btool also can be done by only g++ buildtool.cpp -o tool Make sure you have correct DEVLIB_PATH and MINGW_PATH Thanks Xiangfu

2 1

What is the difficulty in adding text decorations, especially underline?
by mathog 14 Mar '13

14 Mar '13

There have been user requests for underline (a text decoration) going back years and years. Since it has not shown up I assume this means that some library Inkscape depends on cannot do this. I checked the most obvious suspects and did not find a definitive smoking gun: 1. Freetype does support underlining (both location and thickness). http://www.cs.sunysb.edu/documentation/freetype-2.1.9/docs/tutorial/step2.h… 2. SVG supports it with "text-decoration:underline". 3 Pango supports it: https://developer.gnome.org/pango/stable/PangoMarkupFormat.html 4. Cairo - not sure. (Maybe this is the guilty party?) So, is there a document somewhere which explains why Inkscape cannot underline? I looked for a while but all Google turned up were lots of threads about how Inkscape does not have underline, but not _why_ it does not. Thanks, David Mathog mathog@...1176... Manager, Sequence Analysis Facility, Biology Division, Caltech

2 2

GCC version checking
by Alex Valavanis 13 Mar '13

13 Mar '13

Further to bug #992124 <RFE: clang support> [1], is there any reason why we still have a hardcoded check for gcc >= 3.0 in the configure script? 1. It prevents easy use of non-gcc compilers (clang) 2. Version 3.0 is archaic, and is presumably available on every platform that could feasibly build Inkscape (i.e., other infrastructure checks in the configure script would fail, so the test is redundant) 3. Why are we specifically looking for version 3.0? If we're wanting to check for a compiler feature, there are more elegant ways of checking for this than just looking for the version number. If no one objects, I'll remove the version check. Best wishes, Alex [1] https://bugs.launchpad.net/inkscape/+bug/992124

2 1

transform inkscape:original-d into a defs + LPE
by pennec victor 13 Mar '13

13 Mar '13

Hi, Why original path are stored directly into the shape as an inkscape:original-d attribute rather than in a defs (like path-effect) ? Why is there no "simplify/smooth" LPE ? When you draw line (F6), a smooth factor is applied and a certain shape (ellipse, growing triangle) is used. If you'd like to replay different settings on the same input then you're out of luck : the original output is gone forever because the nodes are already discarded. Some we'll be able to remove the pattern along a path by going into the LPE, most will clik/drag the tool controls and ask why they can't modify settings AFTER the shape has been drawn How to submit a blue print ? Thank you for reading

3 2

Your opinion about "Save as expor"
by Maggio Mago 11 Mar '13

11 Mar '13

Hello All, I'm wroking on this blueprint : https://blueprints.launchpad.net/inkscape/+spec/save-as-vs-export I took some liberty and I'd like to have your opinion on the rendering. Regards, PS : About : "Maybe allow several names in the field to export several groups, each to its own file (then default the list to the current selection)" To answerd, I thought to use a perl script more specific.

13 47

filters with background image as source not rendering correctly and lack of support in browsers
by Yale Zhang 09 Mar '13

09 Mar '13

Hi, I created a filter that applies a displacement map to the background image to simulate effects like water caustics or the warping you see around fires, but it doesn't render how I want it to. The filter is being applied to the entire bounding box of the path instead of just inside the path. I tried it in Opera, which does support using the background image as a filter source and it shows the same effect. Is this correct behavior? I've attached an example of my problem. It's a star over a checkboard, but you should only see the checkerboard distorted inside the star. Also, could someone comment on why so few browsers (I only know Opera) support the background image as a filter source? To me, it doesn't seem any more complicated than just a regular transparency. Thanks, Yale

2 3

GdkEventMotion returning wrong (unexpected) x coordinates
by insaner 09 Mar '13

09 Mar '13

in the file filter-effects-dialog.cpp, I'm trying to get auto scrolling to work horizontally in the connections box of the filter editor dialog (for https://bugs.launchpad.net/inkscape/+bug/1062134 ). it already works vertically, and you can test this by dragging a connector down (you might want to apply the patch found there after youve verified how this works on your clean branch). however, im trying to get the scrolling to work horizontally too, but it seems that e->x is returning strange values.. strange in that it seems that when i scroll the box, it uses the coordinates of the primitiveList object.. not the actual visible area (ie, the far left edge is not 0 after i scroll to the right).. so to make it scroll i have to drag past where the left corner would be. so to give you an ascii explanation: __________ | | | : [ ] | | | __________ imagine the [ ] is the primitives box, where the scrolling should occur and connections dragging is done. the : is where i have to drag to for the autoscrolling to kick in while dragging a connector if the area is scrolled a bit to the right.. ie, the : represents where the name of the effect would be if it was visible. hope that makes some sense at all the dragging code is in this function: FilterEffectsDialog::PrimitiveList::on_motion_notify_event(GdkEventMotion* e) i can post a patch of what i have so far if necessary thanks in advance -- View this message in context: http://inkscape.13.n6.nabble.com/GdkEventMotion-returning-wrong-unexpected-… Sent from the Inkscape - Dev mailing list archive at Nabble.com.

