From: "Michael Stewart" <mike AT reggin DOT freeserve DOT co DOT uk>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: definatly an error in either the compiler or RHIDE:
Date: Fri, 11 Jun 1999 08:31:21 +0100
Organization: (Posted via) Netcom Internet Ltd.
Message-ID: <7jqdom$4l0$1@taliesin.netcom.net.uk>
References: <l03130305b385ed37e622@[145.98.116.217]>
NNTP-Posting-Host: hgty.capgemini.co.uk
X-Trace: taliesin.netcom.net.uk 929086038 4768 194.42.240.2 (11 Jun 1999 07:27:18 GMT)
X-Complaints-To: abuse AT corp DOT netcom DOT net DOT uk
NNTP-Posting-Date: 11 Jun 1999 07:27:18 GMT
X-Newsreader: Microsoft Outlook Express 4.72.3155.0
X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3155.0
Lines: 34
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

Dlanor Blytkerchan wrote in message ...
>Small, insignificant, yes. But still worth mentioning:
>
>Writing a sloppy C++ program with a lot of errors in it, the compiler
>returned all those errors to me with the wrong line numbers. For example, I
>was declaring a class variable from main (somewhere beyond line 50 in my
>source):
>
>int main(void) {
>  blabla bla1;
>  blabla bla2;
>  blabla bla3;
>  classType classVar = new classType(bullshit);
>
>  (..)
>
>  return(0);
>} // main()
>
>guess what? It didn't work. (Ofcourse not: I had even declared the class
>itself all wrong, more like a struct than a class..). The compiler returned
>an error message indicating that the fault was in main() (which is right),
>on line number 4 of the file (which is wrong: it is line number four in the
>method, but definatly not in the file). RHIDE promptly pointed me to that
>line. Either RHIDE is wrong or the compiler is (though it was right that I
>was producing errors more than it was, so I stopped programming and started
>E-mailing).


This may be due to optimisations being switched on.

Michael Stewart