- Inkscape Devel - Inkscape Mailman

Failure to load image in non-interactive mode.
by Tavmjong Bah 06 Feb '08

06 Feb '08

Hi, The Image filter works for external images interactively but fails when the command line is used to convert an svg file to a png file. It fails with the following error messages: (inkscape:26505): glibmm-CRITICAL **: Glib::ObjectBase*<unnamed>::create_new_wrapper(GObject*): assertion `wrap_func_table != 0' failed (inkscape:26505): glibmm-WARNING **: failed to wrap type of 'GdkPixbuf' The line in nr-filter-image.ccp that is generating the error message is: image = Gdk::Pixbuf::create_from_file(fullname); It doesn't throw a Glib::FileError or a Gdk::PixbufError. Does anybody have a clue as to what difference between running interactively and using the command line could be the source of this problem? Tav

1 0

Windows Inkscape icon is Crunched
by Joel Holdsworth 06 Feb '08

06 Feb '08

I just installed the Inkscape beta, and I notice a slight problem with the application icon. Notice in the attached screenshot of the start menu, that this 32x32 icon is aliased - I suspect because it's using a 48x48 icon. When was the icon changed, and why? Apart from the aliasing, I can't see any difference in design. Joel

3 2

Making more documentation faster with wiki
by unknown＠example.com 06 Feb '08

06 Feb '08

The documentation is actually 100% docbook based. Docbook are converted to SVG for integrated tutorials if I'm not wrong. The problem is docbook is not the easiest way to make a doc for everyone, and there are lot of people that are not developers and would like to help for some other tasks (as we see lot of good free tutorials around the web and translator in lot of languages). Actually the manuals are here in Inkscape SVN, there is an English and a French beginning of doc : http://inkscape.svn.sourceforge.net/viewvc/inkscape/user_manual/trunk/xml/ The doc didn't move since one month, Elisa-Yemanja (that already amade a paper book manual on Inkscape) contributed in French version 2 weeks ago, but think this is too long process to manage docbook. I think it will be better to use a well defined wiki for manual, as lot of contributors could participate easly as wikimedia projects and some other projects shown. There are already started wikibooks for inkscape on wikibooks.org. I believe there are 2 conditions for this project to be used: * A well defined zone for the wikimanual doc (like doc.inkscape.org, manual.inkscape.org or inkscape.org/doc), actually user have to dig in wiki to find part of user information. * Tools to convert automatically this wiki to docbook. I then searched tools to convert from wiki to docbook and from docbook to wiki, and there are several ones: to integrate docbook to wiki (and don't restart the whole documantation) there are scripts made by blender.org (or at least referenced on there wiki) in perl and python : http://wiki.blender.org/index.php/Meta/DocBook_to_Wiki mediawiki project itself contain a php wiki2docbook, available in the mediawiki svn (and archive, I suppose): http://tools.wikimedia.de/~magnus/wiki2xml/w2x.php Example of conversion to odt et pdf are nice they even contain link to other pages of the manual on the de.wikipedia.org used as example: http://www.nabble.com/Wiki-Konverter-to3627646.html I suppose we could output to SVG from docbook and change to local disk pages or other site obtaining then a full doc separated distribution (as there is gimp + gimp-doc, python + python-doc, ruby + ruby-doc, ...) or choose to make links to generated svg doc on inkscape site or both, to have well done manual for users with interactive features as actual tutorials. I think tutorial is good as is light, allowing a quicktour to Inkscape, the doc should be another documentation separated from inkscape itself, at least in a first time. There are other wiki2XML in Perl (including docbook converter) here: http://search.cpan.org/search?query=wikiformat&mode=all General wiki format, Text::Wikiformat : http://search.cpan.org/~chromatic/Text-WikiFormat-0.79/lib/Text/WikiFormat.… And specialized mediawiki format Text::MediawikiFormat: http://search.cpan.org/author/DPRICE/Text-MediawikiFormat-0.05/lib/Text/Med… CGI::Kwiki and Pod::Simple::Wiki are other Perl parser. There is to dig here for Python parsers/converters: http://pypi.python.org/pypi?%3Aaction=search&term=wiki+docbook&submit=search A Ruby converter from wiki to docbook (and other) : http://rubyforge.org/projects/deplategem/ Ruby wiki parsers : http://rubyforge.org/search/?type_of_search=soft&words=wiki+docbook&Search=… Regards, Popolon

