On Oct 7, 10:11pm, eggert@cs.ucla.edu (Paul Eggert) wrote: -- Subject: Re: [tz] localtime.c patch | I hope the attached proposed patch fixes all these problems; please let us | know if it misses anything. Yes, this looks good. There were other NetBSD changes that were not carried forward, and perhaps they should: - localtime and friends can return NULL. This is not mentioned in the man pages or handled properly. Specially annoying is that ctime() just calls asctime(localtime()) without checking the result of localtime and this can make programs coredump without giving them a chance to handle the error. - timesub() doesn't set errno on overflow so none of the API functions that go through this error path set errno when they fail. It would be nice if the error conditions of all the functions were documented in the man pages and the errors were formalized (i.e. have them always set errno consistently depending on the error condition). Thanks, christos