ANNOUNCE: tkpath 0.2.2
mat...@gmail.com
-----------------------------------
I have added a text item, called ptext, which completes this
package since all items (window disregarded) can now be replaced.
It supports the same drawing model with stroke/fill as the other
tkpath items. The font handling is not the standard Tk one and needs
some more work. See the links below for examples and screen shots.
More fixes:
o rewritten gradients for Quartz (MacOSX)
o serious bugfix when combining -matrix with move
o added -units bbox|userspace option to linear gradients
o minor fixes...
The next step will be to add structure to canvas items, using groups
and inheritance, similar to a tree structure. This will need a new
canvas widget. I'd like to have opinions about this.
The best way to get it is using cvs which also contain builds for
MacOSX,
WinXP, and cairo graphics.
Home: http://tclbitprint.sourceforge.net/
Shots: http://tclbitprint.sourceforge.net/tkpath/quartz/index.html
http://tclbitprint.sourceforge.net/tkpath/gdiplus/index.html
http://tclbitprint.sourceforge.net/tkpath/cairo/index.html
Enjoy, Mats (mat...@gmail.com)
Roy Terry
news:1168433648....@p59g2000hsd.googlegroups.com...
> Announce - tkpath 0.2.2
> -----------------------------------
>
> I have added a text item, called ptext, which completes this
> package since all items (window disregarded) can now be replaced.
> It supports the same drawing model with stroke/fill as the other
> tkpath items. The font handling is not the standard Tk one and needs
> some more work. See the links below for examples and screen shots.
Very nice looking angled text with antialiasing using stroke/fill - nothing
else in the Tk world can do that at a native level. Great work!
>
> More fixes:
> o rewritten gradients for Quartz (MacOSX)
> o serious bugfix when combining -matrix with move
> o added -units bbox|userspace option to linear gradients
> o minor fixes...
Any plan to addesss bugs and RFEs on sourceforge?
>
> The next step will be to add structure to canvas items, using groups
> and inheritance, similar to a tree structure. This will need a new
> canvas widget. I'd like to have opinions about this.
Saying new canvas widget sounds like a good bye to Tk. Are
you planning to make an upward compatible widget or a tkpath
only widget?
I recall you mentioned this item/structure need with v 0.1 and
got puzzlement from myself and others. Is the tree structure stuff
having to do with Coccinella needs? In my work so far I'vve been able
to simple use groups of tag strings to provide all the "structuring" I
need.
>
> The best way to get it is using cvs which also contain builds for
> MacOSX,
> WinXP, and cairo graphics.
Thanks for this excellent package. It's becoming essential to
my projects.
Roy
Kevin Walzer
>> The best way to get it is using cvs which also contain builds for
>> MacOSX,
>> WinXP, and cairo graphics.
>
> to build under XCode (I also can't figure out what the error is -- I'm
> not used to xcode). The configure file errors with a "not implemented on
> macosx" message. Any instructions on how to build this on an intel mac?
>
> Cheers,
>
> -- Neil
I haven't tried this with tkpath, but building some of Mats' other
packages as universal binaries (cf. maccarbonprint) required editing
some project settings in Xcode (under "Edit Project Settings" to set up
as universal binary, and "Edit Active Target" to set the compiler to gcc
4). I also had to tweak a few files to adjust for complaints by
gcc--it's stricter than gcc 3.x. Mats is still using Project Builder
under 10.2.
--
Kevin Walzer
Code by Kevin
http://www.codebykevin.com
Googie
> The next step will be to add structure to canvas items, using groups
> and inheritance, similar to a tree structure. This will need a new
> canvas widget. I'd like to have opinions about this.
What about TkZinc? It's pretty cool, mature and LGPL licenced (free for
commercial and non-commercial usage). Supports OpenGL, but doesn't
require it.
--
Pozdrawiam! (Regards!)
Googie
Ramon Ribó
Hello,
Very nice work.
Do you have any plans for Windows printing?
I assume that if you use GDI this should be more or less
straightforward.
Maybe the packages "print" and "gdi" from:
http://www.schwartzcomputer.com/tcl-tk/tcl-tk.html
can be of some help.
Ramon Ribó
--
Compass Ing. y Sistemas Dr. Ramon Ribo
http://www.compassis.com ram...@compassis.com
c/ Tuset, 8 7-2 tel. +34 93 218 19 89
08006 Barcelona, Spain fax. +34 93 396 97 46
Neil Madden
> The best way to get it is using cvs which also contain builds for
> MacOSX,
> WinXP, and cairo graphics.
The MacOSX build appears to be PPC only. I can't seem to get the project
Arndt Roger Schneider
Hi,
nice work! At least i was able to throw some work into cairo 1.2.6
under OSX(tiger)-X11 ;-)
Are the questions in the README still valid ?
if yes:
Stick with a generic concept for coordinates -- makes it easier to port
canvas's rect to prect also.
