major Rebol programming projects
[1/5] from: jackseay:earthlink at: 28-Aug-2000 15:30
I have been spending the last month trying to learn to create web
pages with Dreamweaver and Fireworks. While both of these are
incredible in many ways, they also show up the defects in the
HTML-Javascript-Flash-Shockwave-CSS-CGI way of creating things. The
tools are huge, complex beasts. You have to understand the
compromises needed to target various versions of different browsers
on different OS's. HTML is user hostile for attempting to create a
page layout. It requires endless plug-ins to make up for its'
deficiencies. The users don't generally get to edit anything you
create to produce their own versions and add comments, etc.
Rebol could help create better tools to edit web pages as interim
projects, but the ultimate goal should be to replace the Internet as
it exists today with something better: a stable data pool interacting
with multi-dimentional browser-editor-camera-microphones. Rebol is a
new generation language, being hardware independent, incredibly small
(and thus fast, easily upgraded, and nearly bug-free), easy to learn
(yet powerful), and has an interactive programming environment. It
also has built-in graphics, including image processing.
What programming projects are likely candidates for major coordinated
efforts?
Start with simple data structures that have no built-in limitations
as to: 1. the type or amount of information they can contain or: 2.
the types of manipulation and interaction with user/creators.
Something like Xanadu could add the ability for anyone to edit
anything produced by anybody. The original version of a
document-picture-audio-video would still be available, the original
owner would automatically get paid royalties for all quotations and
revised versions. Nothing would get lost: not links, documents,
versions, or addresses. An infinite number of document ID's are
available, thanks to an ingenious addressing scheme. Basically, the
whole universe of information is treated as if it were all one
document, and every single character, pixel, movement, edit, and
sound can be addressed. Normally you would address either a full
document or segment (span), but it is possible to get more specific,
down to the bit level. It is also multidimensional, not just 3-D, but
n-D. The dimensions could replace a lot of programming. Just show the
person's name
dimension as "x", the date as "y", pick out a
specific message, then look at it's dimensions to edit the text in
the editor dimension (using any available tool), now shown in "x",
and the grapic editor dimension, now in "y", then roll into the
animation and sound dimensions (with their own dimensions for each
action they can perform), etc. until you are finished. Move to
another dimension and select "send". So instead of writing a program
that will perform these 15 things to a set of documents or images,
just select the items in all the needed dimensions, and take a
snapshot. Then hit "Play", sit back and watch. By selecting a set of
dimensions and choosing operations performed in them, you've created
a program without looking at any hyroglyphics. It is up to the
presentation program to determine whether the user sees
multi-dimensions, an outline, a grid, a 3-d environment, or just a
plain page. The same information could be viewed any of these and
other ways.
No need to load 4 huge programs, using only 1% of the code or less in
each one. That obsolete methodology is similar to the HTML defect of
loading only whole pages even when only a small part is needed.
Xanadu (in it's newest form, called ZX or Floating World) follows the
model of both Rebol and Amiga/Elate of loading only what is actually
used into memory, and loading it only once.
Another possibility for a programming project would be something like
Lifestreams. It has it's own unusual data storage design: throw
everything in one big data ocean, then send out many agent programs
to find everthing matching a set of requirements. The Lifestreams
(Mirror Worlds) type of interface could also be a frontend to the
Xanadu database, since it is also multi-dimensional. Jini, and Java
Spaces use this technology. The Lifestreams website has a Flash
presentation worth looking at: http://www.lifestreams.com/
I've often heard that for a new technology to take over, it must be
10 times better than the previous one. I believe we are beginning to
see that happen in languages; operating systems; user environments;
and data storage, retreival, and analysis models. These previously
distinct domains are in some cases merging as new tools are depeloped
to handle several or all these tasks.
Related web sites:
Lifestreams:
http://home.earthlink.net/~jackseay/jack/2techs.html
Amiga, Lifestreams and Xanadu:
http://home.earthlink.net/~jackseay/jack/AmigaNew.html
Xanadu:
http://www.sfc.keio.ac.jp/~ted/FWdesign99/ZXfrontpage.html
http://www.timecastle.net/v/xanatalk/FrontPage
Xanadu Mirror Worlds mailing list:
http://www.egroups.com/group/Xanadu_MirrorWorlds
*#*#*#*#*#*#
To design the new structures of writing for screens is a profound
issue of literary structure. It is important to provide the best
literary structure that we can, for hypertext, as the literature of
tomorrow, determines in part the new structure of civilization.
Civilization is in large part about, and around, what is written.
This is what we call literature. Literature is an endless river,
connected, like water, in all directions. Document connections go
forward and backward in time, and sideways between documents.
