r3wp [groups: 83 posts: 189283]
  • Home
  • Script library
  • AltME Archive
  • Mailing list
  • Articles Index
  • Site search
 

World: r3wp

[DevCon2005] DevCon 2005

Terry
3-Oct-2005
[1917]
In Nashville, muscians are a dime a dozen.. the money comes from 
writing a big hit.
Maarten
3-Oct-2005
[1918]
Terry: I partially agree. But REBOL is /more/ productive and has 
some very good things coming. Sometimes a big hit comes from a different 
tuning; now may be the time.
Pekr
3-Oct-2005
[1919]
yes, that might be true ... if such new good things are used for 
kind of killer app creation ...
Terry
3-Oct-2005
[1920]
The more I toy with AJAX, the more disinterested I become with desktop 
apps.    I think client/server is the future now..  and apparently 
so does Google.. they're buying up tons of dark fibre and wireless 
'last mile' solutions.. some have speculated that they plan on pushing 
apps to the end user.. and apparently for free   http://www.engadget.com/entry/1234000853054453/
Henrik
3-Oct-2005
[1921]
Sometimes a killer app is an app like any other, just taking up 1/10th 
of the space, has much fewer bugs and was designed and programmed 
much quicker than competing apps by fewer developers. Such apps have 
a better chance to grow into something big, because REBOL offers 
real leverage in programming and deployment. It can make you deliver 
and you can even do it cheaper. A lot of the Java folks, I talk to, 
often have these problems. It just takes too long to do something 
simple and it costs money.

I think this is why many projects run up in millions of dollars, 
are buggy and delayed. Leverage is wildly important if you want to 
deliver.


I haven't worked with AJAX, but I've coded up javascript RPC systems 
and have been working on building web pages that feel more like apps 
than web pages. It's a nightmare, because a web browser isn't really 
built for apps. It's built for document viewing and browsing, and 
by adding a ton of APIs and technologies, it's now possible to build 
something that can perform slightly like a real application, but 
is much slower and much harder to develop. I'm done with that. It 
gives me absolutely no leverage.
Pekr
3-Oct-2005
[1922x2]
Bobik here (half a year novice to rebol), mentioned screen painter 
is badly missing. Typical developer imo looses too much of  time 
designing VID for screen to look acceptable ....
I hope rich-text and rebcode allow for some nice, may it be even 
very simple, but usefull IDE :-)
Henrik
3-Oct-2005
[1924]
about rich-text, what was demoed? was it only possible to display 
rich-text or actually edit it? what about speed?
Pekr
3-Oct-2005
[1925]
hopefully rebcode could help to seepd things eventually up? :-)
PhilB
3-Oct-2005
[1926]
Has Carl said he would implement rebcode ?

The last I heard it was an interesting experiment that Carl toyed 
with whilst on holiday.
Pekr
3-Oct-2005
[1927]
if it was already working, why not to have it inside? :-)
PhilB
3-Oct-2005
[1928x2]
Well, it may not be complete .... or be incompatible in some way 
with existing code .... who knows.


For a speed increase on 30 times one would have thought it would 
have generated more excitment in the community .....
making some things possible that are currently infeasable .... like 
mp3 playback from within Rebol.
Henrik
3-Oct-2005
[1930]
I think we've had plenty of excitement? I just think most of us wouldn't 
see it so soon, plus Carl has a million things in the pipeline...
Maarten
3-Oct-2005
[1931x2]
Terry: you may be right, and you may be not. I like AJAX. But is 
Google Earth Ajax'ed? Or MSN? Skype? Winamp? ...
On rebcode, check the alpha/beta downloads http://www.rebol.net/builds/031/rebcode.zip
 but be sure to read the warning in the docs!!!
Henrik
3-Oct-2005
[1933]
wow! too bad I can't play with it now, but I'm sure pekr would like 
to give it a try? :-)
Romano
3-Oct-2005
[1934x3]
Philb, a realistic rebcode gain in real code is 16-18x, but a lot 
depends from how much is optimized the rebol code and the rebcode 
-- obviously if you use it in loop anly to do math on integer, gain 
can be greater, and if you make a lot of change on string and image, 
gain can be lesser.
For example i wrote a Arc4 algorithm in rebol and rebcode: the result 
is 16x with maximum optimization on both side (i can reach).
I think that the example of Carl: unrolled loop of integer add, is 
just the best case for rebcode performance. Consider it a max.
Pekr
3-Oct-2005
[1937x2]
just run demos. From doc I can understand it is still missing some 
features, so not fully finished?
Romano - you talk about it in a strange mood, don't you? Would you 
guys expected rebcode to speed rebol even further?
PhilB
3-Oct-2005
[1939]
My 30x quote was from Carl's Blog .... he waas just being more realistic 
I think.
Still a 16x increase is pretty impressive ....
Romano
3-Oct-2005
[1940x2]
Yes, impressive, and useful: the arc4 code is almost unuseful in 
rebol code, but acceptable in rebcode. And it can be also more fast 
if rebcode had 1) math on char 2) pick and poke 0 based
i f i find some time, i will implement the blowfish algorithm a lot 
pmore complex than arc4, can be a good test.
Will
3-Oct-2005
[1942]
for the impatients: Gabriele did a great job of taping everything, 
the audio was taken directly from the speackers microphone and it 
will be available in high quality mpeg4/aac
Pekr
3-Oct-2005
[1943x2]
were some pizza tech party technical talks recorded? :-)
my experience is, that the most valuable info comes in just such 
sessions, not presentations and QA sessions :-)
DideC
3-Oct-2005
[1945]
- Pekr : Carl hasprepared his presentation the night before he given 
it (so he was not in great form this particular day ;-)


