You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by "Roy T. Fielding" <fi...@kiwi.ics.uci.edu> on 1999/08/18 04:24:02 UTC

Re: Process abstraction in MPM

>I thought we'd agreed that, and the debate was simply about playing
>games with CVS to make certain people happy? If that's the case, then
>I'd say carry on and do it, and when people who want to play games with
>CVS get around to doing it they'll have a bit more to do.

Playing games?  Why is it that when I suggest a course of action that
will make it considerably easier for everyone to work on 2.0 in the
future, it is now called playing games?  I usually have fun playing
games -- carefully setting up a CVS repository is not fun.  The reason
I do it is because it will save everyone, including me, far more time
in the future than it will cost to create.

Lately, some of you guys have been dead set against anything I do
which makes it easier for Apache to remain a collaborative project,
even when it is painfully obvious that it is a good thing to do.
Take this case for example.  All I need is a merge point.  This is
something any rational person would do when merging trees.  In fact,
Manoj just created the merge point when he tagged mpm as post-merge
up to 1.3.9.  Is this too much for me to ask?

This is really beginning to irritate me.  It is okay for people to be
selfish in developing their own code ideas, pursuing their own goals,
and working on things that they consider to be important.  That is how
Apache is supposed to work.  However, lately people have completely
lost sight of the other part of "enlightened self-interest" -- giving
other people the freedom to pursue their own goals, so that the combination
of all our effort is greater than anything we can do alone.  You can't give
that freedom without incurring some cost, and in this case the cost
was not APRizing the mainline code until after the 1.3.9 release was
merged and tagged.  BFD.

Think about it.  This is the only thing that really characterizes Apache
as a project.  Without it, we just have Ben's project, or Dean's project, 
or IBM's project.  When I say that I can't work on Apache 2.0 without a
clean repository, I am not playing games.  If creating a merge point,
and later swapping repositories (doing a clean checkout after the swap
happens on hyperreal), is too much of a hardship on you guys compared to
the benefit of me working on 2.0, then just tell me so and I'll build 3.0
instead as a separate project.  But stop fucking me over just because a
few months work on mpm and APR seems more valuable than the time I spent
maintaining 1.3.x while you guys were having fun.

....Roy

Re: Process abstraction in MPM

Posted by Bill Stoddard <st...@raleigh.ibm.com>.
> 
> Luckily Manoj got in there before it got blown for you.
It had nothing to do with luck!

-- 
Bill Stoddard
stoddard@raleigh.ibm.com

Re: Process abstraction in MPM

Posted by Ben Laurie <be...@algroup.co.uk>.
"Roy T. Fielding" wrote:
> 
> >I thought we'd agreed that, and the debate was simply about playing
> >games with CVS to make certain people happy? If that's the case, then
> >I'd say carry on and do it, and when people who want to play games with
> >CVS get around to doing it they'll have a bit more to do.
> 
> Playing games?  Why is it that when I suggest a course of action that
> will make it considerably easier for everyone to work on 2.0 in the
> future, it is now called playing games?

Calm down. If we're going to get snippy about language, I could take
exception to being called an "ignorant slut". I didn't, though.

>  I usually have fun playing
> games -- carefully setting up a CVS repository is not fun.  The reason
> I do it is because it will save everyone, including me, far more time
> in the future than it will cost to create.

Well, I'm not convinced, but I'm not going to stop you. My point is that
development shouldn't wait for you to do whatever it is you want to do.
You've stated that I shouldn't object to the work because it isn't work
I'll be doing, but I certainly do object to making developers wait while
you (or whoever) get around to doing it.

> Lately, some of you guys have been dead set against anything I do
> which makes it easier for Apache to remain a collaborative project,
> even when it is painfully obvious that it is a good thing to do.
> Take this case for example.  All I need is a merge point.  This is
> something any rational person would do when merging trees.  In fact,
> Manoj just created the merge point when he tagged mpm as post-merge
> up to 1.3.9.  Is this too much for me to ask?

No, clearly not, and you got it, so what's your beef? I am not against
stuff that helps us work together. I'm just against waiting for things
when there's no need to wait.

> This is really beginning to irritate me.  It is okay for people to be
> selfish in developing their own code ideas, pursuing their own goals,
> and working on things that they consider to be important.  That is how
> Apache is supposed to work.  However, lately people have completely
> lost sight of the other part of "enlightened self-interest" -- giving
> other people the freedom to pursue their own goals, so that the combination
> of all our effort is greater than anything we can do alone.  You can't give
> that freedom without incurring some cost, and in this case the cost
> was not APRizing the mainline code until after the 1.3.9 release was
> merged and tagged.  BFD.

OK, well that's happened, hasn't it? I apologise that I'd lost sight of
your desire to merge 1.3.9 before APRing but what I didn't want to wait
for was your CVS reorganisation, which can happen at any time, surely,
post-APR merge or pre-APR merge.

Luckily Manoj got in there before it got blown for you.

All I'm concerned about is that we don't hit a point where we have to
_stop_ because we're waiting for you to do something. It seems we can
avoid that, so we're all happy, right?

> Think about it.  This is the only thing that really characterizes Apache
> as a project.  Without it, we just have Ben's project, or Dean's project,
> or IBM's project.  When I say that I can't work on Apache 2.0 without a
> clean repository, I am not playing games.  If creating a merge point,
> and later swapping repositories (doing a clean checkout after the swap
> happens on hyperreal), is too much of a hardship on you guys compared to
> the benefit of me working on 2.0, then just tell me so and I'll build 3.0
> instead as a separate project.  But stop fucking me over just because a
> few months work on mpm and APR seems more valuable than the time I spent
> maintaining 1.3.x while you guys were having fun.

Hey, I'm not against any of this, but I don't see what's happened that
stops you from doing it. Ryan wanted to know if he could put APR into
the _current_ CVS tree, and people said he could. OK, I'd forgotten that
there was a sequence you wanted observed but fortiuitously it happened
anyway. Once more, I'm sorry! All you had to do was say so, there's no
need to get steamed up about it!

If there's anything else you'd like done, please say so. There was a lot
of debate, and various courses of action proposed, and as it seems I
forgot one bit, there may be others.

Cheers,

Ben.

--
http://www.apache-ssl.org/ben.html

"My grandfather once told me that there are two kinds of people: those
who work and those who take the credit. He told me to try to be in the
first group; there was less competition there."
     - Indira Gandhi