httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ralf S. Engelschall" <...@engelschall.com>
Subject Re: Apache 2.0 rant
Date Thu, 18 May 2000 19:51:55 GMT

In article <Pine.LNX.4.21.0005180936260.1057-100000@koj.rkbloom.net> you wrote:

> [...]
> "I didn't think 2.0 had a future" -- You made a bad choice then.  I
> started asking in January for alpha and beta releases, and people said
> they hadn't had time to look at the code.  So, we waited and had our first
> alpha in March.  If you still didn't think Apache 2.0 had a future, you
> weren't paying attention, and that's not my fault.

Be careful: if you think people like me don't pay great attention, you
can be wrong, Ryan. Don't think that just because one doesn't post,
this has to mean one doesn't observe the happenings very carefully. For
instance, I personally read really every message, every commit mail and
still follow the evolution of Apache very very closely. The point is
just that I stopped participating actively in discussions and instead
moved myself to the background to avoid trouble.

One of the reasons for this is that I'm no longer able to have a warm
feeling with the evolution of Apache 2.0's development direction and the
development method. This hasn't to mean that this is the fault of the
group as a whole or even the fault of particular members of the group.
Perhaps it is caused by my own personal evolution and my changed point
of views. Currently I think it's a mixture of this.

But at least I imagine me that a drift in the group occurred over the
last 12 months which seem to ask for the current situation. Spoken in
cruel words: a drift over the last four years away from "we are hackers,
let us make Apache as best as it can be done" to "Apache is the best,
we are the best, let us do something with Apache". Yes, I know this is
ugly wording, but at least I feel a strong discrepancy in the imagined
answers of the group members when it would come to questions like this:

o Would I contribute the same amount of code and time 
  if I wouldn't be (at least partly) paid for this?
  [don't confuse this with being paid for things
  one already did before where one still wasn't paid for it]

o Is my total amount of free software contribution and technical
  experience in balance with my attitude?
  [especially my attitude against the suggestions of others]

o Are my contributions partly inspired by my employer's wishes
  or the company goals or do I still act totally independent and also
  allow (industry-wise) unreasonable decisions if they at least are
  justified from a hacker and perfectionism point of view?

o Do I enough basic research and evaluation before I roll my own 
  solution? Do I search through mailing list archives, source code
  comments and documentation for hints from experienced people or do I
  just trust my own experience?

o Do I take group experiences from the past into account enough?
  Especially do I try to understand old goals and decisions?

o Is it really true that license problems cannot be solved
  technically without having to force a reinvention of the wheels? Do I
  really try to solve license problems technically _AND_ by arrangements
  with authors or I do too often just go the easy way of less resistance
  and ignore otherwise perfect solutions?

o ...

Independent what the answers would be, they could be acceptable (at
least I can image two situations where both extremes of answers could be
acceptable). The only thing IMHO which isn't reasonable for a project
is when the answers differ greatly between the members of the project.
And that's my current problem with the ASF and the Apache HTTP server
project.

> [...]
> Bottom line, if you don't like the code, stop complaining and do something
> about it.  

Err... yes and no, Ryan.

Just a typical example: For instance, I forecasted and warned twice and
loudly that doing Autoconf the fast way will result in a mess and a not
well-thought out build environment which later causes great trouble; I
volunteered to do it the more well thought out and slower way for us;
others wanted a quicker solution and ignored my suggestions; the result
is a build environment which still has essential problems (third-party
module integration, DSO support, etc); I got blamed because I only
complained about the current state without trying to fix the problems
myself.

I mean, hey, no one can expect that I fix things which could have been
avoided if my earlier suggestions were treated more seriously. It is
acceptable (although perhaps not very wise) to ignore suggestions of
others, but it is not acceptable to expect that the same people later
feel motivated to fix the resulting problems - especially not if those
problems essentially were caused by the fact that earlier suggestions
were not treated seriously enough.
 
                                  -

So, a few months ago I was convinced that the evolution of the ASF's
HTTP server project is not reasonable and cause trouble in the long
term. Currently I still think the evolution is problematic and will
cause the trouble. But perhaps not trouble for Apache 2.0 itself (hey,
be realistic - we are all hackers and know that every piece of software
finally runs if one just bashes it long enough ;). But certainly trouble
for old developers like me who are used to different goals ("Unix only",
etc.) and different ways of hacking ("best solutions are just good
enough, independent how much effort it requires", etc.)

Or in short: It's already too late to stop and turn around and perhaps
a turn around is no longer reasonable if one takes all aspects into
account. So guys, don't worry any longer about complainers and go
ahead with the current Apache 2.0 development. You just have to accept
that some people no longer can be motivated to work in the changed
environment...

Yours,
                                       Ralf S. Engelschall
                                       rse@engelschall.com
                                       www.engelschall.com

Mime
View raw message