6 6

Re: [Inkscape-devel] [Inkscape-user] fonts disappear in OSX
by jiho 05 Feb '08

05 Feb '08

On 2008-February-05 , at 19:37 , churro wrote: > This is interesting, I have not upgraded office since I started > working on > this in October. The last time it worked fine was last week, before I > upgraded Inkscape. I don't think I changed anything else. I also > found this > font in MSOffice folders on my Mac, so it is an MS font... I'll > convert it > to ttf and see if that works. I can't think of anything else I > changed. I > guess I must have a whole bunch of these fonts from MS, is there a > way to > batch convert them or something like that so I can use them in > Inkscape? > Thanks It has been a frequent request associated with all these font issues. While it does not solve the underlying problem, it may ease the font conversion: I've finally put up a small app based on a script of mine which did this. Drop one or several font files on it and it will convert all of them to TTF, in the same directory. http://jo.irisson.free.fr/dropbox/Font2TTF.zip It uses Fondu[1] to convert Font Suitcases (wether they are dfont or the old Font suitcase format) and FontForge[2] for other types (otf, pfb currently). Fondu is included for 10.5 users (on Intel but I guess there are few 10.5 users on PPC). Fontforge needs to be installed (either via the link below or through MacPorts/Fink. I used MacPorts, it works fine). WARNINGS: - this may just-not-workTM - it overwrites ttf fonts of the same name already present (this is to allow to reconvert fonts if you already did it, without having to manually deleting all of them). - this duplicates the fonts already present on the system. FontBook will put a dot next to their name. You'll probably need to deactivate a version so that OS X behaves sensibly. - if you convert system fonts, you'll not be able to deactivate them, hence you'll need to live with the duplicates - duplicated font can cause rendering problems on Firefox (last time I tried anyway) ADVICE: - copy (not move, copy. this is achieved by holding down the Alt key when dropping the files) the fonts from your libraries (/Library/ Font, /Users/yourname/Library/Font) to somewhere (a folder on your desktop is fine), then select the fonts in this new location and convert them there, eventually select all the newly created TTFs and put them back in a font library folder (or double click each of them to install them) - if you want these TTF versions to be available to X11 apps only and to continue to use the rest idependently, move them to an X11 only location. ~/.fonts for example (the .fonts folder in your home directory). Do to this, in Terminal: mkdir ~/.fonts cd the-folder-where-your-ttf-fonts-are mv *.ttf ~/.fonts - manually rebuild the font cache, just to be sure. In Terminal: sudo fc-cache Let me know if it works/helps/screws everything up. Cheers, [1] http://fondu.sourceforge.net/ [2] http://sourceforge.net/project/showfiles.php?group_id=103338&package_id=111… JiHO --- http://jo.irisson.free.fr/

1 0

Inkscape0802051748: was there and upgrade of GTK?
by unknown＠example.com 05 Feb '08

05 Feb '08

Hi, I have a question about today's Inkscape0802051748: was there an upgrade of the GTK? I'm getting a strange behaviour of the Inkscape window in Windows Vista. When I click on the taskbar to minimize/maximize the Inkscape window, it keep reapearing and never minimizes. Molumen

1 0

Re: [Inkscape-devel] [Inkscape-user] fonts disappear in OSX
by Gail Carmichael 05 Feb '08

