What is proper transform/scaling behavior?
Thanks again for Inkscape, it really rocks. Does anyone know what the proper behavior is for scaling objects via transform attributes? Is the stroke supposed to be scaled or not? I figured there would be just one answer, but the svg-rendering programs are all over the map here:
http://web.mit.edu/jonas/www/anisotropy/
I ran into this when looking at the PS & PDF output of both Inkscape and Batik; at first it appeared to be caused by people assuming SVG's transformation matrices worked the same way as ps, but then I can't quite figure out what the hell the Adobe SVG viewer is doing...
So I'm curious to know what the Right behavior is, or if this is an ambiguous region in the standard...
Thanks again, inkscape is amazing...
...Eric
Eric Jonas wrote:
Does anyone know what the proper behavior is for scaling objects via transform attributes? Is the stroke supposed to be scaled or not?
Inkscape's behavior is incorrect. This is the oldest bug in the tracker: http://sourceforge.net/tracker/index.php?func=detail&aid=851008&grou...
Does anyone know offhand if cairo handles this correctly?
Aaron Spike
On 9/27/05, Eric Jonas <jonas@...415...> wrote:
Thanks, that's an interesting comparison. I didn't know Illustrator has the same problem when editing SVG (if this is indeed a problem).
Can you test it with Illustrator again, but this time by importing the PS file exported from Inkscape into AI?
-- bulia byak Inkscape. Draw Freely. http://www.inkscape.org
On 9/27/05, Eric Jonas <jonas@...415...> wrote:
Thanks, that's an interesting comparison. I didn't know Illustrator has the same problem when editing SVG (if this is indeed a problem).
Can you test it with Illustrator again, but this time by importing the PS file exported from Inkscape into AI?
I just did; while gv displays the .ps properly (vertical lines are wider), importing the .ps into illustrator results in a rectangle with a uniform 1pt-wide stroke.
I thought Adobe, you know, wrote the book postscript or something :) ...Eric
On 9/27/05, Eric Jonas <jonas@...415...> wrote:
I thought Adobe, you know, wrote the book postscript or something :)
Yeah, weird. Can you also try some other PS renderers - such as distiller+acrobat, or photoshop?
-- bulia byak Inkscape. Draw Freely. http://www.inkscape.org
Yeah, weird. Can you also try some other PS renderers - such as distiller+acrobat, or photoshop?
Photoshop renders it with thicker vertical lines, i.e. properly. Distiller on solaris (don't ask) gives:
web.mit.edu/jonas/www/anisotropy/inkscape.postscript.pdf
(for the record, the ps generating all of this is: web.mit.edu/jonas/www/anisotropy/inkscape.postscript.ps)
Now I'm just worried about all the diagrams I have made in inkscape which have this sort of "issue" because of errors in the inkscape renderer. Wherever I have created a group and scaled it, it seems as if the strokes have been scaled as well... this is going to take a bit of thinking to avoid fixing by hand :)
...Eric
On Tuesday 27 September 2005 23:53, Eric Jonas wrote:
On 9/27/05, Eric Jonas <jonas@...415...> wrote:
Thanks, that's an interesting comparison. I didn't know Illustrator has the same problem when editing SVG (if this is indeed a problem).
Can you test it with Illustrator again, but this time by importing the PS file exported from Inkscape into AI?
I just did; while gv displays the .ps properly (vertical lines are wider), importing the .ps into illustrator results in a rectangle with a uniform 1pt-wide stroke.
I thought Adobe, you know, wrote the book postscript or something :) ...Eric
It should or should not have wider vertical lines?
Craig
I thought Adobe, you know, wrote the book postscript or something :) ...Eric
It should or should not have wider vertical lines?
Sorry, this is getting confusing; as I understand it, the ps that inkscape has generated specifies a transformation matrix; for both svg and ps (and pdf), transformation matrices apply to stroke widths as well as the locations of the vertices.
So in this case, ghostview and photoshop work properly, whereas Illustrator CS renders the rect. incorrectly. ...Eric
participants (4)
-
unknown＠example.com -
bulia byak -
Craig Bradney -
Eric Jonas