nicholso@pioneer.arc.nasa.gov (Melvin H. Nicholson -- YBH) (07/07/90)
I was about to upgrade my microsoft compilier to version 5.1 from
4.something, and found out that the current upgrade was to 6.0
Not knowing the bug-status of the new upgrade, I am here looking for
reviews. Does anyone know whether the 6.0 is reliable?
Mel
nicholso@pioneer.arc.nasa.gov
--
Send compilers articles to compilers@esegue.segue.boston.ma.us
{spdcc | ima | lotus| world}!esegue. Meta-mail to compilers-request@esegue.acook@athena.mit.edu (Andrew R Cook) (07/15/90)
MicroSoft C6.0 has many nice features that C5.1 or C4.0 do not have, that make
the upgrade worthwhile. First of all(and most important to me) it includes a
whole bunch more in the way of graphics support. It now has a presentation
graphics library that is good for producing graphs very painlessly. It is not
as flexible as one would like, but is very satisfying none the less. As
before,
all graphics are supported in HGC, CGA, EGA, and some VGA modes. Other new
features include the Programmer's Workbench, an improved CodeView debugger, and
some small increases in compiled program execution speed. The Programmer's
Workbench is a nice idea, combining the compiler, make facility, codeview, and
rather extensive on-line help into a single programming environment.
Unfortunately, the Programmer's Workbench is painfully slow to use unless you
have a fast 386 machine. I use a 16 MHz 286 box, and found using the PWB to
be too tedious to use. The on-line help is a great idea though. It includes
a complete Library reference manual. Unfortunately, there is no printed
library
reference manual included with the compiler, but one is available from
MicroSoft
Press for 20 or 30 bucks.
There are a few bugs to watch out for. I have found and comfirmed with
MicroSoft 3 or 4 bugs with some of the new graphics stuff. Overall however,
I have not felt like the version is bug ridden. MicroSoft has been working
on this compiler long enough that there are not many bugs. There are probably
some bugs in the PWB, but I have not used it much, thus have only found one
bug in it. I wouldn't worry about bugs too much. If you do have a problem,
MicroSoft tech support is quite helpful.
Overall, I would suggest that you do invest the $150 for the upgrade, the new
features make it worthwhile, and MicroSoft has promised to increase the speed
of the PWB.
-Andy Cook, MIT (acook@athena.mit.edu)
******************************************************************
* All the world's indeed a stage, and we are merely players, *
* performers and portrayers, each anothers audience, *
* outside the guilded cage . . . *
******************************************************************
--
Send compilers articles to compilers@esegue.segue.boston.ma.us
{spdcc | ima | lotus| world}!esegue. Meta-mail to compilers-request@esegue.nicholso@pioneer.arc.nasa.gov (Melvin H. Nicholson -- YBH) (07/17/90)
A while ago I posted a querry about Microsoft C 6.0.
For those of you interested, here are the results:
Bugs:
1) The compilers fails on certain large but legal macro expansions.
2) A certain optimization does not treat the stack pointer correctly,
which causes an eventual crash. (reported three times)
3) NULL has been redefined (not really a bug, but annoying)
4) No printed docs.
On the other hand 60% of the repondants reported not having problems,
and most agree that the optimizations are generally better, or at least
will be once the bugs are worked out.
Please note that this is second hand information for which I take no
reponsibility.
Mel
--
Send compilers articles to compilers@esegue.segue.boston.ma.us
{spdcc | ima | lotus| world}!esegue. Meta-mail to compilers-request@esegue.dave@thundercat.com (David Conrad) (07/19/90)
In article <1990Jul14.222613.13835@esegue.segue.boston.ma.us> acook@athena.mit.edu (Andrew R Cook) writes: }MicroSoft C6.0 has many nice features that C5.1 or C4.0 do not have, that }make the upgrade worthwhile. First of all(and most important to me) it }includes a whole bunch more in the way of graphics support. It now has a }presentation graphics library that is good for producing graphs very }painlessly. It is not as flexible as one would like, but is very satisfying }none the less. As before, all graphics are supported in HGC, CGA, EGA, and }some VGA modes. ... Are you familiar with Borland's BGI (included with Turbo C2.0)? How do the two compare? They sound very similar (Microsoft probably added this feature because of its presence in Turbo C). -- Dave Conrad dave@thundercat.com -- Send compilers articles to compilers@esegue.segue.boston.ma.us {spdcc | ima | lotus| world}!esegue. Meta-mail to compilers-request@esegue.
6600sirt@ucsbuxa.ucsb.edu (07/20/90)
>From article <325@tygra.ddmi.com>, by dave@thundercat.com (David Conrad): > In article <1990Jul14.222613.13835@esegue.segue.boston.ma.us> acook@athena.mit.edu (Andrew R Cook) writes: > }MicroSoft C6.0 has many nice features that C5.1 or C4.0 do not have, that > }make the upgrade worthwhile. First of all(and most important to me) it > }includes a whole bunch more in the way of graphics support. > Are you familiar with Borland's BGI (included with Turbo C2.0)? How do the > two compare? They sound very similar (Microsoft probably added this feature > because of its presence in Turbo C). You're right, Microsoft probably did add it because of competition with Borland. But Microsoft's graphics library is much more powerful and easier to use than Borland's. MSC has more presentation-graphics functions, and those that have TC equivalents are faster in executing than TC's. Also, with MSC, you can write in any Windows font. That means that if you have the Windows 3 SDK, you can create your own stroked fonts to use with MSC. The big difference, though, is that MSC does not use the equivalent of BGI files. All of the graphics support (exept fonts) is linked into the EXE. On the plus side, that means you don't have to distribute BGI files with everything you write. On the down side, it means that you can't modify MSC's functions to work with Super VGA or anything. By the way, one MSC 6.0 bug that I didn't see anyone mention: I almost always get an Internal Error when I try to compile a program with a customized memory model. Mike O'Brien 6600sirt@ucsbuxa.ucsb.edu -- Send compilers articles to compilers@esegue.segue.boston.ma.us {spdcc | ima | lotus| world}!esegue. Meta-mail to compilers-request@esegue.
mlord@uunet.UU.NET (Mark Lord) (07/25/90)
In article <1990Jul19.182239.6948@esegue.segue.boston.ma.us> 6600sirt@ucsbuxa.ucsb.edu (Mike O'Brien) writes: > >The big difference, though, is that MSC does not use the equivalent of >BGI files. All of the graphics support (exept fonts) is linked into >the EXE. On the plus side, that means you don't have to distribute >BGI files with everything you write. On the down side, it means that >you can't modify MSC's functions to work with Super VGA or anything. Actually, .BGI files can also be linked into the .EXE by using a turbo utility to first convert the ones you want to use into .OBJ format, and then simply linking them in. Check out the Turbo-C docs for details. -- ___Mark S. Lord__________________ | ..uunet!bnrgate!bmerh614!mlord | | ..uunet!bnrgate!mlord%bmerh614 | -- Send compilers articles to compilers@esegue.segue.boston.ma.us {spdcc | ima | lotus| world}!esegue. Meta-mail to compilers-request@esegue.