[Python-ideas] Iterative development

[anatoly techtonik]

On Wed, Jan 29, 2014 at 4:57 PM, Chris Angelico <rosuav at gmail.com> wrote:
> On Wed, Jan 29, 2014 at 11:29 PM, anatoly techtonik <techtonik at gmail.com> wrote:
>
> Here's a suggestion: Fork Python (that's legal, that's what open
> source means) and start development using the model you advocate. If
> it's massively better than what's happening, (a) developers will flock
> to your model, and (b) the project could be completely handed over to
> you, as happened with GCC.

There is a big difference between people who invent things and do things.
I am a lazy bastard who can not do anything and sustain its job, because
he is constantly inventing new stuff that no one is able to implement. Over
the years I realized that the only good that I can do to humanity is to
develop a sustainable model. So far it didn't happen, because it appeared
that people only work on their own ideas. I don't own my ideas - they are
free for everyone to explore and discuss. So if there is anything valuable -
take it. I don't need power over project or money or anything in between.
Next day there will be another idea and another discussion.

It is nice to see communities that can develop ideas, that can realize that
people are different and use the potential of that people are capable for to
a full degree. It is also nice to see the evolution of people to act in a new
roles that are uncommon for them. You won't like it, but it is also nice to
see how people become worse, because they are human species and to
realize that everyone is imperfect. What is not so nice is to see good
things fail, because people can not reuse technology to help them to deal
with human factor.

> Or alternatively, explain to us here what the real advantages are of
> your new model. So far, what I've seen is "hey, here's an idea", and
> not "here's what this idea will do to benefit Python"; and the idea
> itself looks more suited to a big business than to open source. Maybe
> someone who's actually used Agile will know what's so wonderful about
> it, but unless every core dev *has*, a bit of explanation will help.

Ok. In short. There is only one advantage:

- increased visibility

which in turn results in

- increased interest

which in turn results in

- increased participation.


What problem does agile solve. There is one big problem that "increased
participation" is actually the negative factor for existing contributors,
because it takes more time from them.

Where does this "more time" comes from? In current model:

- increased participation == increased communication


If you constantly communicate, you don't have time for development
(probably the things that you like the most). How does agile help with that?

"agile" means just that - "flexible". If you see the problem, you are not
saying "we are all developers, nobody is interested in communications". No,
instead you're saying -- ok, we have a communication problem, what can we
try?

In current model, you can not try anything, because you can not set goals.
Goals is something that is at least:
- Measurable
- Time-bound

There is no time bounds, there is no measurement. These are not part of the
process, so you don't have even any means to solve the communication and
time deficiency problem. If we have two weeks cycle, we can at least set
goals.