World: r3wp

do we have any date for gui R3 and the beta R3 ?

nop

could we aks for have gui R3 before decembre with Beta R3 ?

We can ask for anything, the question is, if it is possible to happen 
:-)

Carl expressed his will to release Beta this autumn, hopefully before 
Christmass.

before winter ;-)

Current project plan is here: http://www.rebol.com/r3/project-plans.html


Carl expressed his will to adapt the list and its priorities upon 
the needs of community ... I will ask him to wikify it, so that we 
might edit it ...

Do you need R3 GUI for any specific project, or just to play/learn? 
Because other folks might prefer more finished Core, e.g. Multitasking 
...

I am not sure everything on the list can make it into beta in reasonable 
timeframe ....

only for play and learn

i would like to upgrade the some of my app to R3

I thinkt that VID3 needs 2-3 weeks of work, maybe less, in order 
to be safe to suggest, because there are some structural changes/adjustments 
planned ...

I don't know how you can come with an estimate of 2-3 weeks especially 
as the current GUI only works on Windows and doesn't yet support 
Unicode..


Even if it only takes that little, I can't see Carl having 2 to 3 
weeks to dedicate to the GUI this side of the New Year.

Who said beta will span all platforms?

I am also not sure, GUI will support Unicode from the very beginning, 
altough many expressed it being a priority. There is a difference 
between GUI and VID. Carl worked on VID. Once he is back to it, I 
believe Cyphre is going to be contacted to do some work on View part 
...

My guess is 2-3 weeks for VID itself and 1-2 months for the skin. 
Really lowlevel stuff depends on the availability of Cyphre.

Cyphre is waiting for the C code alignment. He does not have recent 
sources for View. If I understand him correctly, he has to wait for 
the Host code release ...

... or for Carl sending him latest version ...

when it's ready, I hope that Cyphre can at least partially work independently 
from Carl.

A windows only beta will nicely establish R3's cross-platrom credentials.

A non-unicode GUI beta will nicely establish R3's internationalisation 
credentials.

OK, so wait another year for the beta ...

then I prefer Core only beta, all current R3 platforms ...

eek... I can use it to send email!

what else

who wrote this?

it looks very nice

who wrote what?

the online rebol console

ahhh  ...  kaj

who said that ?

After the parse enhancement, I really think the extensions improvements 
(other datatypes and devices/callbacks) should be done first... a 
lot of stuff can then start parallel to what carl works on.  not 
just by me... but many of us can work on improving R3 at the capacity 
level through extensions and have the need and will to do it.

bah...

who said what?

Who said that the online rebol console was written by Kaj?

I think Henrik wrote one, Kaj wrote one ... is this a new one ?

we can do the protocols now, by ourself, since a while, but noone 
do that. We are just a bunch of lamers :-)

ahhhh sorry, that is a new one !

And people could come with VID proposals too, to help Carl

freebol is a domain owned by the guy refered to as "hostile fork" 
afaik

I think the VID design is pretty decided upon, it just needs agile 
hands to implement it... and that IIRC depends upon a few enhancements 
to the engine and for some of the design to be finalized by Carl.

I have to change color... blue is too crowded.

looks like Carl wants to visit the problem of empty rules.

Henrik is referring to this blog entry:
  http://www.rebol.net/cgi-bin/r3blog.r?view=0258

Brian, it's possibly forlorn, but I can forsee a pattern of [and 
map! here: change to-block here/1 :here into []] (sorry, rough approximation) 
in order to serialise in one form or other.  Given there could be 
no other possible use for [into map!] perhaps parse could imply that 
conversion?

The new one was written by user known as Fork, or HostileFork ...

Good guy, and smart too. New to REBOL but not new to programming.

Another Brian :)

Chris, we've been careful to make the interaction model of map! more 
like that of objects than series. You can't parse objects either.

Right, objects present a similar pain.  I know for the most part, 
you don't want to do this.  It's a very specific case - when 'parse 
into encounters a map, instead of returning false, it converts to 
a block and parses.  I don't see this as touching the way map! works 
at all, just parse - for those occasions where it'd be useful over 
an old-style workaround.