On Sat, 2004-06-19 at 18:04, Florent wrote:
> So I'm appealing to anyone who can create an autosave :-)
We already have autosave-on-crash already, actually. The one catch is
that for crashes due to failed assertions (like this one), it isn't
(I'm CCing the development mailing list, if anyone's interested in
attacking this right now:
At least on Unix, this is relatively easy to fix -- just set up a
SIGABRT handler similar to the SIGSEGV handler we have now. The one
catch is that the handler will need to disable itself as soon as it is
triggered, because the autosave code raises SIGABRT when it is done.
Also we should make sure we've got SIGBUS covered if we don't already.)
I have the latest debian package, but dunno what that is... can't you
put something in the about dialog, so that I know... I never remember
this terminal stuff.
Anyways, I get a crash when I try to copy a text with a gradient fill.
That is strange on the one hand and annoying on the other. I hope it's
an easy to fix or fixed already.
just wondering if I should file a bug, or if it's so simple that I don't
Especially in the guides and the positioning dialog it would be helpful
to be able to put mathimatical stuff in there. Imagine you have a guide
at 45.6 and want one 7 mm further to the right, you'd just type 45.6+7
and there you go... even more important is something like dividing by 2
or 3, I need that quite often.
Btw: Since my laptop was for repair, I used CorelDraw, and I think for
basic vector editing and layouting inkscape is by now better than Corel.
That is something!
Which also brings me to one other thing I'm missing: Baseline Grid. Even
for simple flyers and brochures it's nice (or rather necessary) to have.
CorelDraw just pissed me of for not having it!
I got my laptop back, with working ethernet card, Yeah!! Take care!
People keep complaining about the unsmooth freehand drawing in
Inkscape and point out Sodipodi's smooth drawing as its advantage over
http://linuxfr.org/2004/07/20/16864.html (see comment "Sodipodi
meilleur que Inkscape ?")
I remember this problem was being attacked quite some time ago, and I
was under an impression that the solution was near. Peter and Nathan
were working on it, as well as the person known as paraprax on jabber.
Can someone please update me on the status of that work? How far has
it progressed and what is needed to finish it? I would really love to
see this issue closed for the next version.
I'm cc'ing my reply to inkscape-devel, where more people will be able
On Sun, 25 Jul 2004 20:30:52 +0300, Anatoly Podlesnuk <beerubeer@...466...> wrote:
> It's good idea to join spalah and inkscape.
> There is some questions.
> 1. I'm not use in spalah "items tree" of SVG nodes from sodipodi. Spalah
> items tree is much simple. So I will need some help with inkscape items
We haven't yet departed too far from Sodipodi's system of SPObjects
tree and repr tree that are synced periodically, though we did a lot
of incremental changes to it (mostly in the direction C++fying it). I
don't think anything simpler than that will actually work for
interactive SVG editing. Of course we will be glad to help you with
> 2. Spalah uses ming library (ming.sourceforge.net) for flash generation,
> so inkscape will require this library.
No problem. SWF export is a good thing to have, even if for static
SVGs at first.
> 3. What is role spalah code in inkscape? It will a part of ExportDialog
> or inkscape will able to add animation to SVG objects? As I know
> inkscape does not support some SVG nodes that responses for animation.
Our goal is to be a complete SVG editor, and that includes animation.
So we'll need to add support for all the corresponding elements and
attributes. As for the UI, what I had in mind is a timeline control
selecting one of the frames for editing, plus a separate preview
window (based on inkview?) with actual animation. Of course, at this
stage all this requires lots of thinking, planning, and experimenting.
> 4. What about release planning, code formatting,
> classes/functions/variables names?
I guess the main development principle is that CVS must always remain
compilable and workable, without bad regressions. Releases are done
when there's enough stuff for a new release, usually every 2 to 3
months. There's a period of code freeze and bug hunt before a release,
but at other times you're free to work on your stuff. For coding
style, we have a few guidelines here:
Let me know what you think.
First off, I wanted to say thank you to Bob for this work. The preview
on open is fabulous! It gives a whole new perspective on my files. A
very important addition.
Still there are problems. I know some of them may not be fixable due
to GTK limitations, but still.
1 The dialog is modal but not transient. It's a very dangerous
combination: it sinks under your document and you can't understand why
the document is unresponsive. Please make it transient as it was
2 The dialog opens in some unpredictable place on the screen. Before,
it opened centered over the document window. Can we restore that?
(Maybe it's a consequence of its not being transient.)
3 BUG: when I press Ctrl+O for the first time, there are no files
listed at all in the directory where it opens, only subdirectories. If
I go e.g. to the parent dir and then back, the files appear.
