Rebol better than EyeOS

You could say it's been tried with REBOL, be it the ViewTop or IOS. It's just not been attempted within a browser window yet - for obvious reasons. There's lots of desktop-within-a-browser attempts out there and they all suck because they're all too slow on anything but the latest hardware and fast broadband. REBOL has the potential to put a fast desktop within a browser window on slow hardware using dialup. And if it's ever available on all the main browsers and all the main OSs, no doubt someone will attempt this. But... But it won't catch on unless REBOL goes open-source. That's just how it is these days. -- Carl Read.

But good-enough is all it takes for a technology to become pervasive, and web-browsers were already pervasive before REBOL was released. REBOL can't replace them, (short of a better browser being written in REBOL itself), and people won't want yet another window open on their screen, (which, if a desktop, you'll have to open before you even get to the app you want to open...), so working within a web-browser window is the only sensible approach.

This is a discussion that's been going on since REBOL began. Why don't web-hosts have REBOL/core available like they do PHP or perl or Python or Ruby on Rails and so on? If it's not because they're open-source and REBOL isn't, then why? What RT needs to decide is whether they want REBOL to have some direct influence on the future of computing or just keep making however much money they're making from it now. (Note that in the past I didn't support the calls to open-source REBOL. But that was in the days when they used to release the latest versions for all the platforms they were then supporting in the same week. And they were supporting many more platforms then than they do now. The pragmatist in me says that since your method of marketing isn't working and the competition's is, then your method's wrong and there's is right.) [small] Yes Joel, you were right. ;-) [/small] -- Carl Read.

I think such languages better fit server-side work. I think Rebol should work over rich-clients. E.g.: try to write with Rebol an application like Zope, or like Apache Tomcat, or JBoss, or IBM Web Sphere, etc... I think Rebol has not yet enough power on server-side. More: too many people like imperative-programming. More: there are already a lot of server-side languages. Client-side: please tell me one example of good programming language to create a full-featured, multi-platform, not too expensive (in terms of resources), fast, small to be distributed (in terms of runtime), etc...I think Rebol could be a killer-application for rich clients. More: possibility to access ALL the resources of a local PC (devices, DLL, api, registry, config files, cache, etc...). I develop programs since when I was 12. I know basic, Pascal and Delphi, PHP, Python, Assembler of 6502/6510, Z80, MC68000, C, Java (great n server-side, but terrible on client-side ;-) ) but only a browser web is a rich-client multiplatform killer app. But browser is VERY limited, VERY incompatible, etc... It is good for documentation. I know very well how is difficult to create a good rich-client app with browser. Actually, in ms windows environment (in the company where I work as senior analyst/programmer) I use Delphi to create single executable, full-featured rich-client apps, and the server is built in java/JBoss. Connection is performed using web services. I think Rebol could replace Delphi-side, not Java-side. --Alessandro On 6/17/06, Carl Read <carl-cybercraft.co.nz> wrote:

I guess that it's just because they are much more popular than REBOL.

[...] IMHO, the only real thing that REBOL lacks for state-of-the-art server-side apps, is multithreading support (multiprocessing with memory sharing). But, even without that feature, you can already write very good and efficient server apps (if you have a good framework). -- DocKimbel