Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!ucbvax!ENG.SUN.COM!Mitch.Bradley
From: Mitch.Bradley@ENG.SUN.COM
Newsgroups: comp.lang.forth
Subject: Public domain RISC assemblers
Message-ID: <9106251924.AA20022@ucbvax.Berkeley.EDU>
Date: 25 Jun 91 15:25:37 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Reply-To: Mitch.Bradley%ENG.SUN.COM@SCFVM.GSFC.NASA.GOV
Organization: The Internet
Lines: 70

>      I have seen that many Forth assemblers are available for diverse micro
> processors (ex. 68K, 80X86, etc.).
>
>       But I wonder there is any assembler existed for RISC chips (ex. MIPS,
> SPARC, PA, etc.).

Of course there is an assembler for SPARC.  You can buy it for a reasonable
price as part of my Sun Forth package.

>      Forth public domain programs seem only favor the traditional chips (I
> suppose this is due to that chip is more available & cheap).

You suppose correctly.

>      But the modern workstations almost equipped with the RISC chips (ex.
> IBM RS6000, DECstation 5000, HP Apollo 700, Sun SPARCstation, etc.). I don't
> know why the Forth experts are still insisting in the microprocessor fields.

Most Forth users cannot afford to buy a workstation *because they give
away their work and put it in the public domain*.

I can afford to have a SPARCstation at home because I ask people to pay me for
the time I spend designing, writing, debugging, documenting, debugging,
porting, and supoorting code.

>      I saw somebody wrote a 80386 assembler, well, this is a really tough
> task. 'Cause the 80386 has a more larger instrcution set. The RISC has a
> more tiny instructions (usually < 100), and - IT SHOULD BE MORE EASIER TO
> CODE -.

Yes, it is relatively easy to code a SPARC assembler.  But to test it, you
need a SPARC machine.  By and large, the people who have access to SPARC
machines are those whose time has some value.  It's hard to justify paying
$10000 for a SPARC machine if all you are going to do with it is
to write public domain code and give it away.

It isn't much easier to document and support a SPARC assembler than a
386 assembler.  It costs the same to hire somebody to answer the phone,
ship orders, and do the accounting, regardless of how hard it was to
write the code.

>      So I would like to find any Forth black-belt guru had written the
> assemblers for RISCs, and put them on public domain. Then I can use them on
> my workstations. /* I really hate to back to PC every time when I want to
> run a Forth system! */

So ask whoever provided you with the $10000 SPARCstation to shell out
an additional $200 to buy Sun Forth from me.

This Forth guru is not about to further ruin the economic viability of
Forth by giving away his work.  In order to make a long-term commitment
to Forth, I have to be able to live.  I have to pay real money for housing,
food, transportation, and a hundred other things.  If I am going to spend
full time on Forth, I have to make enough money from Forth to pay for my
living expenses.  If my company is to create any jobs for other Forth
programmers, the company has to make enough money from Forth to pay for
their living expenses too.  It would be nice to have enough money left
over to actually promote Forth by advertising and marketing.

I think the public domain orientation was great for Forth when it was
just getting started, and the small computer field was primarily a
hobbyist activity.  Now, I think the public domain stuff is hurting Forth,
because it makes it very difficult for companies to make a living selling
and supporting Forth.  In today's global economy, it takes money to grow.

We have to stop treating Forth like a hobbyist toy, and start treating
it like a serious computer language, and that means being realistic about
economics.

Mitch.Bradley@Eng.Sun.COM