Mailing List Archive: 49091 messages
  • Home
  • Script library
  • AltME Archive
  • Mailing list
  • Articles Index
  • Site search

[REBOL] Re: starting to be really late!?

From: petr:krenzelok:trz:cz at: 7-Jan-2004 20:06

Jason Cunliffe wrote:
>>Is rebol ever going to catch up once again? We need Rebol with more >>media capabilities, better memory consumption characteristics and even >>better with general virtual machine in 2004, or probably never .... >> >> >
First - let me add that I almost completly disagree here. So, here we go why:
>While View is nice and capable of more, I really think Rebol's best >strengths are not graphics/media, unless RT got massive investment and had >some very different kind of marketing director working for them. imo Rebol >is best for invisible stuff. >
It has to be defined what does invisible mean? Apache mode? Yes. Few other unix clones - no-way - waste of time. I e.g. did not understand, when Reichart told me, that he would like to see View for Solaris. Why? Because of FtpGadget? Excuse me, but now I call that total waste of energy and resources. Would it bring money to RT? Probably yes - but imo only short time. And Carl created Rebol because of some vision imo. And as for vision - sooner than later the world will be full of wireless mobile devices so I ask openly if we are going to stay behind or not. Something like View for Solaris or even systems like BSD is completly secondary point, if you think in some 3 + years. The only good thing is that such devices are going to be faster and faster, so maybe some Linux View version may run on them.
> As things stand I believe it would be a >relative disaster for RT to spend too much more effort on /View devlopment. >
disaster? And that is why View 1.3 is under development? That is why direct draw modes were/are planned? That is why Carl told us that multimedia is his second name? :-) But maybe first - the problem with me disagreeing is - media - what is media? I don't think rebol should turn into media player - I just mean - ability to work with gfx faster.
>Better to integrate and use make-swf, >
what? Why? :-) What is the single advantage of making flash in rebol? That you would be able to run your "rebol" app inside the browser? Or where can I find flash and not rebol? Unless there is one-button-press conversion of rebol VID app into flash, it still means to program separately for View and for Flash ...
>and work on tight useful Flash >integration. Look at the cool swf2exe prodcuts which have come out >Screenwaever, Northcode, FlashStudioPro. That's easy money and opens up >desktop flash applications to compete with Rebol very quickly. FlashStudio >pro adds 400 fscommands including socket, http, ftp, mail, file i/o.. What >more does one want or need? Even a Rebol-based swf2exe for Windows, Mac and >Linux [in that order] could be a very successful and useful product. And a >viable way to introduce Rebol to much larger and visible audience . >
That can be done without any need for Flash.
> The >mobiles will follow. > >Rebol Projects worth investing in? > >How about a really *great* http server in rebol with integrated tfp, xmpp >[Jabber], mail etc.? >
I can install Apache in 3 mins, configure in 2 - so 5 min to run profi http server on my desktop machine. You mentioned Jabber - so, OK - that is something i agree with, if you think something like more advanced IOS server, with direct communication and also polled communication, ability to use some file-sharing principles, sync to multiple servers ... kind of grid of various protocols, that would be cool. I can imagine rights, roles, rules, workflow engines, which would allow you to model various scenarios - that is REAL business oportunity, and it is a pity RT dropped IOS development - it could be much better.
>How about a really good mod_rebol component for Apache? >
I agree - but then - fast-cgi is not all that bad .... and you also mentioned money - how does RT makes money with that?
>How about rebol for bluetooth/wifi enabling digital video and still cameras? > >Let the others build media engines, let rebol be the modular engine they use >to communicate. >
So now I hope I reached the point which explains it all. I should not mention media - as by media I mean - fast, smooth refresh/movement of elements around screen, fast effectcs, nothing more - I in no way meant RT should duplicate works of others and make native divx, avi whatever codecs inside. I fully agree with integration. In fact - my pov is, that RT should done the most general things ever and the rest should be done by user-base. My top priority is: - general VM dialect - which would allow for to use subset of commands, would be fast and would not require to write external modules in C, as they mean end to cross-platform scripts unless ported to certain platform. TAO did it right with Intent imo. btw - on AltME world Carl told us, that such kind of VM would take one person average month to code. Imo worth real consideration then. - if not general VM, so dedicated VMs then - we call them dialects - currently we heave 'secure, 'remove-each ... we want at least other for fast gfx handling. - language extensions as proposed by RT ... but then who knows what do they mean, no further info was released yet.
>Anyway the mobiles will move moer and more to become embeded linux type >devices I expect. so better to enjoy the broad media sdks avaialbel there >don't you think? >
Yes, it may be right, but I want externall handled image data to be displayable in View face! Period! Are you really interesting in non View windows? Then you can just use kind of Core/Pro product. Cyphre showed nice demo of OpenGL demo - but because we are not able to give face a pointer to external buffer of certain size, we have to use memory copying, which slowes whole thing down. but it was entered into View 1.3 tracker. I think that sometimes the difference lays in details. And that is why we ask for more generalisation of language, so that RT does it, and opens oportunities for others to step-in with feature usage. You may be interested in apache + sql, I may be interested in IOS kind of system, but Cyhre may be able to build Scala killer in half a year, if certain features are implemented/some things changed.
>It is soo much work to develop viable media capabilities - sound video etc. >For it ito be sucesful has to embrace many formats for improt and export. >
Ah, now I completly see - it is misunderstanding - I didn't meant media as a real media - just a bit more faster and smoother blitting, transitions for nicer GUIs ... -pekr-