Date: Fri, 31 Jan 1997 18:10:38 -0500 (EST) From: Dan To: Eli Zaretskii cc: djgpp AT delorie DOT com Subject: Re: DJGPP vs Borland C++ In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Thu, 30 Jan 1997, Eli Zaretskii wrote: :> :>On Wed, 29 Jan 1997 afn03257 AT afn DOT org wrote: :>> :>> >to comment on something that I didn't understand. I use GCC on :>> >different platforms since version 1.4.0, which was about 10 years ago, :>> :>> Not posible unless this is 1999. :> :>Isn't it? ;-) :> :>OK, so it's 8 years (still qualifies to be ``about 10 years'', IMHO). I know a lot of 8 year olds that would be happy to hear that. ;-) :>> Did they debug or just find bugs? There is a difference. Dedicated > :>programmers? You call someone who looks through those sources, having > :>not coded it themself, to find a bug not dedicated? then to make the > :>patch and send it in? I'd call that dedicated. :> :>Dedicated is open to interpretation. Here's mine: a dedicated programmer :>is somebody whose daytime job is to support a given program/package, or :>who invests most of their working week in it. That is certainly NOT the :>case with neither most of the GNU project, nor with DJGPP. Sounds like a job. ;-) :>> >*Any* software has bugs, no matter how long it is developed. In fact, :>> >one of the definitions of software is ``lines of codes with bugs'' ;-). :>> :>> What?? :>> That is exactly what I said, and you said I was wrong. :> :>We seem to agree on more and more points as we go. So why are we still :>arguing? Are we? I didn't think we were arguing. You made a statement about Borland being buggy. I didn't find it to be really accurate considering, so I added my $.04 [inflation]. :>> This is true, however, technically you could patch the comercial :>> software yourself with a debugger. :> :>Incidentally, that's what I did sometimes because I couldn't get the :>vendor to let me have a patched version in reasonable time. But this can :>hardly qualify as a good way to maintain software. Naturally I wouldn't say this is a good way to support commercial software, and HLL src code would be much better. Nonetheless, one point I neglected to mention is that several of the errors reported to Borland are actually windows API bugs which borland can't really do anything about. I had a reply for the newsgroup but my nntp server is down again, as usual.