1 0

[Fwd: Nodes with handles that have other nodes on the end! (That may or may not have handles)]
by Martin Owens 09 Mar '13

09 Mar '13

Forwarding to Inkscape devels. What do you guys think?

2 1

Re: [Inkscape-devel] Your opinion about "Save as expor"
by Alexandre Prokoudine 08 Mar '13

08 Mar '13

On Fri, Mar 8, 2013 at 9:58 PM, mathog wrote: >>> Agreed. Obviously I do not think we need Export at all. >> >> >> <skipped Britannica chapter> >> >>> 4. Region - self evident. The area could by two points (UL,LR as >>> here), or point,W,H. >>> This gets rid of the current "export as bitmap" (or whatever it is >>> called), and generalizes >>> it to all supported output formats. >> >> >> Excuse me, are you proposing to remove the current exporting dialog >> altogether? > > > It is worth thinking about. > > The "Export PNG Image..." is a special case > of the more general "Save as..." I outlined. Most of the features in > "Export PNG Image" are good (by selection, by area) but there is no > reason this sort of cropped output (essentially) should be restricted to > PNG. At least for SUM's "Save" (where "Save" means only to a fully > conservative file type) the choice of the SVG filetype makes sense > for Inkscape since no other fully conservative format is available. > But there is nothing special about PNG. Why not JPG or PDF? Use > the general "Save as..." form and the user can pick whichever target > format is needed for that particular application. I'm literally lost for words. Alexandre Prokoudine http://libregraphicsworld.org

1 0

SpiroLive + BSpline@...2952... Branch
by Jabiertxo Arraiza Cenoz 08 Mar '13

08 Mar '13

Hello to all. I release the new version of BSpline + BPower https://code.launchpad.net/~jabiertxof Fix too the problem tell me ~suv and i dont see at start. Previous releases are declared abandoned. The rasons to abandone previous releases are two: 1.- They are backward incompatible, the bspline node difinition is oposite now, think now is too much logical, and now you can increase-decrease the wheight of the bspline. I have a impotant question, when change the wheight of a curve/node its changed "symetricaly" (Move both handles at same), in the way of "Strict" Bspline. The cuestion is: Give the posibility to change independient handles give more control/power/confusion but maybe isnt strict BSplines. You always can press SHIFT key to move handles "symetricaly" like now. ¿Whats the comunity opinion about this? 2.- The code is fully refactored, more organiced and logical. I thick more easy to undertand. Not comented now, soon, but all my code starts whith //BSpline and end //BSpline End Things to do: 1.-Coment the code in spanish 2.- Translate this comments to english, Guiu maybe could help me again. 3.- Prepare transations to the GUI, the tooltip "SHIFT to CUSP node" The forum is updated whith some packages here:http://www.inkscapeforum.com/viewtopic.php?f=11&t=13496 Thanks to Pieter Barendrecht for start in my sylly head the idea of BPower. And thanks again to all interview in the work, brainless, druban, Guiu, Romain, Sorin, Arshdeep, ~suv, graficalibre, and all i couldent remember. By.

3 2

Strange Padding
by Bric 08 Mar '13

08 Mar '13

Hi, all: I'm getting a strange, annoying padding around my object, and I can't account for it. When I select the object, it seems that more than the object is selected, because various extensions report the error that "no path is selected." I've tried to extract the nodes and repaste them, but the annoying "padding" is preserved across the copy&paste. Here is one such object: http://www.flight.us/misc/strange_padding.svg

2 2

Synthesized font faces.
by Tavmjong Bah 08 Mar '13

08 Mar '13

Pango synthesizes Bold, Italic, and Bold Italic faces when they are not available on the system. Inkscape displays these styles in the UI. I can see a use for the synthesized Italic faces (actually Oblique) but I wonder how useful the synthesized Bold faces really are (it is not a trivial thing to do well). At the moment synthesized Bold faces are listed in the Style menus in the Text toolbar and the Text and Font dialog, and they are displayed in the Text and Font dialog preview area but they are not used on the canvas (probably a difference in how they are displayed, the former uses Pango markup in a Gtk widget while the latter is rendered by Cairo). Question: is it worth trying to get Cairo to render the synthesized Bold faces or should they be removed from the UI. What about the synthesized Italic faces? As an aside, I am struggling to understand how Pango handles font faces... mapping font names to family names and styles. For example, Impact which fc-list reports as Regular or Normal ends up having the style "Condensed"; the synthesized versions don't include "Condensed". Tav

