1. Re: more about serious
Norm Goundry wrote:
It is just my opinion,
>personal and professional business-wise, that the use of programming
>language directed towards PCs of all types, has hit its high-water mark, and
>will soon become second-tier to the combined might of other, lesser thought
>of devices. If one looks at the total amount of memory sold right now,
>embedded devices take the cake. This will only accelerate over the next few
>years.
There is little doubt of this. The only question is what is an interpreter
likely to be doing on this type of system. Other than as a scripting engine for
games, I don't believe that any interpreter will serve to drive high resolution,
real-time, interactive, 3d graphics programs. Interpreters also are unlikely to
form the kernel or system services of any OS. For most business purposes,
probably yes. Now if someone wants to use these engines as Internet appliances,
we have a chance to do something. If someone were to port a copy of Opera and
stick Euphoria onto it as CGI and general programming tool, now we have something
that I consider doable. Opera is on the point of appearing on about five
different environments. A couple more wouldn't be too much of a stretch. It could
provide the user IO that Euphoria really doesn't have. Since Opera already takes
some Netscape plugins, it wouldn't be too much of a stretch to attach Euphoria to
it.
> Consoles (game machines) are a case in point. The Sega Dreamcast,
>for example, is a proper 128 bit machine (64 X 2 data; 128 X 1 graphics),
>and has a two proper operating systems. One of these is Windows CE, said by
>it's Master to be especially tweaked for the Sega box. The other OS is
>Sega's own primitive Game OS, though it still is a legitimate system. Both
>of these stink, of course, to put it bluntly. Especially the CE, as it is
>only a 32-bit setup. The tools that one would use here are 'C' or
>assembler, though CE will allow for use of Visual Basic and DirectX 7.X.
>The cpu for the Dreamcast is a Hitachi Super-H, commonly referred to as the
>SH-4, and it is a very powerful RISC chip. Metawerks CodeWarrior and
>Greenhills 'C' compilers and toolkits have been 'bent' towards programming
>this particular chip; both with Assembler and C/C++. The official status
>TDK, licensed from Sega, giving one the right to design for the Dreamcast is
>8000 dollars US. In contrast, the one for Sony's Playstation 2, which isn't
>even on the shelves until well into next year, costs 20,000 US. These are
>significant outlays of cash, the minimum needed to even sit down at the
>poker table with the rest of the production houses.
>
>Specifically to Bernie: As to APIs, well, you tell me. They do, of course,
>but not in the way that we are commonly used to looking at. Almost all
>cpu's are built from the start with 'C' in mind (btw, don't get me wrong
>here-- I don't care for 'C') and anything from the assembler on up, into the
>next-up compilers, is built on and around 'C'. This is one of the
>'commonalities' about the RISC family (though it is a very tenuous one) that
>can be exploited if you are very clever. And this is where I wish that
>Euphoria could break away a bit, and be directed towards something other
>than the constant drain and strain of Win and Lin.
The drain is from Win and DOS. The Linux effort will get Euphoria into fairly
portable C code form and that can only benefit what you are heading towards.
>
>And may I say something back to Everett Williams here: you accuse me (and I
>don't really take offence; I know you mean well for everyone who EU's) of
>'recruiting', and I should take it off-line. I am a legitimate programmer
>who makes a living writing code. I consider Euphoria to be a serious
>language, and I use it as such. As to 'sucking away talent and time to a
>commercial project', all I did was state MY OWN CASE where I have put my own
>butt on the line for what I think is the not-to-distant future of computing.
>And if I remember rightly, Euphoria IS a programming language, and those who
>use it are programmers, not dilettantes. It never was my intention to draw
>people away from the furthering of EU; and likewise I will reiterate that
>discussing about hardware (RISC; Game Consoles) is certainly a valid subject
>for discussion... the only problem is that people take offense at talk of
>Change. In this sense, this is Change happening right now. I mentioned
>that this would be all 'Open System'; and I will be spending over 30,000
>dollars to even get allowed to do it legitimately (and a lot more than that
>besides) so that others can get it FREE. This must mean something to some
>of you out there. Everett, I am old enough to also remember Mme. Lee, so
>that makes two of us the same somewhat. 
