home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.amiga.programmer
- Path: sparky!uunet!sun-barr!cs.utexas.edu!hellgate.utah.edu!fcom.cc.utah.edu!val
- From: val@news.ccutah.edu (Val Kartchner)
- Subject: I'd use GNU C++ except.... (SAS read this)
- Message-ID: <1992Jul31.175800.18880@fcom.cc.utah.edu>
- Sender: news@fcom.cc.utah.edu
- Organization: University of Utah Computer Center
- X-Newsreader: Tin 1.1 PL3
- Date: Fri, 31 Jul 92 17:58:00 GMT
- Lines: 57
-
- I've been waiting for SAS to come out with their C++ compiler for quite a
- while. I've finally become tired of waiting, and since GNU has a C++
- compiler available, I'd like to use it and NEVER get another commercial
- compiler again. However, there are a few reasons that I haven't yet gone
- with GNU, and may yet cause me to upgrade to SAS/C++:
-
- 1. I have an Amiga 1000. This alone shouldn't be a problem except that
- whomever compiled the Amiga version of G++ made it require the
- ieeesingbas.library which is not available without 2.04. I urge
- whomever is doing G++ for the Amiga to mage a version available for
- us pre-2.04 users. Otherwise, we can't even use G++.
-
- 2. G++ doesn't come with a screen-oriented debugger. This doesn't stop
- me from using G++ (as does #1 above), but a source-level screen-oriented
- does help greatly in debugging. I can still debug the old fashioned,
- but I would prefer not to do so.
-
- 3. G++ requires that object files be converted to Berkley format so that
- you may use their linker. Unless their linker format has special
- support for C++ without name-mangling or other neat features, then
- there is no reason for this. It should be no harder for the compiler
- to produce Amiga-formatted object files than it is for the linker to
- produce Amiga-formatted executable files. Eliminating this step would
- also simplify support by eliminating the need to upgrade the GNU linker
- as well.
-
- 4. The implementer of G++ has an attitude problem. There are certain
- implementation details (besides what has been stated above) with
- which I disagree strongly. These things include implementing a
- method of doing macros (#define's) without side-effects. This is what
- inline functions are for (among other things). To be fair, if I
- were to implement a C++ compiler (as I am considering doing), I
- would do things with which others would strongly disagree as well.
-
- 5. The FSF (Free Software Foundation) and I have a philisophical
- difference of opinion. The FSF, from what I gather, wants all
- programmers to work for free. This means that no programmer should
- ever get paid for any program. (Note: this does not mean that they
- advocate piracy. They are willing to respect others views and I
- respect their distribution license.) However, this is not realistic
- within our economic system. However, I do respect their right to
- hold that opinion, and the ferver with which they stick to it.
- (I have opinions with which others would disagree as well.)
-
- Points one through three are the real sticklers. However, point one should
- be given special emphasis.
-
- Whether SAS/C++ or G++ wins out is yet to be seen. However, my money (pun)
- is on G++.
-
- Just my 2 bits.
-
- --
- |======================== The previous was my opinion =========///============|
- | "AMIGA: The computer for the creative mind."(tm) Commodore /// Weber State |
- | "Macintosh: The computer for the rest of us." (tm) Apple\\\/// University |
- |=================== val@csulx.weber.edu ==================\///=== Ogden, UT =|
-