[REBOL] Re: R: Re: R: Re: R: Re: Who read this mailing list?
From: henrik:webz:dk at: 24-Nov-2007 17:27
On 24/11/2007, at 1.42, Carl Read wrote:
> On Saturday, 24-Novenber-2007 at 1:08:04 Giuseppe Chillemi wrote,
>> I have read everywhere to have a personal opinion about the release
>> Rebol 3 won't be complete for another one and half year. Expect a
>> during the end of the winter/start of the spring; all protocols
>> will be
>> ready by the end of 2008/start of 2009 and then the new VID will
>> see the
>> light. In 2009 rebol will face the major debugging work and in
>> many external modules will be integrated into Rebol new open
> That's a very long three months...
That seems about right for the 2008 part, but VID3 will be available
with the first alpha according to the current plan as approved by
Carl. Having monitored the progress closely (I helped testing the
prototypes), I think it took only a few days to build the first
prototype, about a week more to develop the second one and in about 4
weeks, we got to where we are now with extending the second
prototype. But development has stopped right now, because Gabriele
has been busy elsewhere for about 2 months and it's not certain when
he will return, because he keeps being pulled back to work on
something unrelated to R3. He's crucial for moving VID3 forward. We
also need to develop at least one more prototype before the design
can be deemed good enough. It doesn't take long for Gabriele to do
it, but he is unfortunately busy.
So why can't someone else, perhaps a lazy guy like me do that?
All coders and contributors are specialists in their own area with me
working on VID3 GUI appearance, DevBase GUI, generic testing and the
Cookbook list for DocBase. This is the best I can do and if I were to
try to work on VID3, I'd be ruining Gabriele's work with inferior
code, because VID3 is a pretty good design so far and I can't carry
that torch, so to speak. My face turns white, every time I look at
the VID3 code. :-) Cyphre is the specialist in low level graphics,
GOBs and WinXP graphics integration. He knows how to bring the best
performance out of AGG. Brian Tiffin works on DocBase and knows
things about MediaWiki that we others don't. And so there are great
people working in each narrow field.
The truth right now is that RT needs hard core expert coders, not PR.
From where I'm sitting, it looks to me that all the best coders are
strained very hard to make stuff for R3 right now. They are under
contract for doing unrelated work and that delays all the juicy parts
of R3 for a few months. That's unfortunate, but PR will not help
that. In fact it will make it worse I think, to attract curious
people to R3 at this time, people who are not qualified to work on
these crucial parts and will ask questions that aren't timely to ask
right now. People who are qualified to make critical design reviews,
who write great REBOL code and understand the inner workings of R2 to
be able to judge what would be good code for R3, are needed for R3.
Also people who are able to do a lot of boring tests are needed. Carl
is asking for tests to new code all the time and we are currently
underdelivering, because it's time consuming and boring work. :-)
In order to get to do that, the best thing one can do, is to be a
proven R2 coder and try to get into the AltME world for R3
development. R3 development only happens there, not on forums or
mailing lists or in blogs. It may take some time to get in, because
you need to be approved by Carl and he's busy all the time.
Right now, DevBase is "easier", because it's "just" REBOL code and
more people can directly contribute to that and over 230 code
submissions have been made to its database already, some to R3, some
for DevBase improvements and some to improving REBOL/Services
(something that also benefits R/S under R2).
But yes, REBOL development generally suffers from delays, because one
single coder can hold it all up, due to family or contractual
obligations elsewhere. Had Gabriele been available for the past 2
months, VID3 would be in the late debugging phase now, rocking our
socks off even more than it does now. That's the price to pay, when
the contributions must be top notch.
About the rest of the time table posted above, you are not taking
DevBase into account: As we move further towards completion, simpler,
upper level parts will be accessible to anyone through DevBase, so
the debugging work won't last that long. Combining DevBase with a bug
tracker and a discussion system like AltME works very well. At that
point, it's very much up to you, how fast development goes, so if I
were you, I'd free up some time in 2008 for testing, learning how to
use DevBase and begin some REBOL code scrutinizing. :-) I'm fairly
sure that by the end of 2008, R3's core with all advertized features
are largely done and ready for the really big expansions that will
require a large number of developers.
Henrik Mikael Kristensen