On Friday 12 September 2003 10:42 pm, Jeffrey C Honig wrote:
> Here is some additional info on this.  Does anyone see any problem with
> my checking this in?
>
> Thanks.
>
> Jeff

I'd say the fix is definately missing.

This change works for me, but the patch referenced a very old copy of 'inc.c', 
so the change gets made at line 847 instead of line 756.

The effect of this change is buried deep, and I can't help but wonder if there 
are other related bugs.  By my reading, the problem actually occurs when inc 
calls seq_save(), which calls seq_list(), which determines that the list is 
empty if mp->nummsg is zero.

There are a half dozen or more calls to seq_save() in the code, so the value 
of 'mp->nummsg' needs to be correct in each of those functions as well.

Perhaps a better fix would be to have seq_list() use mp->lowmsg and 
mp->hghmsg, or even better, introduce subroutines to manage the update of the 
values in 'struct msg', a la "object oriented" code.  But frankly, that's a 
lot of work at this point for the code set.

So I'd say commit the change.


_______________________________________________
Nmh-workers mailing list
[EMAIL PROTECTED]
http://mail.nongnu.org/mailman/listinfo/nmh-workers

Reply via email to