- Inkscape Devel - Inkscape Mailman

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