10 20

Bug 1060837
by Arshdeep Singh 08 Mar '13

08 Mar '13

Regarding the bug: I found the piece of code that handles the CMYK values to be scaled as [0,1]. Now making a change to scale values in [0,1]-->[0,100] (though mentioned as a more industry friendly standard) shall require changes to generic codes which also handle the CMS sliders of RGB Color profiles. Also, the interconversions b/w floats of [0,1] to rounded of integers in [0,100] shall cause some loss of accuracy as well. Can people who are more experienced in the designing fields be brought together for poll ? If many people want this change, I am up for tweaking the source file in question. Pardon my audacious attempt of suggesting an idea of taking a vote, but in my respect it does seem important. PS: For the more experienced developers, am I right in thinking that the scaling will cause a loss of accuracy ? Thankyou for the help. --- demicoder

2 1

SVG question, pattern references another pattern but adds a solid background.
by mathog 06 Mar '13

06 Mar '13

EMF and WMF formats have BKMODE and BKCOLOR parameters which are applied to patterns. BKMODE is either transparent (no background) or opaque (BKCOLOR is drawn first, then the pattern is drawn over it). The tricky part for importing these formats and converting them to SVG is that the values of BKMODE/BKCOLOR are NOT employed when the pattern is defined, but when it is drawn. Here is an example of a pattern without a background (transparent) <pattern id="EMFhatch0_3F7FFF" patternUnits="userSpaceOnUse" width="6" height="6" x="0" y="0"> <path d="M 0,0 6,0" style="fill:none;stroke:#3f7fff" id="path15" inkscape:connector-curvature="0" /> </pattern> and here is the same pattern with a background (opaque) <pattern id="EMFhatch0_3F7FFF_FOE0DO" patternUnits="userSpaceOnUse" width="6" height="6" x="0" y="0"> <rect x="0" y="0" width="6" height="6" fill="#F0E0D0" /> <path d="M 0,0 6,0" style="fill:none;stroke:#3f7fff" id="path15" inkscape:connector-curvature="0" /> </pattern> What I really need to be able to generate, when the pattern is used, and BKMODE and BKCOLOR are both known, something like this (pseudocode, definitely NOT the right syntax!) <pattern id="EMFhatch0_3F7FFF_FOE0DO" patternUnits="userSpaceOnUse" width="6" height="6" x="0" y="0"> <rect x="0" y="0" width="6" height="6" fill="#F0E0D0" /> xlink:href="#EMFhatch0_3F7FFF" </pattern> How is that supposed to be written? The idea is to write the solid background color and then the transparent "base" pattern over it. It doesn't look like <path> can do it by referencing the "base" pattern, since style="fill:" is used for the URL to that pattern, so it isn't available to specify the solid background color. That suggests to me that it must be done within the <pattern> - but any technique that works would be fine. Thank you, David Mathog mathog@...1176... Manager, Sequence Analysis Facility, Biology Division, Caltech

1 0

Making devel packages, active extensions in devel
by Jabier Arraiza 06 Mar '13

06 Mar '13

Hi.I have two problems making packages for Inkscape. Im making a inkskape-devel package writting in rules -- --program-sifix=-devel I finaly compile and run the program as inkscape-devel, the problem is i coulden`t have simultaneus inkscape and inkscape-devel. ¿Any body know how to do it? The other cuestion is inkscape-devel have a lot of extensions disabled/empty. ¿Is posible to active it? Hi, Jabier.

1 1

Re: [Inkscape-devel] CMYK/Spot Color (was Synthesized font faces.)
by Jelle Mulder 05 Mar '13

