Do we need this junk?
Bruce M. Simpson
bms at FreeBSD.org
Fri Apr 6 21:30:16 UTC 2007
Nikolas Britton wrote:
> What data do you have to back that claim up? Unless you have real
> quantitative data to support your claims the bsdstats.org statistics
> stand as an accurate representative sample of the population... i.e.
> it's *1000% better the your sample size of one.
Well, they don't, because the statistics which you cite are not
representative -- they certainly do not represent the hardware I use to
work on FreeBSD. I do not own any SSE2 systems. I didn't even know a web
site called bsdstats.org existed until you informed us of its existence
on this mailing list.
I am concerned that a somewhat shallow argument is being put forward
with incomplete statistics from a single source inappropriately being
used to support it; your argument appears to lack real consideration of
the software engineering effort required to put it into production.
Nikolas, to be fair, if you are willing to do the work, or financially
or otherwise support, the work necessary to support the changes you are
suggesting should be made (and I responded to you in private explaining
that I understand some of the motivation behind your suggested changes),
then it would seem to me to be reasonable and fair for the Project to
consider them, and I am sure they will be of interest at the appropriate
Until such time as these systems are more common, however, then I'm
afraid your suggestion has to be abandoned, for the time being, because
your position appears unjustified in terms of the statistics you
provide, and your arguments.
The proposition you make is a very long term proposition indeed. As
such, I speculate most of the people working on FreeBSD will continue to
focus on more important things in the immediate future.
More information about the freebsd-current