bug in sort

[John Watson]

Has anyone else come up with a bug in the built in sort utility that causes
it to crash (SEG.Violation) on large files >10Mb when it begins the mrge
phase.  If a -z150 option is used (record size is 120 at max) the frequency
of failures is reduced, but some files still crash the sort. Am I  the 
first, or is there a known fix floating around.

PS is SCO's sosco machine still on 408-425-3502 ?  I get no answer.

    John Watson