[haiku-development] Re: [PATCH] libroot improvements
- From: "Axel Dörfler" <axeld@xxxxxxxxxxxxxxxx>
- To: haiku-development@xxxxxxxxxxxxx
- Date: Sun, 01 Mar 2009 12:42:45 +0100 CET
Hi Artur,
Artur Wyszynski <aljen-mlists@xxxxxxx> wrote:
> Axel Dörfler pisze:
[...]
> > Is there any reason why you moved the whole libroot, and not just
> > the
> > parts affected by your work, most notably the "posix" subfolder?
> I thought in that way it will be easier to maintain both version
There are no immediate plans to change anything, so there is little
reason for the duplicated maintenance overhead now.
Furthermore, it looks like you've replaced not only glibc but also most
of our own sources with FreeBSDs ones. Looking at the benchmarks you
provided, it looks like at least some of them don't work correctly
anymore (getpid(), strcasecmp(), getenv(), realpath()).
I don't think it's a good idea to throw away code that we have written
ourselves for ported FreeBSD stuff. Instead, we should (in the future)
just get rid of the glibc/ subdirectory, and import everything that is
missing from FreeBSD or write it ourselves if that is the better
option. For everything we really have to port, we don't benefit from
improvements in FreeBSD anyway, other stuff (like strcasecmp() don't
really need updates anyway).
> > Also, you included lots of non-POSIX folder into the POSIX
> > directory
> > (like arpa/ftp.h). I don't really like having them there, as long
> > as
> > there is a different solution that doesn't hinder ports too much.
> Ah, i did it that way just to separate building process for different
> gcc versions, but you are right and i don't know where is the best
> place
> for them
Some of them are in compatibility/bsd/.
> > That doesn't really sound very likely, though. I would be happy
> > with
> > benchmarks proving me wrong, though :-)
> here :)
> http://hitomi.pl/haiku/system_benchmarks.html
> http://hitomi.pl/haiku/gui_benchmarks.html
> (tests were made in vbox, not real hardware)
Benchmarks in virtual machines aren't really usable unless they hint to
problems like realpath() seems to be.
Bye,
Axel.
Other related posts:
