hawley@ucrmath.ucr.edu (brian hawley) (02/20/90)
I'm looking c++ compiler that as closely resembles the ansi standard c++ as possible. I've heard of zortek's compiler, and others, and would appreciate greatly any information that I can get concerning these compilers. Send e-mail replies to hawley@ucrmath.ucr.edu Thanks.
Bob.Stout@p6.f506.n106.z1.fidonet.org (Bob Stout) (02/21/90)
In an article of <19 Feb 90 22:38:26 GMT>, (brian hawley) writes: >I'm looking c++ compiler that as closely resembles the ansi standard c++ >as possible. > >I've heard of zortek's compiler, and others, and would appreciate greatly >any information that I can get concerning these compilers. There is no ANSI standard (yet) for C++. Zortech C++ includes an excellent ANSI-compliant (well, about as compliant as anyone else's) C compiler as well as the only native code C++ compiler currently available for DOS. Borland's Turbo C 3.0 is supposed to be released in April which will also compile C++ code, but here it is almost March and the beta copies still don't work right - we'll see. All other current DOS implementations are ports of AT&T's cfront preprocessor. Other companies with plans to release DOS C++ compilers include Microsoft (late this year or early 1991) and JPI (ditto), with Lattice and Watcom talking but even less schedule information.
sidney@saturn.ucsc.edu (Sidney Markowitz ) (02/24/90)
In article <14468.25E40426@urchin.fidonet.org> Bob.Stout@p6.f506.n106.z1.fidonet.org (Bob Stout) writes: >[...] Borland's >Turbo C 3.0 is supposed to be released in April which will also compile C++ >code, but here it is almost March and the beta copies still don't work >right - we'll see. [...] Just to set the record straight, Borland has not announced a release date for a new version of Turbo C, and has not announced any details about any new features in any particular new version. There was an article in PC Week that was supposedly based on information from a beta tester who violated the confidentiality agreement, but note that 1) beta copies can have bugs and/or performance characteristics and/or features that will not appear in the final product (that's why they are betas), 2) the existence of a beta does not imply all that much about the release date, since there are long and short beta cycles and early and late beta releases, and 3) even if Bob Stout has some inside information from a beta tester, you should realize that beta sites are not told when the final product will be released -- they are just given beta copies to test until the vendor decides that the product will ship. I am similarly skeptical about Bob Stout's claimed knowledge of unnanounced product schedules from MicroSoft, et. al. What *has* been made public is the various DOS language vendor's interest in object-oriented programming languages, and in particular C++. So I'm sure that it is just a matter of time before the major players are in the field. In the mean time, I would like to see less uninformed speculation disguised as hard facts. -- sidney markowitz <sidney@saturn.ucsc.edu>
Bob.Stout@p6.f506.n106.z1.fidonet.org (Bob Stout) (02/28/90)
In an article of <23 Feb 90 20:02:22 GMT>, (Sidney Markowitz ) writes: >Just to set the record straight, Borland has not announced a release date >for a new version of Turbo C, and has not announced any details about any >new features in any particular new version. As early as last October-November, Borland officials were being quoted by name in the popular press that TC 3.0 would include O-O extensions in a manner comparable to Turbo Pascal 5.5. These statements were later ammended after Borland decided to buy Oregon C++'s technology to implement C++ 2.00 features in TC 3.0. The target release date has, as you say, never been released, but is hardly a secret within the industry. >3) even if Bob Stout has some inside information from a beta tester, you >should realize that beta sites are not told when the final product will be >released -- they are just given beta copies to test until the vendor decides >that the product will ship. I would never report information based on the hearsay of a single beta tester. Besides the normal journalistic standard of verifying your sources, it would also make it too easy for a vendor to identify the leak and therefore cause him/her problems. My sources in this case include over a half dozen beta sites evenly divided between the industrial and journalistic community. Fortunately, I'm under no personal restrictions with Borland, so I can pass along anything useful yet difficult to trace (see above). The only compiler vendor I am under a non-disclosure agreement with is Zortech, about whose products I keep my mouth shut and let Walter tell you what he likes. >I am similarly skeptical about Bob Stout's claimed knowledge of >unnanounced product schedules from MicroSoft, et. al. Such is your right - for whatever reasons, I know fewer loose-lipped MS beta testers, but the information I have heard tends to be consistent, therefore credible. My vulnerability in this matter is less subject to errors of reporting, but rather to deliberate programs of disinformation. >What *has* been made public is the various DOS language vendor's >interest in object-oriented programming languages, and in particular >C++. So I'm sure that it is just a matter of time before the major >players are in the field. In the mean time, I would like to see less >uninformed speculation disguised as hard facts. The only exception I take with this is that what I posted was neither uninformed nor speculation, but more in the nature of reporting with verification from anonymous sources. Other than the previously cited quotes from Borland execs stating their intentions of supporting C++ 2.0 with TC 3.0 (would you like to know the exact compiler switch used to do this?), the only announced player is JPI which has been advertising a C++ to be introduced "in 1990". Microsoft's and Lattice's efforts, while known about, have been less publicized - which is, to my way of thinking, to their credit.