gdtltr@chopin.udel.edu (root@research.bdi.com (Systems Research Supervisor)) (02/16/91)
Has anyone noticed that (sizeof "any old string") returns 2 for
any string? Has anyone fixed this?
Gary Duzan
Time Lord
Third Regeneration
--
gdtltr@brahms.udel.edu
_o_ ---------------------- _o_
[|o o|] Two CPU's are better than one; N CPU's would be real nice. [|o o|]
|_o_| Disclaimer: I AM Brain Dead Innovations, Inc. |_o_|wayne@csri.toronto.edu (Wayne Hayes) (02/16/91)
In article <16438@chopin.udel.edu> gdtltr@chopin.udel.edu (root@research.bdi.com (Systems Research Supervisor)) writes: > Has anyone noticed that (sizeof "any old string") returns 2 for >any string? Has anyone fixed this? This is indeed a bug, since "any old string" has the type (char []). It probably thinks "any old string" has type (char *), which is incorrect. However, since ACK cc(1) (the one that comes with PC 1.3 and I think also 1.5) doesn't come with source, it can't be fixed. Well, "can't" is too strong. Apparenty you can *buy* the source from the people who wrote the compiler. Has anyone actually done this? Their address (from Andy's textbook, "Operating Systems: Design and Implementation", p. xvi): In North and South America: In Europe and elsewhere: UniPress Software Transmediair Utrecht BV 2025 Lincoln Highway Melkweg 3 Edison, NJ 08817 3721 RG Bilthoven USA Holland Telephone: (201) 985-8000 Telephone: (30) 78 18 20 And now for a legal question. If someone *did* buy the source for ACK CC, and then fixed some bugs (or maybe the original writers have fixed some bugs?), could the person who bought the source then post the binaries for the new compiler? -- "Dad, what should I be when I grow up?" "Honest." -- Robert M. Pirsig, _Zen and the Art of Motorcycle Maintenence_. Wayne Hayes INTERNET: wayne@csri.utoronto.ca CompuServe: 72401,3525
waltje@uwalt.nl.mugnet.org (Fred 'The Rebel' van Kempen) (02/21/91)
In article <1991Feb15.193820.15619@jarvis.csri.toronto.edu>, wayne@csri.toronto.edu (Wayne Hayes) wrote: > In article <16438@chopin.udel.edu> gdtltr@chopin.udel.edu (root@research.bdi.com (Systems Research Supervisor)) writes: >> Has anyone noticed that (sizeof "any old string") returns 2 for >>any string? Has anyone fixed this? > > This is indeed a bug, since "any old string" has the type (char []). It > probably thinks "any old string" has type (char *), which is incorrect. Correct. I _do_ have the ACK sources (both Pascal and C), and I asked about this just yesterday (listening again, Andy?). There are some other things I would like to fix ("static/extern" bug, 7 char identifier limit and such).... > And now for a legal question. If someone *did* buy the source for ACK > CC, and then fixed some bugs (or maybe the original writers have fixed > some bugs?), could the person who bought the source then post the > binaries for the new compiler? This is what I asked Andy. NO, I (or, for that matter, anyone) am NOT allowed to do so. Beats me why, since Unipress and Transmediair will be making the big tun of money on the ANSI compiler.... So, the situation is: - The current compiler contains some bugs that CAN be fixed, since certain people DO have the sources. However, dus to some silly regulations those people cannot post the fixed binaries. Andy, could you please explain to me why you posted the "asld with separate I/D option" some years ago? Same thing, eh? - The new, "can do it all" compiler (which, by the way, I also have) will be available ONLY from Unipress and Transmediair; they will NOT be posted here. So, forget about 1.6 if you cannot convince Andy to send you the compiler :-) Comments Anyone? Followup-To this group ! Fred This is what I asked Andy. NO, I (or, for that matter, anyone) am NOT allowed to do so. Beats me why, since Unipress and Transmediair will be making the big tun of money on the ANSI compiler.... So, the situation is: - The current compiler contains some bugs that CAN be fixed, since certain people DO have the sources. However, dus to some silly regulations those people cannot post the fixed binaries. Andy, could you please explain to me why you posted the "asld with separate I/D option" some years ago? Same thing, eh? - The new, "can do it all" compiler (which, by the way, I also have) will be available ONLY from Unipress and Transmediair; they will NOT be posted here. So, forget about 1.6 if you cannot convince Andy to send you the compiler :-) Comments Anyone? Followup-To this group ! Fred
bammi@acae127.cadence.com (Jwahar R. Bammi) (02/26/91)
In article <910220513@uwalt.nl.mugnet.org> waltje@uwalt.nl.mugnet.org (Fred 'The Rebel' van Kempen) writes: > > So, the situation is: > > - The current compiler contains some bugs that CAN be fixed, since > certain people DO have the sources. However, dus to some silly > regulations those people cannot post the fixed binaries. Andy, > - The new, "can do it all" compiler (which, by the way, I also have) > will be available ONLY from Unipress and Transmediair; they will > NOT be posted here. So, forget about 1.6 if you cannot convince > Andy to send you the compiler :-) > > Comments Anyone? Followup-To this group ! > trivial, dont use any tool that you care to, unless you have source for it. i have been living quite happily ever since i adopted this policy many years ago for any computer that i owned. in fact, the quality of most tools that i use, that come with source, in general are *far* superior to their commercial counter parts. its not hard to explain why. cheers, -- bang: uunet!cadence!bammi jwahar r. bammi domain: bammi@cadence.com GEnie: J.Bammi CIS: 71515,155
nicolas@cli52or.edf.fr (02/26/91)
> > as to 'use only software where you have the source ready': > > I posted an 68000 C compiler a year ago, and the actual version even > contains a i386 compiler. > > Except for a handful of guys (which I should call 'fan club') who use > my compilers all days, I receive little response. Most people > continue using the ACK compilers despite their weakness (the 68000 ACK > compiler does not contain many bugs -- but it is slow). > > C.v.W> > but it doesn't contain a 86 compiler (Sigh !!!)
