Hi Tim Hutzler, how's life?
09-Jul-96 22:39:04, Tim Hutzler wrote to Marcel Veldhuizen
Subject: Re: Using the 3.2 editor
MV>> Isn't there a way to find out were the runtime error occured?
TH> I'd sure like to know...
Me too...
MV>> What does the 'pgm-ctr' stand for if you get a runtime error?
TH> I think that is the value of PC (stands for Program Counter, a
TH> register that points the processor to the current instruction/OP-code.
I know...
TH> PBC takes a command argument "/OM" that generates a .MAP file that has
TH> xfed info, but they don't seem to correctly correspond to the line
TH> actually making the error in the source code. Further, the xref is in
TH> hex and the error is in decimal.
The HEX-decimal-difference isn't the problem, I hope:-)
On a really small program, the pgm-ctr seemed to make sense, the pgm-ctr
was higher than value of the line on which it occured, but lower than
the next line... But now, I get such a large value that it can't even
refer to an offset in the segment...
TH> Of course, it does. It doesn't use PBD however, but some internal
TH> process that I can not access.
Maybe some secret command-line-option which is available on all
PB-programs:-) Like /ShowClearErrorMessages ...
TH> The IDE is well done in most aspects, but the editor is crap, IMHO.
TH> They really should have made it more programmable, like QEdit.
Yep.. What I REALLY dislike too, is that breakpoint are deleted after a
recompile..
TH> If I create a run time problem, and I can't figure it out, I will go
TH> into PB and let it tell me what the error is. The debugger in there
TH> seems to be okay, though I haven't had to lean on it that much.
You could do that, but of course, it WOULD be really neat if the editor
itself could give the line of the error, it would save you lots of time..
TH> Because the IDE is so crummy, I have been developing my own editor. It
TH> isn't that hard to write one, but it is still time consuming.
I know, I wanted to create my own message-editor once.. It was allmost
finished when I thought of including wrapping, from then on, it never worked
properly again:-(
>>> Marcel Veldhuizen <<<<<<<< * Origin: United Bbs Systems Europe MailGate to -> Fido USA (2:2802/337.0)
|