Zinc uses itemconfigure -position to --mostly-- specify coordinates
for some types. As a result zinc does also support coords... duplicated
work.
On tkpath:
The demo scripts (website) still features -transition instead
of -lineartransition.
I usually create the GUI offscreen, this trriggers some odd
effects in tkpath:
if around ~100 ptext are created wish
crashs with a BUS-Error.
gradients are not properly created --
it does work after a GUI-part was visible.
with tkpath 0.2.2 on OSX-X11(cairo):
Gradients end around 20 pixel left off the right border.
Haven't seen this with 0.1 on aqua.
Are you aware, that linear & radial-gradients are visible
across interpreter bounderies?
I frequently encountered an alloc-trap when using tkpath 0.1 (aqua) and
0.2.2 on X11. This might be related to how you -not-- deal with
interpretes ... I'm not certain about that however -- just guessed it;-(
lineargradient & radialgradient:
The gradient's name is currently tho only way to figure out what
gradient type it is.
I do prefer to specify a gradient-name before the gradient
is created -- there should be a way to ask the gradient which type it is.
TK 8.4.11 &tkpath 0.2.2 tiger-X11 and
Tk 8.4.11 &tkpath 0.1 on aqua (64bit support,threads & shared)
-- didn't compile tkpath 0.2.2 for aqua,yet
Cheers!
-roger
mat...@gmail.com
To all: sorry for my slow response but Google groups which I use
as my news reader didn't update for several days :-(
Roy Terry skrev:
>
> Any plan to addesss bugs and RFEs on sourceforge?
>
Just found it. Will process when I get some time.
> >
> > The next step will be to add structure to canvas items, using groups
> > and inheritance, similar to a tree structure. This will need a new
> > canvas widget. I'd like to have opinions about this.
>
> Saying new canvas widget sounds like a good bye to Tk. Are
> you planning to make an upward compatible widget or a tkpath
> only widget?
I don't know yet. Depends. Input?
> I recall you mentioned this item/structure need with v 0.1 and
> got puzzlement from myself and others. Is the tree structure stuff
> having to do with Coccinella needs? In my work so far I'vve been able
> to simple use groups of tag strings to provide all the "structuring" I
> need.
I was thinking about a tree structure with groups, similar to how svg
documents
are structured, and include some option inheritance.
If an option is set for a group, all items beneith it will inherit it
unless they
have set it themself. TkZinc is using groups but I don't know the
details.
/Mats
mat...@gmail.com
Kevin Walzer skrev:
Sorry, I can't help you there. I really should get one of those
MacBooks Pro
to bring myself uptodate when it comes to building universals.
/Mats
mat...@gmail.com
Googie skrev:
I have looked at it but the code is just full of OpenGL calls.
Although OpenGL is not required ( I don't know what they do without
it)
I thought it was best to go with a native graphics API to get most out
of it.
Also, the APIs I use are pretty closely matched to SVG graphics which
set
a standard, and which I've found is pretty easy to use, yet you get a
lot of freedom.
TkZinc also has a lot of custom items for air traffic control if I
remember correctly.
/Mats
mat...@gmail.com
Sorry for the delay. Google groups arghhhh...
Arndt Roger Schneider skrev:
> Hi,
>
> nice work! At least i was able to throw some work into cairo 1.2.6
> under OSX(tiger)-X11 ;-)
>
> Are the questions in the README still valid ?
> if yes:
>
> Stick with a generic concept for coordinates -- makes it easier to port
> canvas's rect to prect also.
>
copy that
> On tkpath:
>
> The demo scripts (website) still features -transition instead
> of -lineartransition.
>
Fixed
> I usually create the GUI offscreen, this trriggers some odd
> effects in tkpath:
>
> if around ~100 ptext are created wish
> crashs with a BUS-Error.
>
> gradients are not properly created --
> it does work after a GUI-part was visible.
>
> with tkpath 0.2.2 on OSX-X11(cairo):
> Gradients end around 20 pixel left off the right border.
> Haven't seen this with 0.1 on aqua.
>
Do you have a script to show this?
to: mat...@gmail.com
> Are you aware, that linear & radial-gradients are visible
> across interpreter bounderies?
They should be created using a canvas command and belong to a
canvas instance, but so far I have stuck to the item plugin API.
>
> I frequently encountered an alloc-trap when using tkpath 0.1 (aqua) and
> 0.2.2 on X11. This might be related to how you -not-- deal with
> interpretes ... I'm not certain about that however -- just guessed it;-(
>
I've already got a script that makes the whole thing crash.
Perhaps the same. Will fix when I get some time.
There are several mem leakages as well. Fixed two since 0.2.2
> lineargradient & radialgradient:
> The gradient's name is currently tho only way to figure out what
> gradient type it is.
> I do prefer to specify a gradient-name before the gradient
> is created -- there should be a way to ask the gradient which type it is.
>
Good idea. I add this post to the SF bug list so I wont forget.
/Mats