Re: [bug-gawk] gawk with dates before 1970

[Aharon Robbins]

> Thanks.  I hate to beat a dead horse, but why is it a fatal error
> when the timestamp argument is out of time_t range, as opposed to
> simply returning an empty string when gmtime/localtime fails?
> Shouldn't we handle these consistently?

Yeah, yeah, yeah. I'm lazy. :-)  Empty string it is.

> My personal preference might be
> to return "" in both cases along with a warning message.

Warning only if --lint.

Arnold