[CM] Backtrace in s7

[Kjetil Matheussen]

On Mon, Oct 6, 2014 at 3:16 PM, Bill Schottstaedt <bil at ccrma.stanford.edu>
wrote:

> > Do you mean why they started to give
> >these warnings in the first place?
>
> yes -- after googling around for awhile, I think this was introduced
> in guile 2.0, and is not universally popular -- for example gnucash:
>
>     ;; Turn off the scheme compiler's "possibly unbound variable" warnings.
>     ;; In guile 2.0 we get nearly 7500 of them loading the scheme files.
>     ;; This is the default value for auto-compilation-options without
> "unbound-variable".
>     (if (>= (string->number (major-version)) 2)
>         (set! %auto-compilation-options
>               '(#:warnings (arity-mismatch format duplicate-case-datum
> bad-case-datum))))
>
> I still haven't found any rationale from the guile developers.  Anyway,
> it's probably easy to add a switch.
>
>
That would be much appreciated. I guess the usefulness of this varies among
programmers. At least I often write the wrong name of a variable, or do a
typo, and that quite often too. And getting a message about these minor
typos and wrong-namings during compile time is not only a time saver (by
not having to run the program in order to discover the error plus spend
time figuring out where and why things went wrong), but I guess it can also
prevent more long-living bugs. Anyway, I was very thankful to the guile
developers for this feature when I made that software package earlier this
year, since it saved me a lot of time.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ccrma-mail.stanford.edu/mailman/private/cmdist/attachments/20141006/5b2cf79a/attachment.html