httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Paul Richards <>
Subject Re: something to think about for 2.0...
Date Fri, 11 Apr 1997 10:31:20 GMT
Jim Jagielski <> writes:

> Are they setup the same way "we" are (ie: group votes, etc...)?
> If not, then I don't think that's a fair comparison.

No, FreeBSD isn't but that's my point. This project isn't structured
correctly to work effectively. The group voting thing was supposed to
have been re-introduced in it's strictest form because apparently we
were in beta test. 

It's only haphazardly applied anyway, I've seen quite a few commits go
in without any voting taking place and then at other times people jump
all over you for commiting without voting. It seems to depend whether
the "voters" approve of the change or not.

We've been down this road before, I think voting is unecessary and
seriously impedes progress. Voting does not enforce peer review. Not
voting does not preclude it. It's more a question of control, some
people expect to be able to sanction all changes before they take
place rather than allowing development to flow naturally.

All I can say is that Apache is getting to be a worse product not a
better one as it matures (i.e. it core dumps at the end of beta cycles
when it didn't at the beginning) and other projects that have a freer
development environment (e.g. FreeBSD, Linux, NetBSD, OpenBSD,
XFree86, etc, etc) seem to have improving products close to release
and much faster releases cycles for far more complex products.

>From what I've seen over the last few months, voting is based on
whether people see it as being a good idea and not because they've
seriously reviewed and tested the changes.

I firmly believe that in future there should be a branch at release
time and people with release engineering experience are put in control
of the release branch. As an example of changes that absolutely should
not take place during a beta cycle are the bcopy changes. It was a
totally gratuitous change that certainly wasn't a fix for a specific
bug and has caused compilation problems that have required further

  Dr Paul Richards. []
  Originative Solutions Ltd.  []
  Phone: 0370 462071 (Mobile), +44 (0)1865 843155 (Elsevier)

View raw message