- About Internet connection : the scholl connection where we were 
is provided by an Italian telephon companie that ensure the security. 
Mario have tried more than a week before to have some opening on 
it. Lot of faxes, telephon calls and time lost. We were all disapointed 
of that, in or outside the Devcon.


- Rebcode : it will be part of Rebol soon. You can even try it if 
you want (windows builds folder, the zip file). But it still in design 
stage. The form, the capabilities and implementation are sugest to 
changes. Carl has insisted on the fact that rebcode will be a bit 
tricky to use. Kind of assembler code. Not for newbies.


- Rich text : Cyphre show us its work. Actually it's a .dll, like 
he did with AGG in the begining. It's not just a renderer : it's 
editable rich text. Dialect base like Carl's blog proposed. Nothing 
more to say, it's just a prototype, a proof of concept, like AGG 
was at its time. It use AGG as engine, so font can be aliased or 
not.
Pekr
3-Oct-2005
[1946x2]
DideC - thanks a lot for your info ...
I wonder about further AGG integration - you know, AGG added even 
compositing engine lately, so maybe Rebol's native one could be exchanged 
:-) Also there are now effects provided in 'effect level and AGG 
level, that is inconsistent .... what is more - with rebcode coming, 
we can do most effect ourselves and hopefully be fast enough?
DideC
3-Oct-2005
[1948]
Rebcode just allow integer, string/binary and block operation/calculation.
Pekr
3-Oct-2005
[1949x2]
that is enough - the binary part, no? You can already look at image 
example in the rebcode .zip archive.
and according to docs to rebcode, there is a plan to add kind of 
pick/poke or so ...
DideC
3-Oct-2005
[1951]
I have asked Carl about the problem of effect and draw different 
engine. He answered that the effect engine will be replace by AGG 
at terms. No schedule.
Pekr
3-Oct-2005
[1952x2]
ok, that is good to know, thanks once again ....
editable rich-text - how fast it was? Was it fast enough to have 
small, cute IDE for rebol? Simply text editor for color syntax?
DideC
3-Oct-2005
[1954]
We compared my rich text engine and his one. Cyphre richtext display 
instantly, but is a bit slow while scrolling but usable. It's because 
everything in the area is recompute. Mine take some time to display, 
but scrolling is just an offset of the pane to change.
Pekr
3-Oct-2005
[1955]
you use library for that too?
DideC
3-Oct-2005
[1956x2]
It will be enough for making a new Words if you want ;)
No, mine is 100% Rebol, but just display, no edit at all.
Terry
3-Oct-2005
[1958]
Maarten.. Google earth is AJAX I believe.. (or is it Google Maps?)
Pekr
3-Oct-2005
[1959]
not sure - the speed of rebol is sometimes really limiting, let's 
just admit it ... simply - try e.g. here to scroll with altme - really 
slow ...
DideC
3-Oct-2005
[1960x2]
Google maps
It's pretty the same with richtext. But it's a prototype in external 
.dll
Pekr
3-Oct-2005
[1962]
Terry - we do understand your message, but imo rich-apps have their 
place :-) Browser UI's are terrible usability wise. But you may be 
right if you think that browsers become container (canvas :-) for 
various stuff. We should probably managed to get plug-in running 
anyway :-)
Terry
3-Oct-2005
[1963x2]
OK, so Microsoft is embracing AJAX (XMLHTTP, recently pushed strongly 
by Google’s Google Suggest and Google Maps) and they’re putting their 
own abstraction layer – codename “Atlas” – on top of it to make it 
even easier. But why is Charles Fitzgerald, MS’s general manager 
for platform technologies, saying “People who do (AJAX development) 
are rocket scientists
?"
At first I thought the plug-in was more or less useless... not anymore
DideC
3-Oct-2005
[1965]
Did I forget to mention that a Rebol/plugin will be made for Firefox 
?!
Pekr
3-Oct-2005
[1966]
dunno - but AJAX is imo just proving the idea of rich apps we do 
have with Rebol. If not, they would be satisfied with browsers and 
few widgets as is :-)) So AJAX is just kind of rebol aproach inside 
browser. Let's make proper plugins, improve VID and bye bye AJAX 
:-)