httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Roy T. Fielding" <>
Subject guidelines
Date Wed, 01 Apr 1998 08:58:40 GMT
A couple people have mentioned to me that I am being too harsh in
just coming out and vetoing a piece of code, without prefacing it
with a long discussion of alternative solutions.  I do apologize for
that, but I also don't think it should be necessary.  There is no
question that I do need to justify my veto before taking any action,
and hopefully I've done that by now, but asking me to just go along
with the status quo until I've convinced everyone else that I am
right is not going to work either.

Reviewing the Apache source code and keeping up with all the project
discussions is a significant drain on my time, and one which several
other companies would gladly pay me six figures to do for them.
It is probably my most significant contribution to the project; well,
either that or just keeping the project from splitting apart every time
there is a personality dispute.  I have no intention of voting on
something when I am not engaged in the process; people may have noticed
that I spent part of last week going through old STATUS items, and Dean
is painfully aware that I was plugging away at the source because
I roped him into a misguided search for code bloat on Friday.  You can
generally tell whether or not I am paying attention by how quickly the
HTTP questions get answered on the mailing list.

The only way the Apache Project can make decisions without bursting
into flames is if EVERYONE abides by the project guidelines.  If the
guidelines aren't working in the best interests of the project, then
propose a change.  Pissing and moaning every time someone disagrees
with you wastes far more time than just resolving the conflict.
The fact that the guidelines are currently up for approval was very
much in my mind when I decided to send the veto message -- it makes for
a good test case.  Just keep in mind that there are more ways to
contribute to Apache than hacking in new features or fixing bugs.

I've been around a long time and it surprised me how much more shock
is now accorded to a veto than it was when we started.  It isn't the
end of the world folks.  It is an expression of an opinion. Granted,
it is a very forceful and significant opinion and one that cannot be
ignored, but that is exactly what the situation called for.  If you had
the same background as me, then that would be clear, but then I wouldn't
be needed to review the code either.  It is the strength of the Apache
project that we have enough real experts that when I make a stupid
assumption about DNS, Dean can (and will) correct me, or when I try
something even vaguely similar to a buffer overflow, Marc can (and will)
correct me, etc.  You shouldn't be surprised when I have a very clear
opinion about software engineering -- forming those opinions and
professing them is, after all, my real job.


View raw message