Re: [Nmh-workers] GCC 8 pre-releases have escaped...

[Ken Hornstein]

>It could be `compbuffers' guarded by
>
>    /* first-time only initialization, which will always happen the
>       way the code is now, with callers initializing *scanl to NULL.
>       scanl used to be a global. */
>    if (! *scanl) {

It's actually a bunch of things guarded by that, but yeah ... it turns out
for every message now scanl is called initialized to NULL, and a bunch
of things are allocated in there.  But this wasn't in 1.6, and I kind of
think this should be fixed before 1.7.1 goes out the door.

--Ken