Re: [Monotone-devel] Re: RFC: Fake IDs

[Zack Weinberg]

On 7/18/06, Graydon Hoare <address@hidden> wrote:
> I'm not concerned with collision probabilities per-se. I'm concerned
> with using collision probabilities as some proxy argument for what you
> actually want in this case. A "fake ID" is supposed to be "an ID I
> currently don't have in my database, or in the set of fake IDs I've
> currently chosen during this run". Why hash strings and wonder about
> probability when you can just check your database?

Well, as I said, it's a matter of not having the database at some of
the points where these things are required.  work.cc:get_work_rev on
net.venge.monotone.api, for instance.

zw