05 Feb '08

jiho wrote: > On 2008-February-05 , at 19:12 , churro wrote: >>> - select the faulty text, open the xml editor and check what the font >>> attributes are (name, weight etc.) >> these are the attributes: I'm using Culrz MT >> font-size:98.06591034px;font-style:normal;font-weight:normal;fill:#ff8080;fill-opacity:1;stroke:none;stroke-width:1px;stroke-linecap:butt;stroke-linejoin:miter;stroke-opacity:1;font-family:Curlz >> >> MT > > This is a (Microsoft I think) Font Suitcase format, which indeed > causes issues. The strange thing is that it worked in an earlier > version of Inkscape. I am guessing that the difference actually comes > from the font itself. When did you start this piece of work? Did you > upgrade your system or MS Office in between? I've never even heard of these suitcase things (I assume you figured out the format from the file itself and not the above properties). Hopefully this means that the code I did has not adversely affected the situation. I wrote it so that it looks for my new information first, then falls back onto the old tried tested and true way. > >>> - check wether you see the font in the text and font palette (even >>> if it's >> not used in the document)? >> >> The font does not show up in the font palette >> >>> If you get a ttf font in a >>> regular format (ttf extension) Inkscape should be able to see it. >>> Inkscape has issues with fonts in the dfont format and with Microsoft >>> Font Suitcases (no visible extension). If you install the font on your >>> Mac the regular way (double click and click install), Inkscape should >>> have acces to it. To force the refresh of the font library on startup, >>> delete a file via a command in Terminal: >>> rm ~/.inkscape/.fccache-new >>> Inkscape will rebuild the font cache on next startup. >> >> I tried forcing the font library refresh on both versions of Inkscape, >> original dmg and the new one, to no avail. I searched for the font >> Curlz >> under finder and it shows up in the Fonts directory... I noticed >> there are >> several other fonts that are in the fonts directory but dont show up in >> Inkscape. > > OK, that's to be expected with font suitcases unfortunately. They can > easily be converted to their ttf elemental form though. See the FAQ > about that. > > Please send this the needed info to Gail too, she's the font expert. Unfortunately, I'm not familiar with this suitcase thing you mentioned. Normally I would be able to look it up and learn about it etc, but unfortunately I have a heavy course-load for the semester and probably don't have the necessary time to devote to it. I am cc'ing the devel list in case anyone else has the knowledge to have a look. Still, we should try to determine for sure if it could be an issue with my code, and I can (hopefully) fix that if it is. Gail

1 0

Re: [Inkscape-devel] 1.6: a8-mask: Making non-aligned strides illegal
by Carl Worth 05 Feb '08

05 Feb '08

On Mon, 28 Jan 2008 02:17:20 -0400, "bulia byak" wrote: > > The inkscape folks had run into this problem and I had said we'd have > > a fix in 1.6. Instead, I'd like to volunteer to help fix the inkscape > > code to allocate buffers with a properly aligned stride. If some kind > > inkscape developer were willing to help me with this soon, that would > > be great. > > it was so long ago I completely forgot the context :( > > can you please remind me what the problem was, specifically? The problem was in how inkscape allocates buffers before using them as an A8 mask. For example, imagine you want to create a 7x7 pixel mask. Apparently the current inkscape code does something like: width = 7; height = 7; data = malloc (width * height); The problem would then come in trying to use this data as a cairo image surface: stride = 7; surface = cairo_image_surface_create_for_data (data, CAIRO_FORMAT_A8, width, height, stride); In previous releases of cairo, using a stride value like this that is not a multiple of 4 would cause cairo to draw total garbage. Instead, now, as of cairo 1.5.8, the above would result in an error surface with a status of CAIRO_STATUS_INVALID_STRIDE. The fixed could would instead look like this: width = 7; height = 7; stride = cairo_format_stride_for_width (CAIRO_FORMAT_A8, width); data = malloc (stride * width); surface = cairo_image_surface_create_for_data (data, CAIRO_FORMAT_A8, width, height, stride); Here, the cairo_format_stride_for_width call takes care of any necessary alignment, (which in the current version would return a value of 8 for the stride). Obviously, that much of a change is trivial. Beyond that, we'll need to then fixup as much inkscape code as necessary so that all of inkscape's own rendering code properly accounts for the stride not being identical to the width. Make sense? Again, I'm glad to help out with this. Just let me know what I can do. -Carl

