Turbo C vs Microsoft C

[kla at physc1.byu.edu]

Turbo C is much preferred if you have to get down into the bowels of the
DOS world (ie. tsr's operating systems etc) because of the short cuts which
you can't make in MSC. For example _AX = 0x45 translates in turbo to "stick
a hex 45 into the AX register. In microsoft you have to edit a separate
.asm file and link it in. I feel sorry for anyone who has to do this
type of stuff, but if you do Turbo has some real advantages. I wrote a TSR
in Turbo in about 2 hours. It took me (an .asm novice) almost 2 weeks to
get the same thing going with MSC and MASM.

I would be the first to admit that MSC probably puts out better code, but
I believe that Programmer efficiency is often much more important than
program efficiency. With the MSC link taking upwards of 4 minutes on some
of our stuff (= 30 seconds with tlink) there are some obvious advantages.

One reason we have had to use MSC is to link in MSC libraries which have
been sold to us without source. Using Turbo under these circumstances is
tenuous at best, but we worked out a way of putting the library into a
MSC TSR and then accessing it with Turbo C. Works like a charm and 80% of
the students in the lab use this version rather than wait for the $^#^#$*&
linker of MSC.

On the other hand MSC 6.0 is supposedly an incremental compiler. Is this
true or just a rumor? Will this work under DOS or just OS/2. Has it got
anything to break the 640K headaches? If any of these things are true
it would probably beat the pants off of Turbo's current version. But Turbo
is bound to come up with another version real soon too.

Hope this helps or at least fans the flames.