Wildman project - PS3?

I am new to the list, and very new to REBOL (waiting for version 3.0 to begin working with it). The Wildman project caught my eye sometime ago, and while I know it is underwraps, I would make a suggestion. The PS3 console is unusual, it has a Sony's Game OS, and a vacancy - that is OtherOS. In short, because of this it is a general piece of HW unlike any other. I believe current OSes are over the top, far too much doing far too little, for no real benefit. I also believe very strongly in task specific applications and hence the critical importance of a scripting interface to bring various plugins into useful combinations. I am also in favour, which seems attuned to many of the directions REBOL is taking, in generalisied compiled code engines (plugins) bound by scripts into multitudes of different applets and mega applications, modifable and adaptable rather than being cast permenantly into bloated every-function applications found in present off-the-shelf products. Two things cross my mind, one is making REBOL an OS for otherOS on a powerful fixed architecture Cell console. The second is bringing in a "virtual" OS ontop of Sony Game OS loaded as a "game", a game that allows for productivity applications to be run in otherwords. The Cell processor seems ideally suited to a small script interpreter like REBOL. Greg Schofield Perth Australia

Hi Greg, gsina> The Cell processor seems ideally suited to a small script gsina> interpreter like REBOL. I thought the same thing when I read an ACM article on it not too long ago. -- Gregg

Thanks for your reply Gregg, always it is awkward making a first post on a well established and specialist forum. The more I have thought about the Wildman project and read about REBOL, the more interested I become. I am in a position where I cannot disclose some thoughts as fully as I like (no big deal I do some consultancy work with a very small software firm). So please bear with my odd suggestion. Wildman suggest a minimal set of OS functions - basically a very good startegy. However, what I am going to allude to is a significant low level change - re filesystem. The heritage of computer development is locational identity of files. Another old hangover is the uber dot identity system - ie .r .reb etc. The problems with these rather ad hoc solutions from the past are compounded when storage and the amount of data reaches the levels it has, it is further complicated by the web, and the fact that moving a directory can alone lose information amongst the background noise, if not for the designers, then for those trying to access something moved elsewhere. The REBOL header is not a bad basis for a file system (or an XML version of it). That including a mandatory ID along with a Plain English title (a little like URL names and turples), might not be a bad move. I hope people think about this suggestion, there are many implications, expeically for ordinary users, but also for simply navigating through huge amounts of often fragmentary data (much of which gets hopelessly lost). Greg Schofield Perth Australia --- Message Received --- From: Gregg Irwin <gregg-pointillistic.com> To: greg.schofield-iinet.net.au <rebolist-rebol.com> Reply-To: rebolist-rebol.com Date: Mon, 6 Aug 2007 09:10:30 -0600 Subject: [REBOL] Re: Wildman project - PS3? Hi Greg, gsina> The Cell processor seems ideally suited to a small script gsina> interpreter like REBOL. I thought the same thing when I read an ACM article on it not too long ago. -- Gregg

Hi Greg, gsina> Thanks for your reply Gregg, always it is awkward making a gsina> first post on a well established and specialist forum. No need to be shy or awkward here, this is probably the friendliest community I've ever been a part of. gsina> The REBOL header is not a bad basis for a file system (or an gsina> XML version of it). That including a mandatory ID along with a gsina> Plain English title (a little like URL names and turples), gsina> might not be a bad move. Agreed. File systems are including more and more metadata themselves as well, so I think it's important to consider how people use file systems today, as opposed to 20 years ago. Also, because REBOL is all about semantic information exchange, you can consider files to be a type of persistent "message"; the idea being that messages are processed and dispatched based on header information, and you could do the same thing with files. -- Gregg

Gregg: Agreed. File systems are including more and more metadata themselves as well, so I think it's important to consider how people use file systems today, as opposed to 20 years ago. Also, because REBOL is all about semantic information exchange, you can consider files to be a type of persistent message"; the idea being that messages are processed and dispatched based on header information, and you could do the same thing with files." This is very encouraging. If this logic is taken into the heart of Wildman, then things are looking very very good. I am admittedly of a minority position vis a vis current OSes, all I can see is cumbersome and unnecessary bloat. The OS tends to get in the way far too much. Long ago I considered the best course to be a script langauge, used generally (specifically as application glue) that was also THE DOS system, placing the graphic end well up the "application" environment. Maybe REBOL is moving into this gap, and against the tide of bloat. I live in hope and am feeling the gravitational pull of REBOL. Greg Schofield Perth Australia --- Message Received --- From: Gregg Irwin <gregg-pointillistic.com> To: greg.schofield-iinet.net.au <rebolist-rebol.com> Reply-To: rebolist-rebol.com Date: Wed, 8 Aug 2007 07:26:15 -0600 Subject: [REBOL] Re: Wildman project - PS3? Hi Greg, gsina> Thanks for your reply Gregg, always it is awkward making a gsina> first post on a well established and specialist forum. No need to be shy or awkward here, this is probably the friendliest community I've ever been a part of. gsina> The REBOL header is not a bad basis for a file system (or an gsina> XML version of it). That including a mandatory ID along with a gsina> Plain English title (a little like URL names and turples), gsina> might not be a bad move. Agreed. File systems are including more and more metadata themselves as well, so I think it's important to consider how people use file systems today, as opposed to 20 years ago. Also, because REBOL is all about semantic information exchange, you can consider files to be a type of persistent "message"; the idea being that messages are processed and dispatched based on header information, and you could do the same thing with files. -- Gregg

hi all, there is also a possibility that a RebolOS would not actually have a need for such a thing as a "file". It could very well be that a generic concept of data would suffice... Thus there could be no difference between the (in RAM) data you are manipulating within your app, a structured set of data and "files" and conversions of those datasets into other "standard" format structures. They could all live within the same persistent memory space in real time... no need to "save" or load your data... it just IS or IS NOT. imagine you turn off your computer at any time (no need to even log off)... you restart it and its in exactly the same state as when you shut it off... no trick, no RAM dumping to disk to boot up beyond the hardware and rebuilding of cached data like visible graphics and such. Now I am not saying this is how the project is designed or foretold by Carl S. but I would like any RebolOS to go beyond our current limited views of what is . -MAx On 8/8/07, greg.schofield-iinet.net.au <greg.schofield-iinet.net.au> wrote: