> I am not sure I want to. Poul Henning Kamp, of FreeBSD fame, once said
> (freely interpreted) ''I would love to get only 1% of the Linux users,
> provided it is the right 1%''. While I do not always agree with phk,
> I must say there is some truth in this particular statement. 
> I think, emphasis think, that there is correlation between that 
> interesting 1% and those who researches the capabilities of an
> operating system on their own. 

Personally, I would love it if that 1% would be commercial developers
that could spend 1hour or so of their time creating native NetBSD
binaries of their applications.  Another 1% could be OpenSource
developers that can be convinced to remove "linux'ism' from their
source code which make it very difficult to port their software to
NetBSD.

OpenSource is great, but many OpenSource apps commonly used in the
Linux community don't compare to commerical versions.  NetBSD needs
native Macromedia Flash, Adobe Acrobat Reader, Sun Java, wxDesigner,
CrossOver Office, etc.  Yes we can run some of these using Linux
emulation, but not all, and not for NetBSD novices.

Prehaps a NetBSD volunteer committee can contact very companies and
offer assistance to companies with setting up NetBSD boxes for
porting.  For paid software, like wxDesigner & crossover office,
the NetBSD website should maintain lists of people willing to buy a
copy with a native NetBSD port, to encourage companies to spend the
trival effort to port their Linux software to NetBSD.

Prehaps something cool, like a bootable Knoppix like CD/DVD, can be
made for NetBSD, so that developers can pop into their computers,
boot it up, and easily port their code.  I personally used Knoppix
many times to port NetBSD code to Linux, especially since it has a
current gcc-3.3 compiler compared to the very dated gcc-2.9 compiler
in pkgsrc cross-linux.

Linux desktop users suffer from Win32 application envy, and NetBSD
desktop users suffer from Linux application envy.  This needs to
change if NetBSD is to be freed from its ghetto in the server and
embedded space.

And of cource adding critical features to the NetBSD kernel/userland
like ACPI sleep (already in Linux/FreeBSD/Win32) would be really nice.

Alicia.