05 Mar '13

Arshdeep, Great idea.. using 0-100% rather than 0-1 would be bliss for us old school guys ;-) In general using 2 decimals instead of integers doesn't make that much difference for most CMYK designs. Most of the upper and lower values are not reachable for the press anyway. For instance many digital print machines won't print values lower than 10% and stop at 90% at the default settings used by the print shops. On topic of hexachrome and spot colors. Only the very high end of the print work scale as used in my native country of the Netherlands, spot colors are default for any corporation that wants to be held serious. KPN for instance even tried to patent their colour green some years ago and they will pay the extra layer of ink without blinking. Hexachrome is a pretty basic use in glossy magazines that actually often add another layer of varnish to it as well as spotcolor. Direct to plate with stochastic rastering is not uncommon either. I've got an Uncle running a 6 color Heidelberg with an additional varnish extension who specialises in Art related books in Holland and that is mainly hexachrome. It is not a large market, but highly lucrative and relatively stable, unlike the lower end of printing. I personally have been doing print since 1993 and I've seen it all pass the revue. Holland may be a bit of a exception to the rule, but it has been a leader in print technology since the 1600's when all the forbidden stuff was printed there. Design is an export product as much as taters. If Inkscape would support CMYK natively, rather than having to use workaround with the export so you can check while doing the design it would make it a more useful tool. Late binding puts the responsibility of error correction at the desk of your printer as well and most designers I know want to be able to do that themselves. Using color proofing printers and using screen callibration machines to get the whole design department on the same level. Those are not companies that are scared to spend a buck or what on tools and equipment as they also are paid in due for the use of it. Now sure, Inkscape does not have to pretend to operate on the same level, but if it is possible to support the basics I'd say it should be a consideration. I'd love to be able to do a color seperation in Inkscape in a similar way as I can in Coreldraw for instance, so I could check the output raster of my original Magenta raster with what is actually being printed in the print shop. Currently that is a tad cumbersome. My 2 Feng, Jelle

2 1

Re: [Inkscape-devel] Synthesized font faces.
by Jelle Mulder 05 Mar '13

05 Mar '13

Hi all, From a designers point of view I would stay away from generated faux font styles as far as possible. There is a reason why not all fonts have these styles that often go beyond the humongous amount of work that needs to be put into creating a font. If a designer really wants to screw up a font creators work, they can use stroke thickness, skew, stretch or even envelopes. But any experienced designer not satisfied with the look of a font will normally find himself exactly that one that fits his desired balance of whitespace and content. It's part of the skill of a designer to find the correct font that supports the design and typography. Font design itself is a work of dedication and skill that has always been beyond me and goes well beyond what normally meets the eye at 14px. Having tools like Fontforge really didn't do font design all that good if I have a look at modern fonts with their horrible flows, kerning, baseline shifts, corrupt edges and what all more. It's often like everyone with a camera being a photographer or everyone that knows how to push buttons in dreamweaver calling themselves web designer. As long as Inkscape doesn't really support CMYK, Hexachrome, Spot colours and the like, it can never be considered a truly professional tool for print. SVG supports named colours that might be used to simulate spot colours, but it is a rather small palette and the SVG standard just doesn't support anything but sRGB really. Pantone will not give a license to Inkscape for free I gather as they make a living on this and a good one. Color support can only be done using Inkscape namespaces with the sRGB equivalent in the style or attributes next to it and is dependent on output devises as well. Would be a hell of a lot of work to figure out even a few color profiles and quite some wasted resources for the one who does that. Maybe some university might be interested in doing that or some very progressive printer with deep pockets? The latter at this moment not available in Europe at least or will be hard to find. It's a massacre at the moment. If I would like to see anything in SVG 2.0 it would be arbitrary color spaces. Some way to define a color and it's sRGB equivalent in XML and link that up to the SVG file, like is being done with @fonts. @colorProfile or so. I convert the colors in Coreldraw from a Inkscape PDF export. The only reason I still use it and am bound to Windows. My 2 Feng, Jelle

3 4

BUG #1060837
by Arshdeep Singh 05 Mar '13

05 Mar '13

Related to bug: https://bugs.launchpad.net/inkscape/+bug/1060837 How to re-create the scenario being talked about in the bug report. All I get their is A(Alpha) value for my inkscape. The bug mentions the word 'ink'. I am guessing it refers to matching the printed colours and the colors on your monitor, as its a matter under the CMS (Color Management System). As an artist, I haven't worked on very high profile projects such that minor color chnages would make a huge impact so I am totally oblivious to the whole CMS thing. -- demicoder

1 2

GTK Adjustment
by Arshdeep Singh 05 Mar '13

05 Mar '13

What are the page increments and page size arguments in: http://developer.gnome.org/gtk3/3.5/GtkAdjustment.html#gtk-adjustment-new Thanks -- demicoder

1 1

Help with CubicBezier in a closed Geom::Path::const_iterator
by Jabiertxo Arraiza Cenoz 04 Mar '13

04 Mar '13

Hi. I have a problem. Its this: T get the last subpath of a Geom::Path::const_iterator whith a this options: end_open(), end(), end_default(), end_closed() and this subpath is not degenerated at the end. But never get a CubicBezier from the last subpath. Anybody can help me? Thanks in advance.

2 5