1 0

New sp-object question
by Michael Yinger 05 Feb '08

05 Feb '08

Hi, I'm adding a new sp-object; a serial number. This allows the user to drop a serial number on the doc and define the start number, increment, and number of numbers and then the downstream imposition software takes over from there. I've used the sp-skeleton templates and ended up with: sp-serialnumber.h:33: error: expected initializer before 'sp_serialnumber_get_type' sp-object-repr.cpp: In function 'void populate_dtables()': sp-object-repr.cpp:185: error: 'sp_serialnumber_get_type' was not declared in this scope sp-object-repr.cpp:185: error: '"svg:serialnumber"' cannot be used as a function sp-object-repr.cpp:187: error: uninitialized const member 'populate_dtables()::NameTypeEntry::type_id' sp-object-repr.cpp:187: warning: missing initializer for member 'populate_dtables()::NameTypeEntry::type_id' I'm used to working in VisualStudio so I'm getting up and over a lot of learning curves running everything from the command line and a text editor, so forgive me if this is a brain-dead simple question. As I looked at this I started to wonder if one of the problems is that this will introduce a non-standard SVG element. Here is what I've been using. <SerialNumber m_position.left="180" m_position.right="215" m_position.top="-122" m_position.bottom="-97" m_bPen="TRUE" lopnColor="0" lopnWidth="1" m_bBrush="TRUE" lbStyle="BS_SOLID" lbColor="12632256" lbHatch="HS_HORIZONTAL" m_nAngle="0" m_lFont.lfFaceName="Lucida Sans Unicode" m_lFont.lfHeight="12.000000" m_lFont.lfWeight="0" m_lFont.lfItalic="0" m_color="0" m_nLineSpacing="0" m_borderStyle="None" m_horizAlign="Center" m_vertAlign="Center" m_szText="001" m_nIncrement="1" m_nStartNumber="1" m_nPatternLength="3" m_szPrefix="" m_szPostfix="" m_bPrinted="0" m_nLastNumberPrinted="0" m_bDecrement="FALSE"/> Thanks for any help or links to help. -- Mike Yinger www.ticketprinting.com

3 4

Re: [Inkscape-devel] Startup crash on OS X
by John Faith 05 Feb '08

05 Feb '08

On Feb 3, 2008 5:39 AM, Niko Kiirala <niko@...1267...> wrote: > Sat, 2 Feb 2008 09:11:51 -0800 > "John Faith" <jfaith7@...400...> kirjoitti: > > > I found that this small change prevents a startup crash on OS X 10.3.9 > > (also added to https://bugs.launchpad.net/inkscape/+bug/185689): > > > > --- inkscape.trunk/src/sp-filter.cpp.orig Sat Dec 22 12:15:36 2007 > > +++ inkscape.trunk/src/sp-filter.cpp.noStartupCrash Sat Feb 2 > > 08:55:03 2008 @@ -110,7 +110,7 @@ sp_filter_init(SPFilter *filter) > > > > filter->_renderer = NULL; > > > > - filter->_image_name = map<gchar *, int, ltstr>(); > > + filter->_image_name.clear(); > > > > filter->filterRes = NumberOptNumber(); > > > > > > While I'm happy that Inkscape can actually run, it's not obvious to me > > why this fixes the crash, nor if this is generaly an acceptable thing > > to do. Could someone possibly test that this does not cause a problem > > on another platform? > > > > By the way, after applying this patch, I have another crash in > > sp-filter.cpp when opening "Help->About Inkscape", so I wonder if > > there's either a std::map problem with gcc 3.3 or memory is getting > > trashed. Random theories welcome :-). > > That was one of the most annoying bugs I've ever had to debug. SPFilter > is not an actual C++ class, so it's not constructed. This means, that > filter->_image_name is not construced either. Before that line you > changed, filter->_image_name has rubbish state and that line is trying > to set it to sensible state. > > So... That line should just call the constructor for > filter->_image_name. I can see, how my original version may be compiled > as 'construct a temporary map and use operator= to copy it to > filter->_image_name'. > Hi Niko, Thanks for the reply. I wonder how the filter struct memory is allocated? As an experiment, I added 2 vars in sp_filter_init(): static SPFilter* filterNew = new SPFilter(); static SPFilter* filterAlloc = (SPFilter*)malloc(sizeof(SPFilter)); , and in gdb saw: (gdb) print filter->_image_name._M_t._M_header $7 = (_Rb_tree_node<std::pair<char* const, int> > *) 0x0 (gdb) print filterNew->_image_name._M_t._M_header $9 = (_Rb_tree_node<std::pair<char* const, int> > *) 0x71a1fe0 (gdb) print filterAlloc->_image_name._M_t._M_header $8 = (_Rb_tree_node<std::pair<char* const, int> > *) 0x0 , which leads me to believe that the filter->_image_name member is not being initialized correctly. I found that changing the SPFilter struct's _image_name from a std::map to a pointer to a std::map and using 'filter->_image_name = new std::map<gchar *, int, ltstr>;' in sp_filter_init() solves my crashing problems; I can now startup and view the About dialog. , John

3 4

simpletransform.py
by Alexandre Prokoudine 05 Feb '08

05 Feb '08

Hi, We seem to have a simpletransform.py extension that isn't installed by default (bug 187838) and has no accompanying .inx. Is it of any use? Why do we have it at all? Alexandre

4 5