strace broken in 7.0?
timo.schoeler at riscworks.net
Fri Jan 11 05:06:45 PST 2008
Thus Dag-Erling Smørgrav <des at des.no> spake on Fri, 11 Jan 2008
> [moved from -current to -chat]
> Timo Schoeler <timo.schoeler at riscworks.net> writes:
> > Part of the discussion on bsdforen.de was some people complaining
> > that parts of the base system (as well as [important] ports) do not
> > work, and are maintained very badly due to 'no interest'.
> Those people clearly don't understand the FreeBSD development model.
So, the FreeBSD development model does neither care in case the tree
breaks (read: is unbuildable, or in the direction of being unbuildable
soon, or in the direction of becoming of lesser quality than it was in
any time before)? I doubt this.
> This "no interest" bit is completely fundamental. Code gets written
> by the people who have an interest in it, either because they need it
> themselves, or because they find it intellectually satisfying, or
> because someone donated unsupported hardware, or because someone paid
> them to do it.
Setting the last two points you made aside, it's good, but not perfect.
Code should be written and commited if it's needed. The example of the
HD sound was bad, see my other mail (later).
> Earlier in this thread, someone mentioned txp(4) as an example of
> unmaintained code in the base system. It is a very good example of
> the above. It was ported from OpenBSD seven years ago by someone who
> was paid by his employer to do it, and had access to hardware and
> documentation. I doubt anyone else in the project does, because the
> hardware it supports is obsolete (it was discontinued five years ago),
> and was not widely used even when new.
As long as it doesn't break things or causes regressions, I'm fine with
it. But: Even if it doesn't cause any harm, every snippet of code
should be taken care of by a person (or a group) that's liable for this
code. That's a crucial point, IMHO.
> > FreeBSD seems to decay from release to release; don't get me wrong,
> > I *do* like new features, enhances SMP, ZFS, all this stuff, no
> > problem. *BUT* please don't forget the basis. It wouldn't surprise
> > me if ls(1) doesn't work when 8.0 is released.
> You're over-dramatizing; I hope you don't really believe in what you
> just wrote, because it's completely wrong and unjustified.
Sure, this is a rhetorical thing ;)
(I don't ls(1) to break, but I wouldn't be surprised about races on
six core machines, that don't appear on quad or eight cores. That's
what I mean.)
> Dag-Erling Smørgrav - des at des.no
More information about the freebsd-chat