4 After some thinking I'm convinced that (as suggested here) we must
replace the "Autodetect" with two new options, "All images" and "All
files". That makes much more sense and is more consistent with typical
Windows/Mac programs. Both these options would do autodetect and
differ in filtering the list: no filtering for "All files" and showing
only Inkscape-supported images for "All images". The bad thing about
"Autodetect" is that it speaks about what the program does, not about
what the user needs. Autodetect is not something I ever want to be
off; it must always work behind the scenes. What I want to control is
what files are shown and what not, and this is what this menu must
5 The menu should be reordered more logically so that the most
frequently used options are at the top:
All images <-- default
rest of bitmap formats...
6 Also it would be nice to group the same formats with different
extensions as proposed here:
7 If I want to go e.g. to /home/me/inkscape and browse files here, I
go to the filename field, type "/home/me/inkscape" and press Enter.
Nothing happens. In other programs (and I believe in Inkscape with the
old dialog) this caused the dialog to switch to that dir.
8 I get these warnings whenever I open it for the first time:
(inkscape:7505): Gtk-WARNING **: Could not find the icon
'gnome-fs-home'. The 'hicolor' theme
was not found either, perhaps you need to install it.
You can get a copy from:
(inkscape:7505): Gtk-CRITICAL **: file gtksettings.c: line 447
(gtk_settings_get_for_screen): assertion `GDK_IS_SCREEN (screen)'
(inkscape:7505): GLib-GObject-WARNING **: invalid (NULL) pointer instance
(inkscape:7505): GLib-GObject-CRITICAL **: file gsignal.c: line 1726
(g_signal_handler_disconnect): assertion `G_TYPE_CHECK_INSTANCE
I installed http://freedesktop.org/Software/icon-theme/releases/hicolor-icon-theme-0....
but this did not help. Any ideas?
9 There's no way to turn the previews off.
10 The open/close dance of the preview pane as you go through
image/non-image files is a bit disconcerting. Can't we always have the
preview widget on (that is, unless you turn it off, see 8), even if
inactive for the current file, with the same unchanging size?
11 Bob, you mentioned that you did preview for the bitmap files as
well, but this does not work for me, only SVGs.
That's it for now, perhaps I'll find more when I test it more thoroughly.
A long story short. AI7 files aren't in PDF format and AI10 uses PDF 1.4
as native saving format. I Have also heared that Photoshop CS uses PDF 1.4
as saving format. So i looked in pstoedit program and finded out (Author
pointed out) there is GPL svg output plugin.
So it would be neat to have this option in inkscape but is it too much to
ask for http://www.gnu.org/software/plotutils/ and pstoedit be installed
to have this input??
There is also one problem in SVG because it doesn't put fill:none in style
inkscape just fills everything in black:P..
I still look how good plot-svg output is but seems good enouhg to me..
Then i just have to think how to find AI7 and AI10 in ill2svg.pl;)
So should we use it or not? In Gnu/Linux and *BSD (*nix) world it's easy
to install but windoze is biggest problem..
I just read about the new release of the GNOME HIG (Human interface
guidelines) 2 and suddenly I asked myself if Inkscape was taking these
I browsed the mailing list a bit and found some postings mentioning
Thinking a bit about the future and the possibility of Inkscape
becoming part of GNOME, or even if this is not on the roadmap, I think
these guidelines are a "good thing". So, are the developers using these
guidelines as they design the user interface? What do you think about
Lucas Vieites Fariña
Asix Informática <info@...212...>
Tel/Fax: +34 986 54 26 98
Sorry for taking so long to get back to you; emails about compilation
problems will probably get answered quicker on the development mailing
On Thu, 2004-07-22 at 16:05, Florent wrote:
> But maybe, the problem come from the new libs version ?
> Is there anyone who can run the last CVS on a Linux ?
Yes, I've been running it since I did the library conversion.
> And what is his lib versions ?
gtk+-2.0 = 2.4.3
gobject-2.0 = 2.4.2
gmodule-2.0 = 2.4.2
glib-2.0 = 2.4.2
pango = 1.4.0
gtkmm-2.4 = 2.4.4
gdkmm-2.4 = 2.4.4
glibmm-2.4 = 2.4.3
pangomm-1.4 = 2.4.4
> I don't know if it can help but when I run the "ddd inkscape", it gives me :
> Program received signal SIGSEGV, Segmentation fault.
> SPDesktop::_layer_hierarchy_changed(SPObject*, SPObject*, SPDesktop*)
> (top=0x84023d0, bottom=0x84023d0, desktop=0x86bc208) at stl_list.h:397
void SPDesktop::_layer_hierarchy_changed(SPObject *top,
That place for the crash looks a little strange to me. Could you try
compiling without optimization and see if it yields the same backtrace
(and give me some more of the actual backtrace as well..).
The only reason I can think of that it would crash here is if
SPDesktop::_layer_changed_signal were not initialized properly for some