Problems with fakephp

[Trent Piepho]

On Fri, 28 Nov 2008, Matthew Wilcox wrote:
> On Fri, Nov 28, 2008 at 01:21:22PM -0800, Trent Piepho wrote:
> > Sometimes I just want to give up on Linux.  Is there a different interface
> > that isn't wantonly changed without warning?  No?  Do you not see the
> > problem this creates for developers?  Do you not care?
>
> You see, we didn't know that anyone was using fakephp for real work.
> It's marketed as being a way for developers to test their device drivers
> with hotplug slots even if they don't have a real hotplug machine.
> If we'd known, we'd've done a better job.
>
> BTW, cut the crap about "Sometimes I just want to give up on Linux".
> We're putting in a lot of work that you get to use for free.  That doesn't
> give you the right to be abusive.  And if you stopped using Linux, we'd
> have one fewer person complaining, so I don't personally find it a huge
> motivator to drop everything and attend to your whims.

I apologize if I've come off as being abusive.  I think what's been done to
fakephp is a real problem and being told "too bad" in not so many words or
"you're wrong" when I'm not is very discouraging.

I've been contributing code to open source projects for over a dozen years
and have fixed more than a few bugs in the linux kernel.  When I had the
same problem Darrick Wong did with resource assignment, I didn't whine
about it, I made a patch and fixed it.  But when someone thinks removing an
interface, that's been around for years, with no warning is ok, or assumes
that because they don't know someone they can't have a clue what they're
talking about, how can I patch that?

> > So the race condition doesn't matter?  Alex Chiang wasn't even aware
>
> There is no race condition.  You can't remove a fakephp slot.

Please take a glance at fakephp before telling me I'm wrong, as you'll see
that I'm not.

> > So maybe this better interface should be created before breaking the
> > existing one?
>
> Hey, I have a great idea.  Why don't you help instead of just bitching?

When you get told, "your fix is unacceptable because it fixes a problem I
don't care about," it's clear than any fix to the problem will have the
same failing, so what's the point of even trying?

I was under the impression that you and Alex Chiang knew there were "real"
users of fakephp and just didn't care.  But it looks like I was mistaken
about that, so maybe it's not hopeless.

As a token of goodwill, have this spare PCI patch: