- Inkscape Devel - Inkscape Mailman

CVS
by Joshua A. Andler 11 Nov '05

11 Nov '05

Uhm... at least three hours now SF devel CVS has been borked. This is one of the longest outages I've seen and was wondering if we should file a trouble ticked with SF or what. -Josh

2 1

Re: Have you ever thought about security issues?
by unknown＠example.com 11 Nov '05

11 Nov '05

Quoting Maxim Shemanarev <mcseem@...1076...>: > > Well, at least the version I have (v4.1) just gets into an > > infinite loop anddoesn't respond. I don't think it's correct. (assuming you mean 0.41) Hmm, you're right. It catches trivial circular references (e.g. to self or ancestor), but not more complex ones. > Anyway, there's a choice if you control the level of recursion > (and it's a must IMO). You can report an error and stop, or you > can keep rendering until some level of recursion is exceeded. For a circular reference, the standard requires that you do not render the erroneous element or any following elements at all (a naive implementation might have to throw out some intermediate rendering product to do this). I do not believe the standard requires any particular action for references that are deep but not recursive, but stopping rendering at some point is sensible even if not strictly conforming. > I don't see any other elegant way of detecting the loops, because > they can have any level of indirection (a pattern refers to > another pattern that refers to another pattern, etc... and > the first pattern refers to the first one). It's also worth noting that the circularity can happen along several "axes"; for example: ... <style type="text/css">.foo { fill: url(#blah); }</style> ... <pattern id="blah" ...> <rect class="foo" ... /> </pattern> ... > BTW, is that legal to have *nested* patterns/markers at all in > SVG? Yes. See sections 13.3 and 11.6.2 of SVG 1.1: <!ENTITY % SVG.marker.content "( %SVG.Description.class; | %SVG.Animation.class; %SVG.Structure.class; %SVG.Conditional.class; %SVG.Image.class; %SVG.Style.class; %SVG.Shape.class; %SVG.Text.class; %SVG.Marker.class; <== %SVG.ColorProfile.class; %SVG.Gradient.class; %SVG.Pattern.class; %SVG.Clip.class; %SVG.Mask.class; %SVG.Filter.class; %SVG.Cursor.class; %SVG.Hyperlink.class; %SVG.View.class; %SVG.Script.class; %SVG.Font.class; %SVG.marker.extra.content; )*" > <!ENTITY % SVG.pattern.content "( %SVG.Description.class; | %SVG.Animation.class; %SVG.Structure.class; %SVG.Conditional.class; %SVG.Image.class; %SVG.Style.class; %SVG.Shape.class; %SVG.Text.class; %SVG.Marker.class; %SVG.ColorProfile.class; %SVG.Gradient.class; %SVG.Pattern.class; <== %SVG.Clip.class; %SVG.Mask.class; %SVG.Filter.class; %SVG.Cursor.class; %SVG.Hyperlink.class; %SVG.View.class; %SVG.Script.class; %SVG.Font.class; %SVG.pattern.extra.content; )*" > -mental

2 1

Re: Have you ever thought about security issues?
by unknown＠example.com 11 Nov '05

11 Nov '05

Quoting Maxim Shemanarev <mcseem@...1076...>: > I'm afraid nobody has ever thought of it. Are there any > recommendations from W3C? According to SVG 1.1, circular references are an error and should be treated as such -- see the appendix on implementation requirements (F.2 Error processing). It's also referenced in the sections on <use> (5.6, just after the examples) and in the overview section of "References and the 'defs' element" (5.3.1). Any problems with SVG implementations due to circular references is an issue of non-conformance to that normative appendix. Inkscape gets that part of the standard partly right: it at least refuses to render circular references; in general they do not cause harm (if you can find some cases where it does, please let us know). -mental

1 0

WEBFORM inkscape on pocket pc
by Miguel Ozonas 11 Nov '05

11 Nov '05

Message from Inkscape submission form Please note, that the submitter may not read this list. CC your answer to the submitter name: Miguel Ozonas email: miguelozonas@...1075... submitted the following: Enter your message or news here. Hi, i just have downloaded your software and play it, and i have a question. Are you thinking to convert te software for pocket pc?

1 0

inkscape windows icon refab
by Bob Jamison 11 Nov '05

11 Nov '05

Hi, all... Hey, we talked about this a year ago, but nothing happened. The inkscapexxxx.ico file that we use for the Inkscape "Application" icon (resource #300), is a fix of the crappy old .ico that I made myself when we first started working on a 'standard' Windows build. I am in no way an artist, and the thing looks utilitarian and boring. So I ask the question again: is there some guy who can make a new icon which is DIB-like (various sizes) and yet pretty at the same time? Of course there are lots of people who can make nice 48x48 or 32x32 images, but we need a Windows guy who knows and can cope with the constraints of a .ico (limited colors, etc) bob

6 6

NEW: shortcuts, misc
by bulia byak 11 Nov '05

11 Nov '05

- Now you can use Shift+middle button drag in any tool to zoom into an area. This works the same as simple drag in Zoom tool, but is faster because it does not require switching away from your current tool. Together with middle button drag (panning), middle button click (zoom in) and Shift+middle button click (zoom out), this completes the set of canvas navigation shortcuts available in any tool. - In Gradient tool, Shift+R reverses the gradient definition (i.e. mirrors the stop positions) without moving the gradient handles. For example, an elliptic gradient with blue center and red periphery becomes red in the center and blue in the periphery. This works on the gradient(s) of the currently selected gradient handle or, if no handle is selected, on all selected objects' gradients. (Compare with the Node tool where Shift+R reverses the direction of the selected path.) This is especially convenient for radial gradients which, unlike linear, you cannot simply rotate by 180 degrees for the same result. - When toggling one of the "transform with object" buttons (for stroke width, rounded rectangle corners, gradients, or patterns), a message is displayed in the statusbar explaining what has changed in the program's behavior. - Zoom commands in the View menu are moved to a submenu; the Zoom In and Zoom Out commands are added to that submenu. - The contents of the statusbar message are now duplicated as a tooltip that is shown when you hover the mouse over the statusbar. [TODO: need to strip <b></b> from the tooltips.] - Whole thousands above 2000 in the rulers are now displayed as 2k, 3k, 4k etc. -- bulia byak Inkscape. Draw Freely. http://www.inkscape.org

1 0

NEW: Transform dialog redesign
by bulia byak 11 Nov '05

11 Nov '05

Fixes and improvements in the Transform dialog (Ctrl+Shift+M): - The "Apply to each object separately" checkbox is added, allowing you to scale/rotate/skew each selected object by the same amount, around that object's center. When off (by default), the selection is transformed as a whole. The status of this checkbox is remembered across sessions. (It has no effect on Move and Matrix tabs). - The new Clear button resets the values on the current tab to defaults. - The Scale tab now allows you to specify horizontal or vertical size increments in percentage or absolute units. Also there's a "Scale proportionally" checkbox which ensures that scaling preserves the width/height ratio. (If you are scaling several objects proportionally with "Apply to each object separately", you can only use the % unit to specify the scaling; otherwise each object's scale increments will have the width/height ratio of the entire selection, not of that specific object.) - The Skew tab can now specify the skew as an absolute displacement (e.g. for horizontal skewing of a rectangle, that means the shift of the top rectangle side relative to the bottom), as percentage displacement (e.g. a 1% horizontal skew of a rectangle means shifting the top side by 1% of the rectangle height), or as an angle (e.g. horizontal skew by 15 degrees results in the sides of a rectangle being rotated to that angle, while the top and bottom remain horizontal). - The Matrix tab (previously called "Transform") can either edit the current transform= matrix of an object, or post-multiply the transform= with the matrix you specify, depending on the "Edit current matrix" checkbox. As it is now redundant, the matrix in the Object Properties dialog is removed. - The dialog now correctly watches selection changes in the active document window and updates its values accordingly. - The layout of the dialog is simplified, tooltips and mnemonics added for better usability. - Many bugs are fixed, especially in value conversions between units. -- bulia byak Inkscape. Draw Freely. http://www.inkscape.org

1 0

[root@...1070...: [Inkscape-tester] inkscape build report]
by Bryce Harrington 11 Nov '05

11 Nov '05

Some different build warnings than normal... ----- Forwarded message from root <root@...1070...> ----- Date: Thu, 10 Nov 2005 22:50:40 -0800 From: root <root@...1070...> To: inkscape-tester(a)lists.sourceforge.net Subject: [Inkscape-tester] inkscape build report = Analysis Report for inkscape = Report Time: Thu Nov 10 17:37:54 2005 (Run 734) Snapshot: inkscape-20051110-0700.tar.bz2 Status: finished Platform Unpack Config Build Install WRNS ERRS ------------------------------------------------------------------------ debian x86 0 0 0 0 6 0 Test Status Score Info ------------------------------------------------------------------------ sut-1 ?? == Errors == none == Warnings == style.cpp:1254: warning: enumeral and non-enumeral type in conditional expression style.cpp:1261: warning: enumeral and non-enumeral type in conditional expression display/canvas-arena.cpp:87: warning: invalid access to non-static data member `SPCanvasArenaClass::arena_event' of NULL object display/canvas-arena.cpp:87: warning: (perhaps the `offsetof' macro was used incorrectly) display/sp-canvas.cpp:133: warning: invalid access to non-static data member `SPCanvasItemClass::event' of NULL object display/sp-canvas.cpp:133: warning: (perhaps the `offsetof' macro was used incorrectly) == Environment Summary == ------------------------------------------------------- SF.Net email is sponsored by: Tame your development challenges with Apache's Geronimo App Server. Download it for free - -and be entered to win a 42" plasma tv or your very own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php _______________________________________________ Inkscape-tester mailing list Inkscape-tester(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-tester ----- End forwarded message -----

