World: r3wp

Geomol, I agree. And I think it will happen. The first company breaking 
out of the stupid "web-app" path, delivering a product that adds 
so much more value for customers that they use, gaining market momentum 
will win the game.

Robert, Geomol - I am not sure you are right, even if I do understand 
your message. The thing is, that the browser is not just html interpreter. 
It is nowadays a container for native technologies. JS is a glue. 
So, if HTML 5 adds video, it is done in native level, hence there 
is no reason why it should be slow, etc. It is just that so far guys 
did not agree upon media formats.

I think they will never get all the different technologies playing 
together in such a smooth way. It's just to complex, hence to expensive. 
If you can provide a solution in 1/5 of time and budget, showing 
everyone it works as good or even better, youwill win.

I think, that for the sake of the world, it is a bad news, that JS 
is getting better and better. Guys trying to claim web-apps can be 
real-time, are almost true. However - hopefully no matter how they 
try, JS (web) based app will be crappy stuff even for few years coming, 
no matter how fast they get it running, as the problem is overall 
complexity of the whole web aproach ...

See, Graham, this is why there wasn't a JavaScript group before and 
there isn't much discussion in it now: People will discuss JavaScript 
everywhere else, regardless of group topic :)

I can imagine JS group used for real development efforts, JS framework 
chat, and as such :-)

The thing is, that the browser is not just html interpreter. It is 
nowadays a container for native technologies.
And

The Gazelle technology would be slower than the current IE, holds 
an entire operating system for sandboxing.


See, the browser is becoming the OS. I just hope, people will someday 
realize, this is a bad idea. It's better to run the application in 
the OS than running the application in an "OS" on top of a browser 
within an OS.

Really? I thought that's a pretty cool idea, and than wrap it inside 
a VM running on a host OS.

:-)

:-)

It shows that nobody has thought of something useful to do with all 
the CPU power nowadays

Geomol, can you point me to the time when people realized that Windows 
was a bad idea, or that MS Office was a bad idea, or that KDE or 
Gnome were bad ideas?

Henrik, for IE to gain back market share, it has to make the existing 
sites work better than they do now; future sites do not count (FF 
had to make IE sites work after all). They can't do that; their exit 
strategy would be Silverlight (which for sure allows for better web 
apps), but I don't think they're going to win that war.

I just hope, people will someday realize, this is a bad idea. It's 
better to run the application in the OS than running the application 
in an 

OS" on top of a browser within an OS." - Ah, so REBOL was a bad idea 
after all. Who wants to run a OS (REBOL) inside another OS? Better 
to use the native services the OS provides. :-)

REBOL was (is) designed with its primary principle in mind, from 
the very beginning. The web served completly different purpose, and 
is patched over the time. The Web WILL win. We are now facing transition 
era, where to cover its inefficiency, ppl are producing tonnes of 
JS libraries, to make day work easier. And if it gets even more complex, 
you will get it shielded by visual tools, which will allow you to 
do web stuff more easily.

But, there is still a chance. REBOL was made for what it was made 
for - lightweight distributed apps. We are scrapping "the bad", and 
remedying situation with R3. Does web in general throws away bad 
ideas? Mostly not. Where situation will get more interesting is the 
compatibility. In the past, there was IE, then FF, and Opera. Now 
Apple and Google entered the game, with more significant browser 
market share. We will see, what headache all those new browser cause 
to developers - I mean - various incompatibilities between browsers. 
Not to mention mobile browser incarnations, which are often separate 
projects ...

Ah, so REBOL was a bad idea after all. Who wants to run a OS (REBOL) 
inside another OS?


As I see it, REBOL is a language with some OS-related parts. The 
Viewtop for example. But that never became a huge success. Maybe 
because it's not such a good idea with another desktop on top of 
the OS desktop?

Making REBOL a real OS directly on the hardware might be a good idea 
at some point.

Geomol, the Viewtop has nothing to do with it. REBOL file name conventions 
are different from those of the host OS (%/c/file instead of C:\file), 
it does not give you access to the OS functions directly, and it 
has its own graphic rendering engine that does not use hw acceleration 
of anything (faces in R2, GOBs in R3).

if it was just a language, you would be using directx, MFC, etc. 
on Windows, QT or the GTK or Linux, and Cocoa on OSX, regardless 
of whether you're writing the Viewtop or a little game.

