March 2008 - Inkscape Devel - Inkscape Mailman

Re v 17995 - build broken again...
by Krzysztof Kosiński 23 Mar '08

23 Mar '08

The Linux build is broken again due to link errors. There are so many of them that I'm not sure I'm able to sort this mess out by myself. Most of the undefined references are to things in libspwidgets. Can anyone help? Regards, Krzysztof (Chris) Kosiński -- View this message in context: http://www.nabble.com/Rev-17995---build-broken-again...-tp16238429p16238429… Sent from the Inkscape - Dev mailing list archive at Nabble.com.

3 4

Re v 17995 gettext failure - patch for LP bug 170185
by Krzysztof Kosiński 23 Mar '08

23 Mar '08

There are 2 files which attempt to translate empty strings, which is wrong (this msgid gives the translation header - see https://bugs.launchpad.net/inkscape/+bug/170185 for details) Index: ../src/display/canvas-axonomgrid.cpp =================================================================== --- ../src/display/canvas-axonomgrid.cpp (revision 17998) +++ ../src/display/canvas-axonomgrid.cpp (working copy) @@ -411,7 +411,7 @@ "empcolor", "empopacity", _wr, repr, doc)); Inkscape::UI::Widget::RegisteredSuffixedInteger *_rsi = Gtk::manage( new Inkscape::UI::Widget::RegisteredSuffixedInteger( - _("_Major grid line every:"), _(""), _("lines"), "empspacing", _wr, repr, doc ) ); + _("_Major grid line every:"), "", _("lines"), "empspacing", _wr, repr, doc ) ); _rsu_ox->setDigits(4); _rsu_ox->setIncrements(0.1, 1.0); Index: ../src/display/canvas-grid.cpp =================================================================== --- ../src/display/canvas-grid.cpp (revision 17998) +++ ../src/display/canvas-grid.cpp (working copy) @@ -675,7 +675,7 @@ _wr, repr, doc)); Inkscape::UI::Widget::RegisteredSuffixedInteger *_rsi = Gtk::manage( new Inkscape::UI::Widget::RegisteredSuffixedInteger( - _("_Major grid line every:"), _(""), _("lines"), "empspacing", _wr, repr, doc) ); + _("_Major grid line every:"), "", _("lines"), "empspacing", _wr, repr, doc) ); table->set_spacings(2); Regards, Krzysztof (Chris) Kosiński -- View this message in context: http://www.nabble.com/Rev-17995-gettext-failure---patch-for-LP-bug-170185-t… Sent from the Inkscape - Dev mailing list archive at Nabble.com.

1 1

SOC 2008 - project - Transformations Anchors
by Vangelis Katsikaros 23 Mar '08

23 Mar '08

Hello After a brief discussion on the list I did a small research for the Transformation Anchors idea. So, I 'd like to ask about the validity of the things I found. * SVG 1.1 has a "transform" attribute which is used for some of the transformations. Inkscape stores the transformation in the file as following: * "move", "scale" doesn't apply a trans matrix but simply moves, scales the coordinates. It also adds a second inkscape:perspective element. This one include a property "persp3d". This is related to the src/persp3d.cpp and the Persp3D class. * "rotate", "skew" applies a transformation matrix with the SVG:transform attribute. * "matrix" depends on the matrix's values. It could lead to one of the above situations. The Transformation UI is implemented in ui/dialogs/transformations.cpp * sp_desktop_selection() is used in order to get the things the user has selected from the "drawing area". (I don't know if you have a specific name for the area on which the inkscape user draws and selects graphics from. I called it "drawing area"). * The functions Transformation::applyPage(Move|Rotate|Scale|...) are responsible for the transformation of the selection. I then tried to figure out how "move" transformation works: updatePageMove() takes the bounding box from the selection. This is a NR::Rect object (NR::Maybe template). NR stands for "new renderer" right? I follow up to here but then I get a bit confused in this line: double x = bbox->min()[NR::X]; NR::Rect::min() returns a member of the NR::Rect class. It returns a Point instance. This class has overloaded the [] operators. With this overloading we return the [i]-th element of a Coord member. And this Coord is an array with 2 elements (for x and y). I think I think I lost the logic of it a bit... Then we take the value which updatePageMove() had set and we apply it with applyPageMove(). This calls sp_selection_move_relative() and then sp_selection_apply_affine(). That's as far as I have seen. So, it boils down to pass sp_selection_apply_affine() an argument about the center of the tranformation and change things in there so that the argument is used correctly? Sorry for the lengthy mail. I just want to make sure that I get how things work in order to estimate the size of the project and the extend of changes in the base code. Regards Vangelis Katsikaros __________________________________________________ ×ñçóéìïðïéåßôå Yahoo!; ÂáñåèÞêáôå ôá åíï÷ëçôéêÜ ìçíýìáôá (spam); Ôï Yahoo! Mail äéáèÝôåé ôçí êáëýôåñç äõíáôÞ ðñïóôáóßá êáôÜ ôùí åíï÷ëçôéêþí ìçíõìÜôùí http://mail.yahoo.gr