1 0

[cworth@...573...: [cairo] Slipping 1.2.0 out to 2005-12-15, and help needed]
by Bryce Harrington 11 Nov '05

11 Nov '05

----- Forwarded message from Carl Worth <cworth@...573...> ----- Date: Thu, 10 Nov 2005 16:47:09 -0800 From: Carl Worth <cworth@...573...> To: cairo@...278... Subject: [cairo] Slipping 1.2.0 out to 2005-12-15, and help needed Shortly after cairo 1.0.0 was released, I picked an out-of-the-hat date of 2005-11-01 for cairo 1.2.0. That date has now come and gone, so so much for my powers of prognostication. I recently looked over the 1.2.0 roadmap [*] and saw that there's quite a bit still left to do. Personally, I've been focusing on the PS and PDF output quality issues. The hard parts of that are done now, but there's probably at least a week or two of work still to be done there. I had considered finishing those things up to get a 1.2.0 release out as soon as possible (in the next week or two) and letting everything else on the roadmap slip. But it looks like that's not going to be possible. I probably won't have much time to dedicate to this work until the beginning of December. So 1.2.0 won't be happening before 2005-12-15. This does provide more time for other interested people to jump in and fill the gaps for the remaining pieces of 1.2.0. If any of the following sound important to you, please consider getting involved: * win32 backend We basically just need to get it plugged into the test suite. For a long time, the issue here was missing support in the test suite for RGB24 output and for backend-specific reference images. But thanks to Keith's efforts with getting the PS backend into the test suite these two problems are fixed. So, is there a cairo-win32 developer that could fill in the two stubs in cairo/test/cairo-test.c: cairo_surface_t * create_win32_surface (int width, int height, void **closure); static void cleanup_win32 (void *closure); If so, that would be a great help and would definitely contribute a lot toward making sure that cairo-win32 remains stable in the future. * cairo_begin_group/cairo_end_group I sent a half-started patch for this to the list sometime ago. I feel good about the API, (though maybe the patch was before the addition of the current CONTENT support, and if so it should be updated to that). But the real work here is to make the intermediate surface auto-growable, which might require a tweak in the pattern code to get the offset right when the intermediate surface is finally painted. I really like this API, and I'm looking forward to having it available. I'd be glad to help anyone interested in working on this piece. * Performance fixes There's always more to be done here. I'm going to make that a big focus of my efforts post-1.2.0 by working on a performance regression suite and tackling some long-standing issues. We probably don't need much here before 1.2.0 since people are claiming that cairo 1.1.1 is already performing much better than 1.0.2. (Of course, there's still a lot of room for improvement, but this isn't something that should slow down 1.2.0 I think). * SVG backend The roadmap has this as experimental, so I think it can be thrown in at pretty much any point. It just needs someone to look at it first. I will do this after the PS and PDF stuff is done if nobody else has by then. * Bug fixes There are a couple of fairly serious bugs listed in ROADMAP already that I think really need to be fixed. A couple of the XFAILs obviously need to be fixed as well, (such as the *ahem* segfault from extend-reflect). Meanwhile, there have been a bunch of little bug reports, (slight build problems on various platforms for example), that should be pretty easy picking for almost any willing volunteer. As usual, there's almost no end of fun work to be done. I appreciate everything that so many of you are already doing. And this is also a great chance for some new faces to jump in and make some contribution. I'm glad to provide more details and guidance on any of the above as needed. -Carl [*] http://cairographics.org/ROADMAP _______________________________________________ cairo mailing list cairo@...278... http://cairographics.org/cgi-bin/mailman/listinfo/cairo ----- End forwarded message -----

1 0

newest gtkmm-dev necessary to compile HEAD
by Ralf Stephan 11 Nov '05

11 Nov '05

Hello, So it appears one needs the newest gtkmm libs to compile HEAD. I did not foresee this when using some features of the library, and, now that people have upgraded, it's nonsense to revert ;) What's unclear to me is which version is fine, here it is libgtkmm-2.4-dev 2.8.0-0ubuntu1, can others add info? thanks for your patience, ralf

9 15