Re: encode-time vs decode-time

[Lars Ingebrigtsen <larsi@gnus.org>]

Eli Zaretskii <eliz@gnu.org> writes:

>> Actually, while I agree that it's possible it will break something
>> somewhere, it's rather unlikely.
>
> I think our record of making such predictions is rather poor.
>
> And anyway, suppose someone comes up with complaints about this
> change: what exactly do we say in response?  The reason for the change
> is quite weak: it's basically an enhancement required by somewhat rare
> niche use cases.  How can we justify breaking someone's code on such
> weak grounds?

Having sub-second time resolution in calendar dates doesn't seem like a
minor thing to me.  I think if you say "now Emacs is ISO8601 compliant",
then it's all the justification anybody could possibly need.

I also think this is probably less likely to break anybody's code than
when Emacs changed the output from `current-time' from being a list of
two integers to four integers.

The only possible real world problem I can see being slightly likely is
somebody writing the output of `decode-time' to a file, and then reading
it back in again.  If the writer is a new Emacs, and the reader is an
old Emacs (that calls `encode-time' on the data), then things will
probably break.

But I don't know of anybody doing stuff like that.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no