Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!cmcl2!rutgers!mtune!codas!killer!richardh
From: richardh@killer.UUCP (Richard Hargrove)
Newsgroups: comp.sys.ibm.pc
Subject: Re: QuickC
Message-ID: <1535@killer.UUCP>
Date: Tue, 15-Sep-87 21:00:20 EDT
Article-I.D.: killer.1535
Posted: Tue Sep 15 21:00:20 1987
Date-Received: Sat, 19-Sep-87 12:54:20 EDT
References: <1485@killer.UUCP> <3320052@hpsrlc.HP.COM> <3826@watdcsu.waterloo.edu>
Organization: The Unix(R) Connection, Dallas, Texas
Lines: 49
Summary: bugs and Microsoft

In article <3826@watdcsu.waterloo.edu>, broehl@watdcsu.waterloo.edu (Bernie Roehl) writes:
> In article <14563@watmath.waterloo.edu> kcollinsthom@watmath.waterloo.edu (Kevyn Collins-Thompson) writes:
> >In article <3320052@hpsrlc.HP.COM> darrylo@hpsrlc.HP.COM (Darryl Okahata) writes:
> >>
> >>     I've also heard that Quick C is a MEDIUM-model ONLY compiler. YUK!
> >>
> >If it has less bugs than Turbo C it seems like it might be worth
> >throwing out Turbo C in favour of Quick C
> 
> It's *very* unlikely that Quick C will be less buggy than Turbo; Turbo
> has what, maybe four known bugs?  Compare that to some of the stuff
> Microsoft has put out, and I'll stick with Borland.  

I'll stick with Borland also. They've gotten a bad rap for the bugs in
Turbo C 1.0 (how many bugs were in MASM 1.0 or MS C 2.0?). Though, I have no
idea when a new release is coming from Borland, I suspect they are going 
to make sure it is well debugged.

A short story about compiler bugs: at work (Texas Instruments) we are 
purchasing an Ada compiler targeted for the 1750A. It costs somewhere
between $1000 and $100000 (I don't know how much - I just know it's the 
high-priced spread). I say purchasing it because: Guess what ? It's got
bugs! Amazing! Our software people are working very closely with Tartan
to find and fix them.  

These programs are horrendously complex, involving literally thousands
of trade-offs, engineering decisions, and judgement calls. There will
be bugs! Live with them, work araound them, and wait for the next 
release. Oh yes, bitch like hell to the vendor. That's the only way
you stand a chance of having them fixed. But don't throw out a fine product
because it has a few bugs!

Note that it's the middle of September and we're still wondering what 
Quick C and MS C v5.0 are going to be like. Borland really turned up the 
heat on Microsoft back in May. Given Microsoft's history of buggy products
(we use MS Pascal at work, v3.something or other; we still have modules
that we must compile with optimization turned off or otherwise we get
bad code generated) combined with the pressure of releasing something in 
September, we have the potential for a real buggy set of products.

Borland has been very responsive to bug reports. We've gotten nothing but
grief from Microsoft.

Like I said, I'll stick with Borland.

regards,
richard hargrove
...!ihnp4!killer!richardh
-------------------------