Improving the handling of PR:s
Matthew D. Fuller
fullermd at over-yonder.net
Sat Jan 12 03:03:04 PST 2008
On Sat, Jan 12, 2008 at 10:27:47AM +0000 I heard the voice of
Matthew Seaman, and lo! it spake thus:
> As I understand it, I think the reason for this difference in
> performance at resolving PRs is because there is a body of ports
> committers that basically expect to spend a lot of time committing
> other people's work, whereas src committers are generally focussed
> on their own projects and tend to commit what they or people closely
> associated with them have developed.
That's certainly a good part of it, but I think there's another
important contributing factor, which is the assumed responsibility.
If I submit an update to my port and anything's wrong with it, of
course, it's my problem. If it screws up INDEX builds or something
like that, whoever committed it is probably in for some heat too. But
if it works as a port just fine (which is the rough equivalent of "it
compiles" in src/), and happens to just be a completely broken release
of the application, *I*'m sure in for it, but nobody attaches any
responsibility to the committer for it.
There's no expectation that they vet all the functionality of programs
submitted. It's not their job to make sure I know what I'm doing;
just that my mistakes don't mess up other people's ports. I'm free to
screw up mine :)
In contrast, if you commit something into src/ that fails somehow,
your feet are on the fire as much (or more) than the submitter's.
There may be nobody that stands up and smacks you with the Shame
Stick, but the general culture tends to inculcate you with an
awareness that you screwed up, in a way that a ports committer
commiting a "correct" update of a port submitted by the maintainer
that happens to be a broken version of the application wouldn't.
That has to have a disincentive effect on your willingness to put
something in, especially if it's in an area of code you don't know
well enough to be confident about all the side effects. And even if
you could wave a magic wand and get rid of that cultural inclination,
I don't think you _should_; it's a proper difference in orientation,
considering the needs and properties of the two different realms.
It's not really solvable because it's not exactly a "problem"; but it
is a reason for the difference.
Matthew Fuller (MF4839) | fullermd at over-yonder.net
Systems/Network Administrator | http://www.over-yonder.net/~fullermd/
On the Internet, nobody can hear you scream.
More information about the freebsd-bugbusters