My point was that if you did not have legitimate intentions of using Euphoria,
you should move your discussions off the forum. I applaud the "Open System"
aspect of what you are proposing, even if I am uncertain of your heading. Caldera
has already indicated that they are working on thinned down versions of Linux for
appliances and hand-held devices and others have already produced a prototype 64
bit version for Intel's new chip. What benefit is there in repeating all the work
that has been done in that area? Even a Euphoria compiler would require a
complete reworking of the module to module interfacing and a major rewrite of the
data structure area before further work as large as an operating system could be
attempted. The garbage collection in that case would be a "concept" that would
have to be completely rewritten for the non-dynamic case. That is not to say that
some hybrid of compiler and interpreter with the best features of both could not
be produced, but in what time-frame. The Dreamcast is out now and the Sony is
coming next year. Tabula rasa(clean slate) does not seem promising as a way to
meet the timelines inherent in what I just described.
>
>(Back to Bernie) About creating a common library. It seems that at a
>certain level, it is possible to do this, even high language enough into
>'C'. And as this can be shown to be the case, then it would be also
>theoretically possible to cobble a restricted portion (at first) of EU to do
>the same. But what this would mean would be for several capable people to
>try and forget about how EU exists in the conceptual world of Win API, and
>visualize how a completely different environment might look. RISC is very
>much not Microsoft territory, for one thing, and EU has demostrated that it
>can make the transition from there to Linux. But these are still using the
>same Intel x86 base, so nothing has changed much there. So the fact is that
>the cpu's (CISC vs. RISC) have nothing in common at all, except this
>connection through 'C'.
>And yes, I do have propriety info about the Hardware.
Don't fool yourself. The current Pentiums and Athlons are RISC under CISC and it
would not take much to peel off the CISC layer. I believe that it can be accessed
to some degree now. The new 64 bit units are very much RISC with CISC in
parallel.
>
>The various players (nintendo, sega, sony) do NOT have an 'intended' common
>interface, but as I mentioned above, this very much depends on a
>purposefully carried out difference on the propriety level-- but an awful
>lot has been overlooked about this being the same on an 'unintended' level.
>Much food for thought in this...
>
>As to there existing a common specific OS, the answer would, once again,
>seem to be 'certainly not'.
>
>And finally Bernie, your last question, about there being a common low level
>instruction set... the tantalizing answer is YES, but I will not go into
>that because of considerations with my partners. Perhaps I should clarify
>that a bit: the consoles do not, but the chips do, and this has been not
>been forgotten, just overlooked.
>
>And one of the reasons I would like to see EU there is... it has garbage
>collection (which takes out 'C' any day in my humble opinion).
>
>Sorry if I have bored anyone, or enraged others, but that is the way that I
>am. I will try and answer whatever I can (or am allowed to under my
>contract), so keep it coming! Just remember that sometimes Food For Thought
>needs a bit more chewing than as usual.
>
>Regards,
>Norm
I have layed out some serious ideas above, and would be interested in your
thoughts. I don't think that any here fear change...just poachers. If there is a
legitimate use for the language, then let it be. It just seems that the language
is at such an immature state and those with the copyrights so few in number, that
getting where you want to go would require either their full attention or their
contractual release of some significant rights to you to allow you to use your
resources to extend the language in the directions that you seem headed. Too much
sentence, that, but understandable, I think 
Again, that would seem to
necessitate your communication with Rob and Junko before going much further. So
far, they don't seem enamored of the open source model that you seem to be
espousing in this project. I truly think that there should be some response from
them on this forum before this proceeds much further for both practical and
ethical reasons.
Everett L.(Rett) Williams
rett at gvtc.com
