binary protocol time representation

[Daniel Farina]

On 7/12/07, Dustin Sallings <dustin at spy.net> wrote:
>
> On Jul 12, 2007, at 12:45 , Marc wrote:
>
> > Are there any places where we either require absolute time or a time
> > interval greater than 49 days?  For the latter, we could simply use a
> > uint64_t version of timestamp, but I'd rather that be the exception
> > than the
> > common case.
>
>         Well, I'm wondering if anyone has an interval that long where they
> don't mean ``forever.''  It might be enough to say 0 == no time-based
> invalidation.
>

I think this is a good option to have (although not essential) if
someone simply wants to not worry about time invalidation and just
rely on LRU and cache size constraints. In a way that's even more
classical cache-y than time based invalidation.

>         As a cache, it should be considered possible that an entry may
> disappear at any time and you'll have to regenerate it.

Agreed.