[Monotone-devel] Re: nvm.stripped versus botan

[Zack Weinberg]

On Tue, Jan 20, 2009 at 2:31 PM, Zack Weinberg <address@hidden> wrote:
> I noticed a severe slowdown in the time for nvm.stripped to run the
> testsuite (from five to twelve minutes on my big beefy quad-core
> machine) which *seems* to be all down to initializing botan's RNG.

Most subcommands don't use cryptographic random numbers at all.   I
just pushed changes to .stripped that make us wait to instantiate a
RandomNumberGenerator object until we actually need one (under the
>=1.7.7 API).  That takes the testsuite runtime back down to a little
more than five minutes.  I consider the performance with botan 1.8.0
acceptable now.

I'm pretty happy with .stripped.  I was able to muck out an incredible
amount of gunk from the configure script, and if we get rid of netxx
even more can go.  The binary's smaller, the build is faster, and the
Debian security team will be happy with us.  What remains to be done
before we can land it?

Folks with exotic systems (non-Linux, non-*BSD) might wanna check what
I did to the configure script, btw.  I might have been overoptimistic.
:-/

zw