World: r3wp

I would like to know when the problem with Agg Matrix will be solved 
and when Carl plans to integrate in REBOL/AGG a radiant function 
that can better feet with SVG datas  ? I started 3 month ago an SVG 
engine and since then I'm stuck because of those 2 problems... SVG 
without transformation matrix and without gradients it's like nothing 
.... ( see greoup SVG renderer for more informations and test code)

shadwolf: please see my reply in 'Carl Only' group.

It's 1:04AM here, not a time for serious work, so I'm fiddling little 
bit with some eye-candies in new DRAW and I must say it's SO powerful. 
I'm just starting to see the potential...whooo! Thanks

ft: make face/font [size: 60] 
view center-face layout [
	origin 0
	box 400x80 white effect [
		draw [

   fill-pen linear 0x0 0 100 30 1 1 black white black white black white
			font ft
			text 0x0 "Wavin' REBOL" vectorial
		]
		emboss
	] with [
		rate: 0
		n: 0
		feel: make feel [
			engage: func [f a e][
				f/effect/draw/3/x: n: n + 2
				show f
			]
		]
	]
]

That's very cool.

Wow. Can you made an animated gif from that? :)

'save doesn't support GIF, so I doubt it... Unless, of course, someone 
who understands GIF files hacks something together  >.<

I guess Kru can save multiple images and use some external converter? 
:)

Or that.

Cool Kru! :-)

thanks :)

In the new DRAW dialect, should SCALE have influence on LINE-WIDTH?

that is a very hard question, because in most vector drawing apps 
it doesn't, while it is also arguable that it should.

also, i don't know if AGG is able to scale a line width if the scale 
is not the same for x and y.

That seems like a frustrating issue.

Cyphre thank you for your answer You need more infos on Gradient 
here is the official webpage on the topic hopes it  could help you 
to make a svg gradient complying fonction in draw  ...:)

http://www.w3.org/TR/SVG/pservers.html

Cyphre there is no missing thing in AGG  gradient the coordinate 
datas given by SVG gradient data block is different. To make it short 
and simple SVG work with area pourcentage of filling (what ever the 
 size of the area to colorize is you will allways have red 20% blue 
40% black  40% if red blue black are your 3 colors that are used 
to make the gradient effect) Then we have a vectorial information 
that  give the orrientation of the gradient effect. SVG gradeint 
are linear, circular or parterned (in most case linear is applyed 
...). REbol AGG gradient use area offset you have a start offset 
a end offset and the  colors are degraded along this vector  nothing 
more nothing less you have no area coloring infos. This is if  my 
recalls are good the differencies betwin AGG gradient and SVG gradient.

and the proposed solution is? Can we have some helper parameters 
in rebol AGG to allow SVG kind of parameters? Are AGG gradients subset 
of SVG? Is there 1:1 equation, so that we can transform between the 
two? Maybe the problem is a bit more complex?

Pekr and the solution proposed by Cyphre  was to insert to agg in 
built fonction complying with SVG datas ...

yes, that should be enoug, no? If Cyphre can use AGG internals to 
simply make it working, so he will, no?

Pekr yes we can do adaptation but this will enslow dramatically the 
SVG rendering. My goal is to make a SVG widget capable for REbGUI 
 to dispaly graphic content like tool bars, logo etc so faster is 
the render better it's

he can create e.g. gradient-svg command ...

I am not talking about adaptation in rebol level code, but directly 
in rebol AGG layer = C code - that should not be slow, no? Cyphre 
would then expose it via e.g. gradient-svg new command ...

we have e.g. new natives for rgb-to-hsv and hsv-to-rgb already ...

yes that was what he had in mind and that perfect feet my need  ^^. 
But he asked me in Carl's only  group some more detailled informations 
about SVG way to handle gradients. Th'at why i give him the link 
to the gradient SVG official documentation and why i make this simple 
explaination ;:)

it was simulated in rebol level, was found being slow, so Carl reacted 
upon the request and made those functions native, as they became 
usable for us ...

yes the actual svg render have been made in one week it's pretty 
fast and the code is very tiny ( 400 lines for a svg to AGG/Vid rendering 
engine that not so mutch  ^^)

basically SVG is a XML  structure but for several reasons i choose 
to pass thru a XML 2 Object converstion and then I  treat the SVG 
infos converted to rebol objet structure ;)

main reaseon to act like that xas svg recursion structure is a nightmare 
if you use parse or xml-parse

xml-to-object? Are you talking about Gavain Mckenzie's parsing XML 
stuff?

yes exactly ;)

http://shadwolf.free.fr/svg-demo-shad05.zip
http://shadwolf.free.fr/svg-demo-shad06.r

I found it being really excellent ... but - it is based upon old-good 
'parse itself of course :-) anyway - way much more usefull than xml-parse, 
which is imo absolutly insufficient for real work - but it is long 
time I looked into it ...

see those link  05.zip countains an sVG sample file and the 05 code 
 06.r is the more advanced version i made

I try as often i can to retake other ones ideas to enlight them and 
make my own production being very smooth ;)

05.zip url is wrong probably ... I downloaded 06.zip

http://shadwolf.free.fr/svg-demo-shad06.zip

yes you are right 05  have been remplaced by 06 ...

Shadwolf: so you are missing the percentage ratios of gradient colors? 
I have been thinking about this enhancment too. I can't tell you 
when this will be added but I believe I'll try to add it during the 
next gradient redesign phase.

Geomol: in the next DRAW update there should be scalable line-width 
available.

Okay! :-) I'll contact you privately. I got some questions, if you 
have a little time.

There is some 1.3.1d or so called 1.3.2 beta VIEW version. Are the 
AGG fixes going to be part of it?

No, this version fixes couple of binding bugs afaik.

Cool :)

I think, there's a bug with colour keying in DRAW, when doing an 
IMAGE. Grey-tones seems to work, and also the green colour 0.136.0, 
but most other colours don't work!?

The colour keying bug is known issue and was fixed for the next release. 
Anyway thanks for reporting ;)

:-)

I also read someone reporting problems with different characters 
rendered using AGG. This should be also fixed in the next release.

ah, I just wanted to ask! Thanks ...