Scholarship and fiction, political speeches and criticism,
advertising, journalism and technical reports-- all affect each other
and evolve in a constant flow of ideas and writings. ... Ted Nelson...
Jack Seay [jackseay--usa--net]
[2/5] from: petr:krenzelok:trz:cz at: 29-Aug-2000 1:09
[jackseay--earthlink--net] wrote:
> I have been spending the last month trying to learn to create web
> pages with Dreamweaver and Fireworks. While both of these are
<<quoted lines omitted: 14>>
> (yet powerful), and has an interactive programming environment. It
> also has built-in graphics, including image processing.
Don't bet on REBOL's built in image processing. It's not enough in any
way. It has limited set of effects, it just supports only three types of
image formats, with ability to save just as an bitmap. REBOL lacks general
mechanisms. 2.0 showed the way and now imho Commodore Amiga mistake
repeats - basic success of technology is influenced by business decisions
(see REBOL/Express). There's nothink bad with /Express itself. so just let
me explain.
You want REBOL technology to step in and replace complex and let's say
bloated technologies of today Internet. You mentioned Flash for e.g. Just
one question - can we replace flash? Uh, we surely can't. We have just
REBOL/View - it's compositing engine, not drawing one. We have no sound
support too. My friend told me applying effects to pictures is just kind
of mathematical combination of two matrices. A handy set of funcs would be
probably enough for us to implement own 'wipes, probably fast enough. We
have dialecting, just imagine something like stating:
view layout/size [
button "QUIT" [unview/all]
scene [
arrives 0x100 to 200x200 %bay.jpg 0:05
appear %photo.gif 100x100 immediate
fade bla bla bla
]
] 640x480
With similar media stuff REBOL couldn't be beaten .... imagine tones of
dialects, custom effect grouped into 'wipes etc etc.
So we've got /View. It's not enough media based, but it's great. Carl
started company and has to bring money back to investors. I just wish they
had infinite resources of income to concentrate on Core stuff. I know Carl
likes multimedia, small companct apps - an engines.
Let's say I am a little bit disappointed about intermediate steps RT has
to take to "rule the world", but can't do anything about it.
What do I like about REBOL in last few weeks? - REBOL experimentals - it
just shows Core technology is being advanced (see one of the latest posts
of Holger Kruse) by very nice additions.
So ... conclusion ... it's not time to replace current web technologies,
not yet
((c) Ridley Scott, Gladiator :-))) ...
-pekr-
[3/5] from: holger::rebol::com at: 28-Aug-2000 16:31
On Tue, Aug 29, 2000 at 01:09:52AM +0200, [Petr--Krenzelok--trz--cz] wrote:
> Don't bet on REBOL's built in image processing. It's not enough in any
> way. It has limited set of effects, it just supports only three types of
> image formats, with ability to save just as an bitmap.
The next experimental version of View will have full PNG support. All PNG
formats for reading (including transparency and alpha-channel), and 24-bit for
writing.
> I just wish they
> had infinite resources of income to concentrate on Core stuff.
*grin*. Yes, we all wish that :-).
--
Holger Kruse
[holger--rebol--com]
[4/5] from: alex:pini:mclink:it at: 29-Aug-2000 16:20
>- Open Your Mind -<
Quoting from [holger--rebol--com's] message (29-Aug-00 01:31:24).
h> The next experimental version of View will have full PNG support. All PNG
h> formats for reading (including transparency and alpha-channel), and 24-bit for
h> writing.
:-9
Will it be possible, in the future, to have a CGI-script prepare a /View-like picture
and send it back as a PNG, GIF or JPEG?
Does anyone on this list have browsers *not* capable of displaying PNGs? Do you know
anyone who has one? In other words, is PNG good for everyone *now*? (I hope so...)
Alessandro Pini ([alex--pini--mclink--it])
All in good time, Captain. All in good time.
(Delenn)
[5/5] from: hen:flamefew at: 30-Aug-2000 11:28
PNG's are still poorly supported as far as I'm aware.
Especially on the more mainstream PC/Mac machines. Although browsers and
programs will claim to support PNG's, they only partially support them and
don't handle the more complex areas of gamma correction and partial
transparancies.
I believe they can sometimes crash when they meet a PNG they can't handle.
However the PNG support is improving and maybe my information is even now
out of date. If not, hopefully it will be by this time next year.
Also probably issues with .gif files compression patent.
Hen
On Tue, 29 Aug 2000 [alex--pini--mclink--it] wrote:
Notes
- Quoted lines have been omitted from some messages.
View the message alone to see the lines that have been omitted