On 2013-03-27, at 3:16 PM, David Levine wrote: > Sure, 64-bit ints still have that risk. But I don't have > a Cray :-]
What you really want is a PDP-10. Throw some 36-bit ints into the mix and watch people start excreting large bricks :-) Anyway, I'm with David on this one. Leave the sequence bitmap at 32 bits. There's no way we can propose making this platform dependent and admit it could lose user data. The short term fix should be to ensure we use a uint32_t instead of an int, to make it very clear what's going on, and document the limit. The next step is for someone who really feels 27 sequences is a hindrance to replace the uint32_t with a variable-length bitfield. I can't believe there are less than a few hundred license-compatible implementations out there to chose from. --lyndon _______________________________________________ Nmh-workers mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/nmh-workers
