Subject: Re: gdb
From: "L. Peter Deutsch" <ghost@aladdin.com>
Date: Sun, 14 Dec 97 10:58 PST

> I can't figure out what you are talking about.  First you say that you
> prefer to purchase proprietary software, because it is cheaper.

I said that proprietary software is often cheaper, and I said I had
purchased certain pieces of proprietary software (Windows 95, Frame).  Don't
put words into my mouth, please.

> Then you tell me that you, in fact, do not generally purchase such
> products, because your development environment is a freed one, and those
> tools are adequate.

No, I said that I currently use freed development tools and that on the
platforms where they are available, I am generally satisfied with them.  I
did not say "I generally do not purchase some products", any more than I
said "I prefer to purchase such products."  Please respond to what I said,
not to what you wish I had said.

> It seems to me that you are happy and satisfied with
> the freed software that you use, and that you aren't spending inordinate
> amounts of time getting "significant" bugs fixed in gcc, gdb, and emacs.

For the development tools I currently use, the level of bugs in the freed
and the proprietary software is about the same, and I work around them.  I
do not choose to exercise the option of getting the lesser bugs fixed in the
freed software, even though I know that option is available, because the
hassle of finding someone competent to fix them, explaining what I want
fixed, shepherding them through reproducing the problem, and verifying that
the result actually works in my environment would be higher than the benefit
I would get from doing this.

There *have* been serious bugs in both the freed and proprietary software I
use.  There was a bug in gcc from 2.7.0 through 2.7.2 that caused it to
generate incorrect code on Intel machines (at least) unless I disabled
'const'.  I patched around this by turning off 'const' explicitly.  There
have been bugs of equivalent seriousness in the Borland, Watcom, IBM
RS/6000, H-P, and other compilers.  Now, it is true that the gcc maintainers
were kind enough, when I inquired about this problem, to send me without
charge a source patch that fixed this bug: but that kindness was useless,
since it wouldn't have helped the thousands of users on the net (and the
dozens of commercial customers) who wouldn't have had that patch to compile
the source code I distribute.  Paying $$ to fix the problem wouldn't have
helped.

I think I will stop here.  I have felt frustrated by the high ratio of
opinion to verifiable fact on this list, and I don't want to increase it
further.

-- 

L. Peter Deutsch         |       Aladdin Enterprises :::: ghost@aladdin.com
203 Santa Margarita Ave. | tel. +1-650-322-0103 (AM only); fax +1-650-322-1734
Menlo Park, CA 94025     |        http://www.cs.wisc.edu/~ghost/index.html
*Oppose bulk-mail abuse of your mailbox and newsgroups: http://spam.abuse.net*