bchin@umd5.umd.edu (Bill Chin) (04/17/91)
Anyone seen/have such a beast? I've heard lots of rumors about broken optimization and warning levels creating bad code. Specifically, will using -W3 or -Oas create bad code for DOS or Windows programming?? Thanks. -- Bill Chin internet:bchin@umd5.umd.edu PC/IP, Computer Science Center NeXTmail:bchin@is-next.umd.edu University of Maryland, CompuServe:74130,2714 College Park *Standard Disclaimers Apply*
n8541751@unicorn.cc.wwu.edu (Where there is darkness, light) (04/18/91)
bchin@umd5.umd.edu (Bill Chin) writes: >Anyone seen/have such a beast? I've heard lots of rumors >about broken optimization and warning levels creating >bad code. Specifically, will using -W3 or -Oas create >bad code for DOS or Windows programming?? >Thanks. I don't have a bug list, but we're switching back to version 5.1 at work because of bad problems with warning levels and slower compiles without a noticeable gain in execution speed. Kris. -- Kriston M. Bruland | . . . . . . . . . . n8541751@unicorn.cc.wwu.edu | . . . . . . . . . 8541751@nessie.cc.wwu.edu | . . . . . .
nengle@copper.ucs.indiana.edu (nathan engle) (04/18/91)
In article <1991Apr18.070131.1892@unicorn.cc.wwu.edu> n8541751@unicorn.cc.wwu.edu (Where there is darkness, light) writes: >bchin@umd5.umd.edu (Bill Chin) writes: > >>Anyone seen/have such a beast? I've heard lots of rumors >>about broken optimization and warning levels creating >>bad code. Specifically, will using -W3 or -Oas create >>bad code for DOS or Windows programming?? >>Thanks. > >I don't have a bug list, but we're switching back to version 5.1 at work >because of bad problems with warning levels and slower compiles without >a noticeable gain in execution speed. I was personally very disappointed with the state of first release of 6.00. It had lots of things that just didn't work, and several that did work but very slowly. For about 4 months I did a complete backtrack and started moving all my stuff over to Zortech (they send you their bug list for free). However, eventually 6.00a came out and most of my complaints were cleared up so I'm running with it right now. I have to agree that MSC5.1 is/was about twice as stable as C6.00; 6.00a seems to be more on an equal footing with 5.1 as far as reliability goes. 6.00a IS slower than 5.1 and the output code isn't that much better, but I'm sticking with 6.00a because of the inline assembler. Also, if I'm ever rash enough to take on any OS/2 projects 6.00a supports OS/2 better than 5.1 did. -- Nathan Engle Software Evangelist Indiana University Dept of Psychology nengle@copper.ucs.indiana.edu
mrs@netcom.COM (Morgan Schweers) (04/20/91)
Greetings, I too would like such a beast. MSC 6.0's supposed 'optimization' badly broke my code recently. The strange thing is that when I used /Ozax it failed, but when I used /Ozacegilt /Gs it worked. These two are supposed to be the same, according to the online help. Am I seeing things? -- Morgan Schweers
demillo@porter.geo.brown.edu (Rob DeMillo) (04/26/91)
In article <1991Apr18.133035.15827@bronze.ucs.indiana.edu> nengle@copper.ucs.indiana.edu (nathan engle) writes: > > I have to agree that MSC5.1 is/was about twice as stable as C6.00; >6.00a seems to be more on an equal footing with 5.1 as far as >reliability goes. 6.00a IS slower than 5.1 and the output code isn't >that much better, but I'm sticking with 6.00a because of the inline >assembler. Also, if I'm ever rash enough to take on any OS/2 projects >6.00a supports OS/2 better than 5.1 did. > I have to agree with the opinions experessed here. My partner and I have stuck to MSC 5.1 for that very reason. Speaking of which, when is the next release of MSC coming out? - Rob DeMillo | Internet: demillo@juliet.ll.mit.edu Mass Inst of Tech/Lincoln Lab | Also: demillo@porter.geo.brown.edu Weather Sensing Project-Group 43 | Reality: 401-273-0804 (home) "I say you *are* the Messiah, Lord! And I ought to know, I've followed a few!"