[Bf-python] Introducing the BPyC (a script repository)

[Levi]

Warning: This is a long e-mail!


My name is Levi Schooley (masquerading on the elysiun and blender
forums as "reD_Fox"), and I am attempting to rally support for an
online script repository that I have been calling the "Blender
Python Consortium" (BPyC).
My credentials? I don't have many. I'm a 20-year-old computer
science student that has been with blender since the summer of
2000. That's about it. I haven't familiarized myself with the
blender development (downloaded the source for the first time last
week), am not an exceptional coder or artist, and have very little
to show for the many hours I've spent with blender.
Anyway, after that wondrous introduction, now to the real reason
I'm sending this e-mail.

I have posted an initial proposal on both the blender and elysiun
forums:
http://www.elysiun.com/forum/viewtopic.php?t=46772
http://www.blender3d.org/forum/viewtopic.php?t=6475

Perhaps some of you have already seen them. Based on a comment on
the blender topic by levon, I decided to join the mailing list and
broadcast a refined proposal to a more "focused" group. This
proposal is created mainly because of the following circumstances
(which have been posted on the forums; sorry to parallel content):

1. Disarray. Currently, Python scripts created for Blender are
scattered across forums, authors' websites, and other web (and
non-web) locations. There is no central location for retrieving
even half of these scripts. This leads to...

2. Circulation of out-of-date versions. Often, links to certain
scripts from web sites will be out of date, or will be hosting a
version of a script that has since been updated by the author.

3. Duplication of effort. Without an organized system, many
scripts are created "twice" because one of the authors was not
aware that a script had already been created to fulfill a
particular need.

4. Fragmented authorship. Very few scripts have been created as a
"team effort." Most script authors work (and struggle) on small,
fragmented scripts, or are working on scripts that could be
benefited by a multi-person team.

5. Limited script inclusion with the Blender releases. The
official Blender release cannot and should not include every
single script. The download would be too large, and the
maintenance overwhelming. Not only that, but it is often hard for
the authors to "perfect" their scripts for release status.

6. Lack of synchronization. There have been some efforts to help
users keep their "official release scripts" up to date. However,
there is no real mechanism in place to help users stay up to date
with any of the other scripts they might have chose to add to
their collection.


I have been working on the "theoretical design" of a system that
will combat each of these problems. It has (as I see it) mainly
three parts:
1. The committee
2. The online database
3. The authors

Of these, the database will be the most apparent/important.
Everything should hinge around a central repository. It will
consist of a database of scripts tagged with metadata such as
version numbers, authorship, screenshots, and so on. The database
will need a web-based interface, and so I also include the people
who maintain the website in this category. Also, the database will
need some mechanism for the user to be able to download and update
scripts.

The committee will need to be a specially qualified team of people
experienced in Python (and other) programming, and familiar with
Blender's workflow and development. I was at a bit of a loss for
how to find these people since I haven't been involved with
Blender's development. However, as I browse the bpython mailing
list archive, I see that many of you who are active on this
mailing list may already form a type of "committee." In other
words, I'm assuming that as the repository is created, the
committee will form naturally.

The authors will be the "work force" of the repository. Too many
attempts have been made to provide a "comprehensive list of
scripts" that have blossomed for a short time, and then fade away
without anyone knowing the difference. I think one of the reasons
these efforts have not been successful is because it was the
responsibility of a single person to maintain the list. This is a
recipe for burnout. The solution that I see is to make it the
AUTHOR'S responsibility to maintain the scripts. They're the ones
who wrote the script. They want others to use it. So, they should
be given the opportunity to maintain and update it in a place
where others can readily access it. The trick will be to advertise
the repository enough so that the majority of script authors will
accept and use it. If only a few authors contribute and update
their scripts, the repository will be useless.

So far, most of what I have thought up has been fairly
"high-level" in detail. I do have some lower, implementation-level
details that can be presented later.
Initially, I wanted to get some feedback on the feasibility of
such a project.

1. Do I have the support of those (you know who you are) actively
involved in the blender-python development? Does anyone have any
objections/resistance?

2. Do I have support from script authors? Would you be willing to
maintain your script in the repository?

3. What steps do I need to take to get this integrated with the
foundation?


That's about all I really need to know for now. Issues regarding
the web system, download manager, and other implementation
specific details can be discussed as well.

Thanks!
Levi



 

________________________________________________________________
mail.eaglecom.net powered by Eagle Interactive