COMMENTARY: The War, or Has Anyone Actually Read ESR?

[Chris Jantzen <chris maybe net>]

I hate writing long posts. However, I care about GNOME too much. I feel 
that without some shifts in perspective we are all Doomed. After some
of today's traffic, this commentary may not be strictly necessary, but 
this subject will come up again. Better now than later, I guess.

Let me say, first, that it is obvious from most of the posts, that
hardly anybody here has actually read esr's Cathedral and the
Bazaar. I strongly recommend that you do: 

http://earthspace.net/~esr/writings/cathedral-bazaar/

First, let me quote from Chapter Nine:

"It's fairly clear that one cannot code from the ground up in bazaar
style. One can test, debug and improve in bazaar style, but it would
be very hard to originate a project in bazaar mode. Linus didn't try
it. I didn't either. Your nascent developer community needs to have
something runnable and testable to play with."

To me, it is fairly obvious that we have none such, and yet we are
hearing clamerings that Bowie's not playing Open Source. This is one
of the bigger misconceptions I see. Bowie was trying to run a Bazaar
model as best he could before everything seems to have degenerated
into Open Chaos. Yes, he appeared to be secretive and it seemed like
he was hiding things. This is NOT a Bad Thing. As we can now see, he
was working very hard on a document. Due to popular request, all
progress has been slowed as it now appears he ratifies each point by
committee. (That's a touch of sarcasm, by the way.)

In my experience, if you feel the need to create a web page and
promote your project into the ground as Open Source before at least
three months of solid work have transpired, then your project is
doomed to failure. Period.

A quote from Fred Brooks, The Mythical Man Month, Chapter 4:

"I will contend that conceptual integrity is *the* most important
consideration in system design. It is better to have a system omit
certain anomalous features and improvements, but to reflect one set of 
design ideas, than to have one that contains many good but independent 
and uncoordinated ideas." [emphasis in original]

He goes on to discuss that any successful program or documentation
should be the product of one or two minds, with the input of many. esr
discusses this, and as I have pointed out already Linus provides a
good example. In the end, they are each In Charge. If you have a pet
patch that you want in fetchmail or Linux and it doesn't fit into the
whole design and scheme according to Linus or esr, then it doesn't go
in. What do you do if this ticks you off? I'll get to that.

Next issue. I heard somebody justify the RSG in the following manner
(paraphrasing): "That's what Open Source is! Look at Apache! They just
took the source and started modifying it!" Further clarification is
needed here. In the coder community there is an evil thing, which I
have seen referred to as the Code Fork. If you fork an existing and
active project for what are viewed as unnecessary and complicating
reasons, you will be shunned and hated around the `net. Apache did no
such thing. They took existing code and modified it, yes. But they
then respectfully submitted the patches to the original
maintainers. When it became apparent that the original maintainers
were no longer interested in the project as a whole and it was clear
that the license allowed them, they formed a new project. esr
addresses this as well in his paper (url referred above) in describing 
the role and responsibilities of a maintainer.

Has this happened here with the RSG? I am not sure. I highly doubt
it. I think what we really have here is confusion resulting from two
projects in one mailing list. Imagine for a moment all the Linux and
FreeBSD people were required to discuss development of their
respective programs in the same mailing list. Or, if OpenGL and
Direct3D developers were required to ask and discuss OpenGL and
Direct3D questions in the same mailing list. Pure chaos would
ensue. Pure chaos is ensuing here. See any similarities?

>From a certain perspective it may be valid to have two or three SG
projects. May the best man win, and all that! We've certainly seen
that before in the Open Source community. What we do not need are two
or three SG projects in the same discussion channel. Especially with
radically different development metaphors.

I'll bring up an earlier point again, to reinforce it, and then close
with a dose of realism. Where the issues at stake are philosophical or
of the mind, democracy has never worked--even the United States is not
a democracy, but a republic. Another way of saying this is, think
about the words "design by committee". Isn't that one of the most evil
phrases in modern society? Look carefully over the literature and the
examples. Every successful Open Source project has ultimately been the
work of one mind or one entity. If Netscape dropped the source on us,
and said: "Okay, go to it! Make it better! We wash our hands of it
all.", Mozilla would be dead and bloated and we'll all be running
Internet Explorer on NT7 in a few years. Instead, they maintain
control and a directing influence over it.

Okay, the promised wrapping up. A quote from Franklin D. Roosevelt:

"It is common sense to take a method and try it. If it fails, admit it 
frankly and try another. But above all, try something."

Two things to be said here. First, this is all about building
something greater than us all. It's not about who's dad has a bigger
car, or who can yell the loudest. Let's all be rational in our
discussion.

Second, let's just get to work and produce something. Let Bowie hunker
down and pump something out. I strongly believe that Tom and his group
should create their own mailing list to develop what is, in reality, a
seperate project. I've offered Tom his own mailing list (heck, even
webspace and semi-vanity hostname if he wants) on my server network so
that he can continue his project with what he feels is the best
development model, without disturbing the UISG's development. I'm not
saying no communication should go between. If you want, subscribe to
both mailing lists. Gleen ideas from both, but keep comments relevant
to the building up of the repsective projects focused in their own
arena.

My two cents.

-- 
chris jantzen kb7rnl =-> 
---------------------------------------------------
systems administrator |    __O   |          student
maybe.net             |  _`\<,_  |     oregon state
possibly the best     | (*)/ (*) | computer science
---------------------------------------------------
  http://www.maybe.net/ - mailto:chris@maybe.net