On 12/8/15 1:06 PM, Robert Haas wrote:
> I don't really want to get into an argument about this, but is the
> reason we haven't updated config.guess and config.sub in the past that
> it presents an actual stability risk, or just that nobody's asked
> before? Because the first one is a good reason not to do it now, but
> the second one isn't.
We had some incompatibility issues with these updates in the very
distant past, but I don't think this would be an issue anymore. The
updates themselves are better and smaller now, and the buildfarm
coverage is quite good. It would be prudent, however, to do a manual
verification of the changes, especially in the distant back branches.
I think there is a slippery slope argument, but I don't think we have
that many new platforms and such all the time. If people start wanting
a new arm variant every six weeks, the we'll have to put a stop to it,
perhaps.