rossj@cognos.UUCP (Ross Judson) (01/17/91)
I am currently spending some spare time writing an application for my ST, and I seem to have come across some strangeness. Does anyone know why the MWC CSD debugger might be having trouble with global variables? I am running TOS 1.4, 2.5 meg, csd 1.0, MWC 3.09. Apparently CSD is casting the variables incorrectly (I think). When I view variables they are printed incorrectly, but the code seems to run. Am I running the most recent versions, or should I try and contact Mark Williams? BTW, I've compiled and run the same code on Laser C with no problems (under the debugger as well). -- Ross Judson ++ uunet!mitel!cunews!cognos!rossj + hunger is the fuel of greed Cognos Inc. ++ rossj@cognos.uucp + enough is less than you need
gcarter@gumby.cs.wisc.edu (Gregory Carter) (01/19/91)
MWC's CSD is flaky as all hell. I am surprised your actually trying to use that thing. If your doing something that complex, that you need a source debugger, I would trash MWC and get Borlands package. THERE are a LOT of strange things that compiler system does for the ST, if you don't at least have version 3.0.6 I really question whether its useful for anything but light and medium-light jobs. Gregory
tony@tacky.cs.olemiss.edu (Tony Reynolds) (01/22/91)
In article <1991Jan18.225523.22934@daffy.cs.wisc.edu> gcarter@gumby.cs.wisc.edu (Gregory Carter) writes: >MWC's CSD is flaky as all hell. I am surprised your actually trying to use >that thing. If your doing something that complex, that you need a source >debugger, I would trash MWC and get Borlands package. I once called Mark Williams Co with the question "Where is the break key?" They said, "What do you mean? Pressing F1 will stop a program...." "But not a run-away program!" I once worked on a 386 cross-assembler: it was going to run under MINIX, but MINIX was so unstable then (summer 88) and had the same (flaky)trouble. Program runs fine under TOS, but the debugger causes mistakes, or even better, crashes under TOS but runs fine under CSD.... !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Please note! All mail directed to me should be sent to my new address: cctony@4dsci1.ocis.olemiss.edu Thank-you very much. This includes USENET Mail, please check the "To:" field for replys
rossj@cognos.UUCP (Ross Judson) (01/22/91)
In article <1991Jan18.225523.22934@daffy.cs.wisc.edu> gcarter@gumby.cs.wisc.edu (Gregory Carter) writes: >MWC's CSD is flaky as all hell. I am surprised your actually trying to use >that thing. If your doing something that complex, that you need a source >debugger, I would trash MWC and get Borlands package. >THERE are a LOT of strange things that compiler system does for the ST, >if you don't at least have version 3.0.6 I really question whether its >useful for anything but light and medium-light jobs. These are fairly strong statements. It would be great if you could provide a few concrete examples of problems. There are several problems with switching compilers. The main one is that my program is uses a lot of the MWC specific features in the library. These would all have to be converted to a new compiler. Second, it costs to buy a new compiler. I already have one and I'm very satisfied with it, except for the debugging strangeness. But I'm curious. Where can one get the Borland C package in Canada? Is anyone aware of a profiler that is compatible with MWC version 3 binaries? Some time ago (1987) David Rowley released a profiler for MWC v2 and alcyon/dri format executables. Has this ever been updated for v3 executables? -- Ross Judson ++ uunet!mitel!cunews!cognos!rossj + hunger is the fuel of greed Cognos Inc. ++ rossj@cognos.uucp + enough is less than you need