Gabriele - wrong - Qt is not native toolkit. You consider it being 
one, just because it is cross platform and widely adopted? By naming 
Qt, you could admit, that even View is native one, whereas VID is 
not. It does not matter that View uses its own UI elements, as much 
as GTK on Windows looked in the past (when I checked) non-native 
too (dunno if they changed it or not ...)

So, what's the native toolkit on Linux?

How do you define a toolkit "native"?

I guess, one that doesn't pretend to be a toolkit :-)

this seems to be a pretty huge announcement - http://blog.wolfram.com/2009/03/05/wolframalpha-is-coming/

Reichart, does your friend at AdaptiveAI know about any of this do 
you think? Is he a friend of Stephen Wolfram by chance?

Adrian, there was an article about it that points out how it's not 
really AI, but rather a system that can compute answers. it is surely 
a breakthrough and it's going to change our lives maybe... but that 
is sort of like google, in that it is a support tool for intelligence, 
not intelligence. that is, it is a tool that intelligent beings use. 
so, AI systems will find things like Google and Wolfram Alpha extremely 
useful, because they allow independent learning (without humans having 
to teach).

I didn't mean to imply that it was AI - I don't know if what AdaptiveAI 
uses "real" AI despite their name (though it's hard to tell from 
the info on the site). I was just thinking that they might use similar 
approaches and was curious to know if Reichart's friend who works 
there was possibly an acquaintance of Stephen Wolfram's.

OK, I'll sit and wait for the revolution to happen :-)

Wolfram has to have good PR - even our main Czech news portal is 
reporting on it :-) There was even link to already existing - http://tones.wolfram.com

I wonder if this is going to be a parser similar to Inform-7 ... 
without the gaming fun.  ;)

Don't know, I'm having dinner with him soon, I'm sure it will come 
up.

Mazatech release AmanithVG 4.0, supporting OpenVG 1.1 http://www.amanithvg.com/project.html

Wouldn't this be a good library to be used in replacement/combination 
with AGG?

I tried their Tiger example. GLE crash on my system. SRE give me 
138 fps at full quality on my system.

Is it so small as AGG? Does it have good license for us? I thought 
that Amanith is indeed based upon AGG, but maybe I am wrong :-)

IIRC Cyphre had some old comparisons, and AGG was the fastest one. 
There is still some place to optimise stuff. E.g. bitmap/image node 
caching is not done in View yet ...

The GLE version does not crash here, but is 2x slower then the SRE 
(with the best quality). It really looks that the hardware acceleration 
is not the best way.. at least if you don't want to make another 
Quake like shoot-them-all clone.

Studies sponsored by NASA show that an embedded software bug introduced 
in the requirements phase is 130 times more expensive to fix during 
integration and 368 times more expensive after rollout of the embedded 
device.

Do they give examples of such bugs?

I found it on a blog, so I don't have the original article.

The wrong semicolon in the FORTRAN program that made an explorer 
miss Jupiter by 100,000 kilometers is a good candidate

I'm not sure this is "tech", but sort of....


http://www.techcrunch.com/2009/03/26/california-may-ban-black-cars/

http://www.google.com/ventures/

Google looking for apps to invest in ...

I'm going to put a proposal to Google.  What would one expect to 
pay for a full time top-notch REBOL programmer?  A web designer? 
 Some to do documentation?

Graham: I have a old Amiga buddy working for Google/I could ask him 
what the scale is

Alan, I was meaning market rates and I guess the only one who knows 
that is Reichart!  ( and his employees ... )

Sounds like they're following Paul Graham's VC model. 


How much people charge depends on a lot of things. Where they live, 
whether it's their main gig, and how long the contract is being important 
criteria. I've been quoted from USD$30/Hr to USD$125/Hr when I was 
looking for contractors.

It is a complex formula.... 


What hours total, length of contract, freedom to work which ever 
hours in the window, etc.

You might make $500 an hour, but you get one job a month doing this. 
 While there is an open slot to get $10 an hour, you can work when 
ever and where ever you want, and the work is stupid simple.  you 
might take both.