[plt-scheme] "procedure ..." in mzscheme error messages

From: Matthew Flatt (mflatt at cs.utah.edu)
Date: Thu Mar 5 09:13:23 EST 2009

I don't think the error-print width affects that. A location-based name
is generated at compile time, and it uses some arbitrarily chosen width
--- something like 30 characters. The 30-character cut-off often has
the nice effect of keeping a machine-specific path out of bytecode,
though it's a hack for that goal.

To support backtraces, there's a more complete and portable path stored
with procedures (which was added long after the procedure-name
support). So, I could change the way that the procedure name is
generated and build it on demand from the more complete path, instead
of keeping just 30 characters of the original filename. But then this
needs to be generalized to work with other forms, such as `class'...

I'll put it on my list as something to clean up, but it probably won't
happen soon.

At Thu, 5 Mar 2009 13:24:15 +0000, Noel Welsh wrote:
> This might be down to the setting of error-print-width (which you can
> easily change).
> N.
> On Sun, Feb 22, 2009 at 7:04 PM, Neil Van Dyke <neil at neilvandyke.org> wrote:
> > Occasionally, an "mzscheme" error messages looks something like this:
> >
> > procedure ...s/funcelit-scrbl:18:9: expects 1 argument, given 3: #<pcdata>
> > #<pcdata> #<pcdata>
> >
> > Any chance the error message could be changed to not elide the path to the
> > source file?
> >
> > Thanks,
> > Neil
> _________________________________________________
>   For list-related administrative tasks:
>   http://list.cs.brown.edu/mailman/listinfo/plt-scheme

Posted on the users mailing list.