gordon@alberta.UUCP (Gordon Atwood) (03/18/86)
I would like to obtain as much data as possible on Aztec C, produced by Manx
Corporation. Details of other C compilers would also be appreciated.
I would appreciate useful comments. Not "I liked it", but "I like it because
it did this..."
I need to know specific information in the following areas:
1) Compatibility with "Apple Compatible" microcomputers
2) Operation on an Apple IIe w/ 128k. ie, speed, does it use all available
memory, size of source code, size of executable, speed of compilation, etc
3) Interaction with Prodos, and Dos 3.3.
4) Compatibility with "Standard C" [whatever that may be], (I prefer as close
to Kernighan as possible).
5) What is the company like, responsive, uncaring, useless, ...
6) Any other information that will help in my making a decision concerning
the purchase of this Package.
G.H.A.gwyn@brl-smoke.UUCP (03/23/86)
In article <853@alberta.UUCP> gordon@alberta.UUCP (Gordon Atwood) writes: >I would like to obtain as much data as possible on Aztec C, produced by Manx >Corporation. Details of other C compilers would also be appreciated. > >I need to know specific information in the following areas: > > 1) Compatibility with "Apple Compatible" microcomputers Unknown, although if they run unmodified DOS 3.3 and ProDOS I would expect them to run Aztec C also. > 2) Operation on an Apple IIe w/ 128k. ie, speed, does it use all available > memory, size of source code, size of executable, speed of compilation, etc There is no special treatment of the extended memory. Under ProDOS, one can use this /RAM space to speed up loading of executable code, or for temp files, etc. Generated code is fairly good; one can select either native 6502 (65C02 optional under ProDOS) or interpreted code: native code is fast and large, interpreted small and slow. Both types of code can be linked in the same executable. For real speed I directly access peripheral registers rather than sending I/O through the standard I/O hooks. The compiler itself is fairly slow, so if you program by trial-and-error you'll probably not like it. > 3) Interaction with Prodos, and Dos 3.3. The ProDOS release is "beta testing" (it actually has a ways to go yet); both versions of course use the OS for file I/O (what interaction are you talking about?). Both versions provide a "shell" environment much like a UNIX subset, plus several utilities including a screen editor. The ProDOS version is supposed to eventually have support for cross-development of DOS 3.3 target applications. > 4) Compatibility with "Standard C" [whatever that may be], (I prefer as close > to Kernighan as possible). Aztec C is very close to K&R, and the ProDOS version includes later features a la X3J11. Floating-point support IS provided. Bit-fields were missing last I looked. The library is closer to 7th Ed. UNIX than to X3J11 or the SVID; this can be fixed easily enough. > 5) What is the company like, responsive, uncaring, useless, ... It varies. Usually they seem rather unresponsive. There is a tech support bulletin board for customers. > 6) Any other information that will help in my making a decision concerning > the purchase of this Package. The ProDOS version is taking much longer to get to the release point than Manx at first indicated. I get the feeling that Jim Goodnow is their main developer and he has been busy getting Mac & Amiga releases ready. I've found the compiler itself to be pretty reliable, but the ProDOS shell is not ready for serious use yet.
nazgul@apollo.uucp (Kee Hinckley) (03/26/86)
In article <2006@brl-smoke.ARPA> gwyn@brl.ARPA writes: > In article <853@alberta.UUCP> gordon@alberta.UUCP (Gordon Atwood) writes: > >I would like to obtain as much data as possible on Aztec C, produced by Manx > >Corporation. Details of other C compilers would also be appreciated. > For real speed I directly access peripheral registers > rather than sending I/O through the standard I/O hooks. And if you really want to cruise, you can always talk to the screen directly! > The ProDOS version is supposed to eventually have support > for cross-development of DOS 3.3 target applications. They have that out in beta now, but I haven't tried it yet. > > 4) Compatibility with "Standard C" [whatever that may be], (I prefer as close > > to Kernighan as possible). > Aztec C is very close to K&R, and the ProDOS version > includes later features a la X3J11. Floating-point > support IS provided. Bit-fields were missing last I > looked. The library is closer to 7th Ed. UNIX than > to X3J11 or the SVID; this can be fixed easily enough. Bit fields are in the new prodos version (least-wise the latest update warned that arithmetic with bit fields didn't work yet, which I assume means that bit-fields are there!) > > 6) Any other information that will help in my making a decision concerning > > the purchase of this Package. > The ProDOS version is taking much longer to get to the > release point than Manx at first indicated. I get the > feeling that Jim Goodnow is their main developer and > he has been busy getting Mac & Amiga releases ready. > I've found the compiler itself to be pretty reliable, > but the ProDOS shell is not ready for serious use yet. I agree with that. I doubt that the Apple ][ market is a major money maker for them, so it doesn't get priority. On the other hand the only other choice is the ORCA implementation of SmallC, which by their own admission is not a production system (but you do get the entire source code!). -nazgul -- ...decvax!wanginst!apollo!nazgul Little Bo-Peep Has lost her sheep, The radar has failed to find them. They'll all, face to face, Meet in parallel space, Preceeding their leaders behind them. A Space Child's Mother Goose