2 1

GSoC Idea
by Afief Halumi 23 Mar '08

23 Mar '08

Hello everyone, I would like to submit this idea as my summer of code project and of course any feedback/piece of mind from the inkscape community is greatly appreciated. The Live Path Effect Stack seemed to be *the* thing I always wanted to have in my graphic programs. I have used similar technologies(don't know if this is the right word) before in 3D Studio Max and in Blender, but the way LPES works from what I read is far from what I was hoping for. The main functional fault of LPES is that it is a stack in my opinion. Consider this example from the LPE-bluprint wiki page: http://wiki.inkscape.org/wiki/index.php/Image:Lpe_need_rich_input.png The Live Path Effect would need two objects, the cloud and the sun(let's forget for a minute that each of those is itself an object with an LPE applied,) but how can this be done in a stack environment? The obvious way would be to apply the LPE to one object and add a reference(pointer/id...) to the other object as a parameter. This way would most certainly work, but the second object would remain somewhere outside the object with the LPES applied to it, which will make working with a lot of LPES awkward. How I would like to do it: Instead of using a stack I would like to have an LPE Tree. Each tree node can be either an effect or an object. This way each object can contain a whole hierarchy of editable effects(Similar to Blender and Maya's Node System, but more simple). This would essentially make on-canvas editing of any part of the object(the outline of the cloud, the spike or the circle of the sun) more intuitive as the object itself would contain all the information necessary to build them. This is a (very) simple mockup for the interface as I imagine it(lots of elements taken from 3d studio max) http://img187.imageshack.us/img187/5057/mockuprv7.png Wow, this turned out longer than I thought, thank you for reading through it. Afief Halumi

3 5

Inkscape 0.46
by Bryce Harrington 23 Mar '08

23 Mar '08

The Inkscape 0.46 release has been tagged and a tarball is in the process of being created. The next step will be packaging and finalizing PR materials. Thank you everyone for your hard work in producing this release. We will be embargoing PR announcements until March 24th, to provide the Windows port with additional time to identify fixes to some critical bugs, so please hold off from publicizing this release for now. Bryce

7 7

NEW: visual bbox enlarged for filters
by bulia byak 23 Mar '08

23 Mar '08

The visual bounding box (which is the default bounding box type used by Inkscape) of an object with a filter applied now includes the expanded area of the filter. For single blur filter (such as the blur you apply with a slider in the Fill and Stroke dialog), this expands the bounding box by 2.4*radius; although theoretically, blur is infinite, this is the distance at which the opacity of the object drops below the perceptibility threshold of our renderer. For all other filters, the area is expanded by the relative amounts you specify on the "Filter general settings" tab of the Filter Effects dialog. Only visual bounding box is affected; if you use geometric bounding box, you will notice no change in most cases. However, the Export bitmap dialog always uses the visual bbox for selection export area; this means that you can now export a blurred object to bitmap without any clipping of the blur. -- bulia byak Inkscape. Draw Freely. http://www.inkscape.org

