1. Euphoria Compiler
- Posted by Nate Brooman <nateb at LOG.ON.CA>
Jul 29, 1998
-
Last edited Jul 30, 1998
Rob, why don't you make a compiler for Euphoria? Bind is too big (at least
300KB per EXE) and you just can't distribute Euphoria programs that do
practiclly nothing with that size. Look at Micro C, I can compile a 1,000
line C program (most of em are 1,000 lines 
and the EXE is only about
10KB.
How about a Linux Euphoria? I use Linux more than M$-DOS
(gasp and it's very stable and fun.
- Nate Brooman
2. Re: Euphoria Compiler
- Posted by Pete Eberlein <xseal at HARBORSIDE.COM>
Jul 28, 1998
-
Last edited Jul 29, 1998
Nate wrote:
>
> How about a Linux Euphoria? I use Linux more than M$-DOS
> (gasp and it's very stable and fun.
>
There are two (count'em, two) portable Euphoria interpreters in the
works as we speak. David Cuny is working on his Gnuphoria interpreter,
aka Project X. Inspired by David's work, I too am writing an Euphoria
interpreter in ANSI C. It compiles successfully with DJGPP for MS-DOS,
as well as the Gnu C Compiler for Unix on my ISP.
And now a question for Rob... what is the object() function supposed to
do? "? object(x)" displays x, not a true/false value like the other
type functions: atom(), sequence(), integer()... I guess types are just
functions... try this:
type foobar(object x)
return x
end type
object hey
hey = {1,1,2,3,5,8}
? foobar(hey)
? object(hey)
weird, huh?
But an object() function would be useless because everything is an
object, so it would always return true. Boggles the mind :)
Later,
--
_____ _____ _____
________ /\ \ /\ \ /\ \
/ \ \ / \____\ / \____\ / \____\
/ _ \____\ / / ___/_ / /____/ / / ___/_
/ / \ |____|/ / /\____\ / \ \ / / /\____\
\ \_/ / / \ \/ / ___/_\ \ \ \ \/ / ___/_
\ /____/ \ / /\ \\/\ \ \ \ / /\ \
\ \ \ \ \/ \____\ \ \ \ \ \/ \____\
\ \ \ \ / / \ \____\ \ / /
\ \____\ \ / / \ / / \ / /
\ / / \ / / \/____/ \ / /
\/____/ \/____/xseal at harborside.com\/____/
3. Re: Euphoria Compiler
Pete Eberlein writes:
> And now a question for Rob... what is the object() function
> supposed to do? "? object(x)" displays x, not a true/false value
> like the other type functions: atom(), sequence(), integer()...
> I guess types are just functions...
In 2.0 and earlier there's a bug in object(). It should always
return 1, but it doesn't. I came across this bug
and fixed it a few weeks ago. No one ever
reported it. object() wasn't even listed in LIBRARY.DOC
(until I added it just recently).
> But an object() function would be useless because everything
> is an object, so it would always return true. Boggles the mind :)
That's true. It exists just for completeness.
Perhaps in the future it might
be used to distinguish initialized variables
from uninitialized variables. Just a thought - I'm not
really interested in doing that.
Regards,
Rob Craig
Rapid Deployment Software
http://members.aol.com/FilesEu/
4. Re: Euphoria Compiler
Pete Eberlein wrote:
Inspired by David's work, I too am writing an Euphoria
interpreter in ANSI C. It compiles successfully with
DJGPP for MS-DOS, as well as the Gnu C Compiler for
Unix on my ISP.
----------------
so.... it's PEU... is that Portable euphoria interpreter,
or Pete's euphoria interpreter... orrrrrrrrrrrrrr......
simply 'PEU' pronounced "peeeee-ewwwwwww" as in stinky?
:)
couldn't help being juve'.... i sowwy :)
i am greatly looking forward to linux euph, Xwin euph...
opens more possibilities than it closes doors...
linux supports multi-processor heterogeneous networking.
*engwish pwease*
<english>
you can have a network of ~oo (approx infinite)
computers linked together over a net (ipx/tcp 4 ex) that
(get this) share memory and processor power...
someone talking about 8GB structures? no prob... no limits
via causeway cuz its circumvented...each machine could
have 1gb mem, get 8 machines, you've got a 8gb mem pool,
and *eight processors* working to solve your problems.
</english>
anyone seen the report that the nist built a multi-proc
symmetric machine this way-- linux box becoming one of
the worlds fastest computers for $150K US in off shelf
parts... blows away multi-multi-millon $ machines...
sweet...
win95 programming giving you headaches? Xwin...KDE...
good potential answers...
the best part? with the status of emulators today for
linux, (from what i've seen in my research) there is
about 90% compatibility for win3.1 programs off the
bat (using "wine"--windoze emulator for linux) and
a very high chance that a win95/nt prog will run
directly if it doesn't use directX.... and some
directX programs (about 20% of the ones tried)
actually run...
anything else that doesn't run, or doesn't have a direct
replacement (i.e.. need browsing? XwinNetscape... need
a WP? spreadsheet? image prog? all those types of
programs are ready and waiting...) we can then
program in PEU or GNUphoria....
can't wait... can't wait... (hehe, already got a 540M
partition, linux kernel recompiled, started getting
Xwin current version DL and installed...vroom!)
</endbabbleing>
got long winded/excited...sowwy...
take care all-- Hawke'
ps. to who wrote about XWin/320x200/s3Virge prob...
ditto... same setup here... same prob... working on it...
i think the solution is XF86.3.3.2.2 (xwin 3.3.2 patch2)
it has much better default svga support for us...
been DL it in spare bandwidth time, but work calls often...
will let you know if that update was worth it b4 you
DL for nuffin...k?