HBO043%DJUKFA11.BITNET@cunyvm.cuny.edu (Christoph van Wuellen) (02/27/91)
I regard the 80386 as the first usable INTEL processor.
mwm40@duts.ccc.amdahl.com (Marcelo Mourier) (02/28/91)
In article <45839@nigel.ee.udel.edu> HBO043%DJUKFA11.BITNET@cunyvm.cuny.edu (Christoph van Wuellen) writes: >I regard the 80386 as the first usable INTEL processor. Until you try to do something "fancy" with the Paged MMU in it..! :-) -- Marcelo Mourier Internet: mwm40@duts.ccc.amdahl.com UUCP: ...!duts.ccc.amdahl.com!mwm40 -- -- Marcelo Mourier Internet: mwm40@duts.ccc.amdahl.com
waltje@uwalt.nl.mugnet.org (Fred 'The Rebel' van Kempen) (03/03/91)
In article <45818@nigel.ee.udel.edu>, HBO043%DJUKFA11.BITNET@cunyvm.cuny.edu (Christoph van Wuellen) wrote: > as to 'use only software where you have the source ready': > > I posted an 68000 C compiler a year ago, and the actual version even > contains a i386 compiler. Yes, we know... :-) > Except for a handful of guys (which I should call 'fan club') who use > my compilers all days, I receive little response. Most people > continue using the ACK compilers despite their weakness (the 68000 ACK > compiler does not contain many bugs -- but it is slow). Yes, because (as far as I know) c386 canot generate code for the 8086. So, people HAVE to resort to either BCC or ACK.... BCC is a good compiler (hey, it's fast!), with reasonably few bugs. I personally think it should have an EXTERNAL preprocessor, and possibly a code optimizer, but that's future music I guess. So, maybe we can convice Christhoph to write a 8086- code generator for his compiler? Fred.
HBO043%DJUKFA11.BITNET@cunyvm.cuny.edu (Christoph van Wuellen) (03/03/91)
Oh, please no, PLEASE, *NO* 8086, *NO* 80286. Perhaps you find a sado-maso guy who will add a 8086 code generator to c68/c386. I have often said, the the INTEL processor line begins with the 80386 for me. The others are not even existent. C.v.W.
hp@vmars.tuwien.ac.at (Peter Holzer) (03/05/91)
HBO043%DJUKFA11.BITNET@cunyvm.cuny.edu (Christoph van Wuellen) writes: >Oh, please no, PLEASE, *NO* 8086, *NO* 80286. >Perhaps you find a sado-maso guy who will add a 8086 code generator to >c68/c386. I have often said, the the INTEL processor line begins with >the 80386 for me. The others are not even existent. The problem is that the 386 starts up in the so-called ``real mode'', in which it acts just like a 8086. The switch to protected mode is done later. So you cannot recompile the whole kernel with a 386 compiler. You need a 8086 compiler, too, for the 2 or 3 C-files which are used before the switch is done (Maybe we could move the switch into to an earlier point or even into shoelace, so that all the C files of the kernel can be compiled with a 386 compiler, but I think Andi won't like that. -- | _ | Peter J. Holzer | Think of it | | |_|_) | Technical University Vienna | as evolution | | | | | Dept. for Real-Time Systems | in action! | | __/ | hp@vmars.tuwien.ac.at | Tony Rand |