World: r3wp

Is hardware acceleration being considered at all for 2D rendering?

I suspect the answer to whether hardware acceleration will be provided 
will depend on whether AGG supports it. (I think not).

Pekr: I guess some of the crap in the IMB of Cairo is to support 
hardware acceleration and printing.

Peter - yes, it might be hw acceleration or printing. But - I think 
Carl would not like to have REBOL at 2 MB, at least not yet :-)

AdrianS: talked with Cyphre xy times. AGG does not support HW acceleration 
out of the box. But Cyphre plans to add it. Either as regular release, 
or as his own commercial project. Remember - whole View is going 
to be open-sourced. After all - it is mainly an AGG code plus events. 
There is a little bit problem with HW acceleration though - that 
stuff is not 100% compatible amongst the platforms - you might not 
get 100% pixel precise rendering ...

I think we still have some homework to do even in current state. 
What I think is suboptimal is timers. Fast View actions are still 
being a CPU hog. We discussed it several times, and I found some 
resources about timers. There are 4 - 5 types of timers in Windows 
itself, compatible with various Windows versions (or not). Windows 
is so much crap, that it does not support vertical blanking period 
=  forgot smooth not jerking not smeary animations, without DirectX.

It's funny that Maxim changed AGG to GPL for new development, then 
didn't do any.

http://www.codeproject.com/KB/system/timers_intro.aspx

I think Queue timers might be good for us. Such low level changes 
might speed up View and make it more resource friendly. Not sure 
Carl wants to experiment, so maybe we will have to try once View 
is open-sourced.

I didn't think that View  was going to be open-sourced. I thought 
that it was inside the closed-source "kernel".

no, inside the closed source kernel is only rebol.dll = rebol interpreter. 
Things like threading, networking, events, view, everything else 
platform dependant is in host code. So - R3 is going to be some 80% 
open-sourced ...

Steeve: as for your note about not thinking only about AltME. I am 
going to look into MS OCS (Office communication server) for some 
inspiration in few months. I hope we can greatly improve AltME, or 
clone it, as we can't influence the code. E.g. IOS discussions allowed 
to post even drawings. We should think about sharing dashboards, 
video/audio codecs, etc., maybe even different organisation of "rooms" 
vs "groups", more threading capabilities to discussions etc.

AGG Versions: I really don't understand why Carl is not asking for 
a Rebol license? I'm doing this quite often for GPLed stuff and it's 
not big deal. Most authors are proud if what they have done and just 
want to avoid that the dark side is using their code.

Henrik: Cool stuff so far. But, aren't you the guy doing the list-view 
stuff?


I think the current VID should really be stressed on a complex real-world 
example: A Table/Spreadsheet that can handle 1 Mio. rows with high 
speed a la Excel. It's a widget type you can use to show stuff or 
even let people enter something.

Robert, yes, I did LIST-VIEW. With current VID, do you mean R2 VID?

Robert, I think that a good reason why Carl hasn't asked for a REBOL 
license for the AGG 2.5 is that the license is the only difference 
between it and 2.4 so far, and further development of the GPL'ed 
version seems to have stopped there. :(

I wish RT would hire Maxim as a consultant to do further development. 
Other companies have done the same (according to the site).

Actually, we may be wrong in one of our assumptions: REBOL may *already* 
have a commercial license for AGG. Carl hasn't said anything to the 
contrary, just that it was "no problem" when the subject was brought 
up years ago.

There should be absolutly something done about the font display quality. 
In following screenshot, look for letters like w, s, G, D, B - it 
ruins overall impression from the new GUI look.

http://rebol.hmkdesign.dk/files/r3/gui/110.png

Henrik - are you planning to experiment with buttons too? I too find 
their weight beigh too high. I mean - nice shape, nice gradient, 
but letters are way too boold. Would like to see, how some thinner 
letters would do?

Pekr, I tried and it looks horrible. This size and shape makes the 
buttons most readable.

I have not yet experimented with other fonts yet, however.

Henrik - what I dislike about current GUI though, is its too big 
metrics overally. I especially hated it with Gnome - those guys are 
crazy. To display one dialog, it takes xy percent of your screen 
with no added value, just to display some info to users. Too much 
spacing for my liking. When I look at my current desktop - 1280x600 
notebook with Vista, and compare VID 3.4 scroller and button - it 
is unnecessary big imo, and I am not sure it should be a standard 
... Dunno if I can judge it upon screenshots (as who knows what resolution 
was it taken with), but UI elements as buttons, scrollbars are more 
than 1/3 bigger than they should be ...

Well, I don't think they are too big.

What system do you develop under? OS-X? Can you post a screenshot, 
of your system button and scroller along with R3 GUI ones? :-)

A single button is 24 pixels tall, scroller is 22 pixels wide. It 
takes a few seconds to adjust that.

(so I'm not so interested in arguing about element sizes until later 
;-))

Petr, that's why in my vid you could not only resize the windows 
up, but also down, making them smaller and removing unnecessary white 
space. And of course I had scaling too, though that needed improvements 
(eg. hinting, and making it so you didn't have to explicitly support 
it in the style code)

.. so I hope we will have it too. I need to use the code on various 
LCD panels, with various resolutions ;-)

Happy now, Pekr? :-)

maybe i'm wrong, nut Gabriele said he done it in his vid but not 
in the current vid

*but

If anyone is interested in testing (lack of time here), but I've 
found that VID3.4 hangs when I want to use a .fon font, such as "Modern" 
or "Roman" in WinXP. Perhaps it's the same in the R3 public release?

Henrik - yes, happy, thanks :-)

Henrik, I for some reason (unemployed) have time to test if needed.

Paul, ok grab a copy of the public R3 alpha. Some bugs need to be 
distilled into simple problem, so Carl can better see what goes wrong. 
I think the public alpha will produce the same issues as we have 
here in the internal release.

Could be a FreeType configuration problem

Henrik - any code example, of when/where it hangs?

Pekr, that's why I wanted to delegate that work...

Don't have time to dig into it.

yes, but some starter code at least? :-)

btw - what is happening in VID3.4 land last days/week?

Well, you would need to create a layout or a gob with a draw block 
and some rich text with the incriminating font.

Working on slider again to make it scalable. I guess Carl is busy 
with PARSE.

ah, so parse changes already happening?

I think he's been working on them for the past week

parse proposal is really complex. Will every proposal make it? Or 
does Carl choose only some of proposed featuers?

That is unknown at this time.

He hasn't been thinking out loud.

From Carl, in case those who are in the r3-alpha group are not noticing:


Test Strategy: I would like to build a group of REBOLers who can 
do different parts of this. For example, someone who will manage 
the discussion boards, someone to manage the bug tracker, someone 
to help with source archive, etc.

There are many other areas too.

Anyway, 
this week it will happen. So, if you can help now, or later, please 
let me know.