NetNews Usenet Archive 1993 #1/NN_1993_1.iso/spool/comp/sys/mac/programm/20948

Text File | 1993-01-07 | 6.1 KB | 137 lines

Newsgroups: comp.sys.mac.programmer
Path: sparky!uunet!mcsun!chsun!bernina!bernina!neeri
From: [e-mail] (Matthias Neeracher)
Subject: Re: Sozobon (or other free C's): would this strategy work?
In-Reply-To: [e-mail]'s message of Thu, 7 Jan 93 17:58:34 GMT
Message-ID: <[e-mail]>
Sender: [e-mail] (USENET News System)
Organization: Integrated Systems Laboratory, ETH, Zurich
References: <[e-mail]> <[e-mail]>
    <[e-mail]>
    <[e-mail]>
Distribution: comp
Date: Thu, 7 Jan 1993 23:23:47 GMT
Lines: 121

In article <[e-mail]>, [e-mail] (Radu) writes:
> quoting from [e-mail] (Matthias Neeracher)
>
>> Another reason against porting gcc is that there are already at
>> least two powerful, well-integrated compilers around. I doubt that you could
>> get gcc to compile even as fast as MPW C, let alone Think C, especially if
>> you consider that gcc needs a separate assembly step. If you add the
>> necessity of writing a decent user interface for a standalone compiler, I
>> doubt that people will be willing to spend a lot of effort in writing the
>> 3rd fastest and 3rdmost compatible C compiler for the Mac
>
>>> But I think a lot of people would benefit from a port,
>>> given the large numbers of Mac users.
>
>> My parents use Macs, too, but I doubt they would benefit from a port.
>> Seriously, though, I think the target audience is not that big. gcc is too
>> big for amateur programmers, too slow for many non-amateurs, and some
>> programmers wouldn't like to use a standalone compiler (I want my MPW :-)

> I take that whole post as a direct insult!

Wasn't meant to be one, but to paraphrase Guns 'n' Roses, if you feel offended,
you might want to check out rec.music.newage :-)

> Sure your parents might use a mac, but so do I, and i'm not by any means
> related to you.

My point was that many Mac users have no intention of programming a Mac, so
the target group for gcc would be significantly smaller than "the large number
of Mac users".

> Just because you have
> the money to go ahead and buy MPW or Think C, doesn't mean that everyone can.
> I for one can't even come close to spitting out the $150 for Think C.

I doubt that.

> I'm
> currently a student in colege so that could explain the lack of funds.  On
> the other hand I'm using a Mac + which i bought for less that 200.  So
> before i spend more money on a compiler than I spend for my comuter i'll
> much rather buy another computer.

This is plausible, but may I point out to you that there is no way to expect
that a gcc port would run on a Mac Plus ? (You *might* get it to crawl along
on a 4M Plus, but I don't think that anybody in 1993 would want to experience
something like this).

> I hapen to agree with FSF drive against apple,

So why did you want gcc on the Mac ?

> I'm all for it, and i think they have a good reason for it as well.

I don't.

> Apple has had the bad habbit of ripping people off since the bigining
> of time.

You got it all wrong, it was the snake that led Eve astray, the apple was only
a passive participant :-) You are of course entitled to your opinion, but
then there is little point for you in hanging around this newsgroup.

> recently i was debating buying a  mac CI for 2000, noting how much prices
> came down, but then I saw a 486 66 for only 1000.  Now lets face it, a 486
> 66 will cream a 68030 25.

Yup, especially if your preferred hobby is to run big bang simulations. If
it isn't, you should judge based on the applications you want to run and
on the reputation of the manufacturer (And the '486 might still come ahead).

> Put linux on it, and your set, not to mention
> that i'll have all the code that has ever been written for unix.

Except for a few lines of proprietary code, that is. Linux seems to be
real nice, though.

> I strongly feel that someone should write a port for GCC, something that
> is independant of MPW, or any other comertial product.

- You haven't made the slightest point why. The FSF is against it, and
  it wouldn't run decently on the low end machines that the target segment
  might own (Like my own machine, BTW). You can't do real mac programming
  without a *lot* of manuals, and a new set of Inside Macintosh is probably
  more expensive than Think C.
- Many programmers (such as me) would think it sort of cool to have gcc, but
  I doubt that is sufficient motivation for a standalone port.
- It's close to impossible to make a good compiler port *totally* independent
  from any commercial products, unless you want to spend enormous amounts
  of time on writing and maintaining Toolbox header files and glue code. I
  suspect that a gcc will mainly be a second compiler for people who own
  already Think or MPW.

> I can't make it happen since I don't know enough, but i hope someone
> that knows enough can and will do it.

This was another point I was trying to make. Those who'd most like to have gcc
don't have the resources to port it, and those who have the resources lack the
time and motivation.

Stan has made it clear that we will soon have a running MPW version of gcc, and
somebody *might* do a standalone version of that, but it's close to impossible
that you will get free headers and libraries with it (legally, that is).

> Make it happen!

To end my posting on a somewhat more positive note, there is an ANSI C compiler
frontend called lcc (lccfe on princeton.edu). There is no freely available
backend for the 680X0, but decent documentation on how to write one. lcc is
rumored to be less resource hungry and just as well optimizing as gcc. If
anybody is considering the "gcc lite" strategy that was mentioned in this
thread, have they thought about lcc ? It would, of course, not change the
header/library problem in the least, but it might alleviate a lot of speed
and memory problems.

Matthias

-----
Matthias Neeracher                                      [e-mail]
  "Macintosh System 7.1 -- just say no.  If I'd wanted to pay for OS
   upgrades, I'd have bought an IBM." -- [e-mail]