Re: [Bloom] review starts on 13th of May

From: Alexander Grund (alexander.grund_at_[hidden])
Date: 2025-05-23 08:21:04

Am 22.05.25 um 22:15 schrieb Arnaud Becheler via Boost:
> One example is that k is a compile-time constant. As a user this is far
> from optimal. I do not care about the number of bits used for each key,
> only that it is optimal. In practice, the optimal value for k is a function
> of the target FPR. In other cases the target is the memory overhead (m/n)
> and k is derived from that. I can choose m, n, and the FPR at runtime, but
> I have to decide on k in compile-time. I understand that there are
> implementation considerations in favor of making k a compile-time constant,
> I just wish to have that explained in the docs.
An idea: Other containers use "-1" as template arguments for "runtime size".
If this isn't too hard to support it might provide the extra flexibility 
for power users.