- Inkscape Devel - Inkscape Mailman

one more :)
by bulia byak 18 Mar '04

18 Mar '04

The bug I wrote you about earlier today is fixed, thanks :) Now there's another: https://sourceforge.net/tracker/index.php?func=detail&aid=918458&group_id=9… Besides the scrolling by Ctrl+arrows is now rather uneven - it used to accelerate monotonously but now it moves jerkily. Do you think this is something related to rendering too (e.g. slowdown)? _________________________________________________________________ http://join.msn.com/?pgmarket=en-ca&page=byoa/prem&xAPID=1994&DI=1034&SU=ht…

1 0

dialog delete handlers
by bulia byak 18 Mar '04

18 Mar '04

Peter, in align.cpp and transformation.cpp, in delete signal handlers, you wrote static gboolean sp_transformation_dialog_delete(GtkObject *, GdkEvent *, gpointer data) { if ( data != static_cast<gpointer>(dlg) ) { g_warning("possible bug: dlg differs from passed data"); /* E.g. maybe we shouldn't update the global x,y,w,h, or maybe we should use dlg instead of dead_dlg below. */ } GtkWindow &dead_dlg = *static_cast<GtkWindow *>(data); gtk_window_get_position(&dead_dlg, &x, &y); gtk_window_get_size(&dead_dlg, &w, &h); instead of the previous code which simply used dlg, ignoring data. I don't quite understand the reason for this change. Besides: - Why only these two files? The same code is used in all dialogs. - Attempting to close Inkscape with any of these two dialogs open will crash (bug 917122). So for now I restored the old code to fix the crash, but I'm just curious why you deemed this change necessary. _________________________________________________________________ Add photos to your messages with MSN Premium. Get 2 months FREE* http://join.msn.com/?pgmarket=en-ca&page=byoa/prem&xAPID=1994&DI=1034&SU=ht…

2 1

rendering bug
by bulia byak 17 Mar '04

17 Mar '04

Hi, I think it's your recent caching changes that caused this: http://sourceforge.net/tracker/?atid=604306&group_id=93438&func=browse Can you please have a look? _________________________________________________________________ MSN Premium with Virus Guard and Firewall* from McAfee� Security : 2 months FREE* http://join.msn.com/?pgmarket=en-ca&page=byoa/prem&xAPID=1994&DI=1034&SU=ht…

1 0

853858 patch comments
by bulia byak 17 Mar '04

17 Mar '04

1. - GSList *sl; - sl = g_slist_copy ((GSList *) sp_selection_repr_list (selection)); - sl = g_slist_sort (sl, (GCompareFunc) sp_repr_compare_position); + const GSList *items = sp_selection_item_list (SP_DT_SELECTION (desktop)); I.e. you removed sorting the copied list of items. Why? The result is that the copied items now do not preserve z-order when pasted; they're pasted in the order in which they were selected, which is wrong (try it). And you cannot sort the list returned by sp_selectiion_{repr_}list, only a copy of it, which is what was done in the original code. 2. You only copy gradients applied to the fill, not to the stroke, so gradient-filled stroke is lost when copying between documents. 3. Here you check that the fill is either pattern or gradient, and then, without any further check, cast it into a gradient: + if (style && style->fill.type == SP_PAINT_TYPE_PAINTSERVER) { /* Object has pattern or gradient fill*/ + SPGradient *gradient = SP_GRADIENT (SP_OBJECT_STYLE_FILL_SERVER (items->data)); The result is quite predictable - it crashes when you try to copy an object with pattern fill. (Probably with bitmap fill also.) 4. When you paste copied objects with gradients within the same document, it still adds a copy of the gradient (with different id), although the pasted object still refers to the old gradient. This will lead to an unchecked growth of defs each time you do copy/paste. This should be prevented. E.g. check if the current defs contain an exact copy of the gradient you're about to add (not only same id but also same content), and if so, skip adding it. 5. Same thing when you copy an object with gradient to another document and that document already contains a different gradient with the same id. The result is that the pasted object uses that other document's gradient (different from its original gradient), while the correct gradient is added to the defs with a different id (and thus left unused). Fixing this is more complex: if the document you're pasting to contains a gradient with the same id but different content, you must change the id both of the gradient you're inserting to the defs _and_ the reference in the object style. The difficulty being, to check for identical content, you must check not only the content of the repr, but all other reprs it is linked to, and their links, and so on until the tree is exhausted. _________________________________________________________________ MSN Premium helps eliminate e-mail viruses. Get 2 months FREE* http://join.msn.com/?pgmarket=en-ca&page=byoa/prem&xAPID=1994&DI=1034&SU=ht…