1 0

GSoC '08 proposal
by Miklós Erdélyi 22 Mar '08

22 Mar '08

Hi all, In the last two years I have participated in GSoC working on Inkscape's PDF export/import features. This year I would like to participate again, wrapping up these :) My project proposal would be based around the previous two summer's work, mainly improving the PDF import and export extension. Regarding PDF import I propose: - to establish an embedded font framework, - to implement CMYK support as soon as we decide how we're going to need it :) - to fix bugs related to the PDF import extension (free time activity?). The embedded font framework for the PDF importer would allow importing of text as either text, either path or as text with embedded SVG fonts. Using embedded SVG fonts could join Juca's SoC project of implementing this crucial missing feature in Inkscape. Some of the design work of this framework has already been done. Regarding PDF export I propose - to incorporate Cairo's user font support (soon to be released) to export SVG fonts, - to optimize the renderer for the now-improved Cairo PDF backend, - to fix various stuff related to not passing the SVG 1.1 test suite either due to missing stuff in Inkscape (like not handling the preserveAspectRatio attribute for svg:pattern and other elements) or due to bad implementation of the Cairo-renderer for PDF export, - and to fix bugs in LP in connection with PDF export. What do you think about these? Which ones are worth a SoC project? I would appreciate any suggestions :) Best regards, miklos

3 3

lib2geom integration - team effort possible?
by unknown＠example.com 22 Mar '08

22 Mar '08

Hello all, it seems there is quite a bit of interest to work on lib2geom integration. If I count well, there are 4 (Daniel Kraft, mgsloan, Marco Cecchetti and me). How much work do people think it will be? Perhaps we can use this many people working on it, but maybe it's hard to split it up in 4 well defined areas. On the other hand, with 4 people the chances of completely switching to 2geom are greatly increased. Searching for NR::Point gives 3000+ hits and NR::Matrix 1000+ hits... (only a few hits are from libnr). So just replacing libnr types with 2geom types is 4000 places that need to be checked; most of them just need a short glance I guess, but still. I like the team approach a lot, but don't know how to split up the work. Njh mentions that perhaps it is inefficient to initially split things up. Regards, Johan

1 0

Bugfix Patch for the Print Dialog
by joel＠...1709... 22 Mar '08

22 Mar '08

Hi All, I've just fixed a very minor problem with the print dialog which allowed it to fall underneath the inkscape desktop window in Win32. I've already applied the change to trunk, but I'm just wondering should this fix be applied to the 0.46.X series of branches? If so, can someone do that for me. Thanks Joel

3 4

Toolbar problem: action missing
by unknown＠example.com 22 Mar '08

22 Mar '08

I had troubles with the toolbar, so I decided to get a clean svn checkout, get the latest devlibs and do a clean build. Still I get the following warnings when starting Inkscape: (inkscape.exe:43288): Gtk-WARNING **: ZoomIn: missing action ZoomIn (inkscape.exe:43288): Gtk-WARNING **: ZoomPage: missing action ZoomPage (inkscape.exe:43288): Gtk-WARNING **: ZoomNext: missing action ZoomNext (inkscape.exe:43288): Gtk-WARNING **: ZoomPage: missing action ZoomPage (inkscape.exe:43288): Gtk-WARNING **: DialogFillStroke: missing action DialogFillStroke (inkscape.exe:43288): Gtk-WARNING **: DialogText: missing action DialogText (inkscape.exe:43288): Gtk-WARNING **: DialogAlignDistribute: missing action DialogAlignDistribute And the toolbar is not populated with all buttons (it stops after "zoom to fit drawing in windows"). I downloaded the latest build from inkscape.modevia.com which does not show these warnings. Does anybody have an idea what is wrong on my end? Thanks alot, Johan

3 6