svn commit: r187332 - head/tools/regression/usr.bin/jot
M. Warner Losh
imp at bsdimp.com
Mon Jan 19 18:15:18 PST 2009
In message: <20090120012256.GB29741 at dragon.NUXI.org>
"David O'Brien" <obrien at FreeBSD.org> writes:
: On Mon, Jan 19, 2009 at 06:14:35PM -0700, M. Warner Losh wrote:
: > In message: <20090119165030.GA18409 at dragon.NUXI.org>
: > "David O'Brien" <obrien at FreeBSD.org> writes:
: > : On Sat, Jan 17, 2009 at 02:00:55PM +0000, Robert Watson wrote:
: > : >
: > : > On Sat, 17 Jan 2009, Stanislav Sedov wrote:
: > : >
: > : >> Was this discussed somewhere? I don't understand why we should restrict
: > : >> our filesystem naming because of limitation of auxilary filesystems.
: > : >> There're even more restrictive filesystems available, we couldn't support
: > : >> them all. While previous modifications looked harmless this one seems
: > : >> questionable to me.
: > : >>
: > : >> I understand, this is a bikesched, but I really don't see a reason. You
: > : >> can't build FreeBSD on windows anyway.
: > : >
: > : > Many of us would *very* much like to be able to cross-build FreeBSD from
: > : > both Windows and Mac OS X, which would be highly desirable for embedded
: > : > systems and appliance shops.
: > :
: > : Run VMware. Seriously. Or raise a serious discussion on developers@
: > : about this and get conciseness that cross building on MacOS X and
: > : MS-Windows is a goal.
: > I've tried this on my mac. It doesn't work very well at all, and it a
: > pita for file sharing.
: Sorry - I don't quite follow. What is "it"?
The point I've made several times: minor renaming isn't much of a
pain. It has big benefits to those people that wish to use other file
systems that have different restrictions than we have.
Yelling 'run a VM' isn't a productive suggestion, frankly, and given
the low level of pain for a rename, shouldn't be yelled at all. I'm
glad JUNOS can build inside its own VM. Kudos. But given that the
world is much more than a tightly controlled development environment
at a well funded company, I think that we need to be more
accommodating. Especially for the embedded world.
While a 'vm solution' is OK for today, I think it is too short-sighted
and leaves some significant opportunities on the table. We shouldn't,
as a project, say "nothing can change at all" for these systems.
Rather, we should do a cost benefit analysis. For renaming, the cost
seems very low, and the benefit seems very large. For other things,
we'll cross that bridge when we come to it...
More information about the svn-src-head