2 1

Re: [Inkscape-devel] inkscape crash on startup
by bulia byak 17 Mar '04

17 Mar '04

> > Even if I try to start it without loading a file. It also > > produces a dialog saying "Inkscape has encountered an internal error and >will > > close now". > > -- > >That's what I meant, when saying on startup... Can you do a backtrace? _________________________________________________________________ Tired of spam? Get advanced junk mail protection with MSN Premium http://join.msn.com/?pgmarket=en-ca&page=byoa/prem&xAPID=1994&DI=1034&SU=ht…

1 0

Re: [Inkscape-devel] inkscape crash on startup
by bulia byak 17 Mar '04

17 Mar '04

>I get precisely the same error, using an "inkscape-current" downloaded and >compiled today. Even if I try to start it without loading a file. It also >produces a dialog saying "Inkscape has encountered an internal error and >will >close now". Please file the bug to the tracker, with all relevant details. Especially useful would be a backtrace. _________________________________________________________________ STOP MORE SPAM with the MSN Premium and get 2 months FREE* http://join.msn.com/?pgmarket=en-ca&page=byoa/prem&xAPID=1994&DI=1034&SU=ht…

1 0

Fileselector
by Ted Gould 17 Mar '04

17 Mar '04

Hey, I was looking through the new fileselector code and found a couple of issues that I think should be worked out. Namely, with the inclusion of input and output formats that are done through extensions. Previously, the file dialog called the extensions code to get a menu that contained all the input/output formats that could be handled, and each one had a key attached to it that specified which extension should be used. I'm not saying that this is a good thing, there is too much GUI code in the extensions system with that solution. But, I still think that the extensions system needs to be queried on the matter, and get that key passed into the save functions. Probably the best thing would be a list of verbs, but I don't know if that is the best option as these aren't actions as much as selections. So, I'm proposing this: GSList * Inkscape::Extension::get_input_list (void); This will return a list of items that are: gchar * name gchar * extension gchar * key And then the GUI code can do what ever it wants with them, so long as it returns the key when it finishes (along with the filename ;) Thoughts? --Ted

3 3

RE: [Inkscape-devel] inkscape crash on startup
by bulia byak 17 Mar '04

17 Mar '04

>Yesterday I tried to open the attached document and it wouldn't, but >cause inkscape to crash. >Since I created it with inkscape I thought, well, get the newest cvs and >it'll be working again. but having installed todays cvs inkscape crashes >on startup without any error messages of use: I can open your attached file with no problem using today's CVS. Except that it warns it can't find the font AF PEPSI and the embedded .tif image is also missing. Maybe it would crash on that .tif or that font, but I can't check that. If you still can't open it, submit a bug and attach all relevant files to it. Crash bugs are usually fixed quickly. _________________________________________________________________ Add photos to your e-mail with MSN Premium. Get 2 months FREE* http://join.msn.com/?pgmarket=en-ca&page=byoa/prem&xAPID=1994&DI=1034&SU=ht…

1 0

my win32 patch does not work, why (splivarot.cpp)
by Adib Taraben 17 Mar '04

17 Mar '04

Hello, I tried to start to fix those (win32) locale problems with the float numbers. In the file splivarot.cpp which is the "simplify path" command I replaced the printf's with g_ascii_formatd. This function is locale independent. Unfortunately it does not solve the problem. Anyhow what is the desired method generating locale undependent number output? Am I on the right way? Thanks, Adib. --- splivarot_org.cpp 2004-03-14 17:09:02.000000000 +0100 +++ splivarot.cpp 2004-03-16 14:57:26.000000000 +0100 @@ -1542,34 +1542,45 @@ if (typ == descr_moveto) { Path::path_descr_moveto* nData=(Path::path_descr_moveto*)(path->descr_data+theD.dStart); - g_string_sprintfa (result, "M %lf %lf ", nData->p[0], nData->p[1]); + gchar cX[33], cY[33]; + g_string_sprintfa (result, "M %s %s ", g_ascii_formatd(cX, 32, "%f", nData->p[0]), g_ascii_formatd(cY, 32, "%f", nData->p[1])); } else if (typ == descr_lineto) { Path::path_descr_lineto* nData=(Path::path_descr_lineto*)(path->descr_data+theD.dStart); - g_string_sprintfa (result, "L %lf %lf ", nData->p[0], nData->p[1]); + gchar cX[33], cY[33]; + g_string_sprintfa (result, "L %s %s ", g_ascii_formatd(cX, 32, "%f", nData->p[0]), g_ascii_formatd(cY, 32, "%f", nData->p[1])); } else if (typ == descr_cubicto) { Path::path_descr_cubicto* nData=(Path::path_descr_cubicto*)(path->descr_data+theD.dStart); float lastX, lastY; + gchar cX1[33], cY1[33], cX2[33], cY2[33], cX[33], cY[33]; { NR::Point tmp = path->PrevPoint (i - 1); lastX=tmp[0]; lastY=tmp[1]; } - g_string_sprintfa (result, "C %lf %lf %lf %lf %lf %lf ", - lastX + nData->stD[0] / 3, - lastY + nData->stD[1] / 3, - nData->p[0] - nData->enD[0] / 3, - nData->p[1] - nData->enD[1] / 3, nData->p[0],nData->p[1]); + g_string_sprintfa (result, "C %s %s %s %s %s %s ", + g_ascii_formatd(cX1, 32, "%f", lastX + nData->stD[0] / 3), + g_ascii_formatd(cY1, 32, "%f", lastY + nData->stD[1] / 3), + g_ascii_formatd(cX2, 32, "%f", nData->p[0] - nData->enD[0] / 3), + g_ascii_formatd(cY2, 32, "%f", nData->p[1] - nData->enD[1] / 3), + g_ascii_formatd(cX, 32, "%f", nData->p[0]), + g_ascii_formatd(cY, 32, "%f", nData->p[1])); } else if (typ == descr_arcto) { Path::path_descr_arcto* nData=(Path::path_descr_arcto*)(path->descr_data+theD.dStart); + gchar cRX[33], cRY[33], cA[33], cSX[33], cSY[33]; // g_string_sprintfa (result, "L %lf %lf ",theD.d.a.x,theD.d.a.y); - g_string_sprintfa (result, "A %g %g %g %i %i %g %g ", nData->rx,nData->ry, nData->angle, - (nData->large) ? 1 : 0,(nData->clockwise) ? 0 : 1, nData->p[NR::X],nData->p[NR::Y]); + g_string_sprintfa (result, "A %s %s %s %i %i %s %s ", + g_ascii_formatd(cRX, 32, "%f", nData->rx), + g_ascii_formatd(cRY, 32, "%f", nData->ry), + g_ascii_formatd(cA, 32, "%f", nData->angle), + (nData->large) ? 1 : 0, (nData->clockwise) ? 0 : 1, + g_ascii_formatd(cSX, 32, "%f", nData->p[NR::X]), + g_ascii_formatd(cSY, 32, "%f", nData->p[NR::Y])); } else if (typ == descr_close) {

2 1

RE: [Inkscape-devel] graphics troubles between SODI and Inkscape
by bulia byak 17 Mar '04

17 Mar '04

>Okay, this file looked fine the other day in Inkscape, and is fine in >SODIPODI 0.34. It is smeared in Inkscape now. Any quick thoughts on >this? Yeah :) Forgotten break in line 541 of src/svg/svg-path.cpp. Fixed now. Thanks for reporting. _________________________________________________________________ MSN Premium with Virus Guard and Firewall* from McAfee� Security : 2 months FREE* http://join.msn.com/?pgmarket=en-ca&page=byoa/prem&xAPID=1994&DI=1034&SU=ht…

1 0