[Bf-docboard] Blender Proceedings

[Roger Wickes]

Sorry to have caused some ill will between people. But sticking to the 
facts,,,Regarding video, having authored hundreds of tutorial videos available 
for subscription and for free, and having watched tons, they can be good, but 
they can be bad and annoying, depending on how they are done, and what they 
are billed as versus what they actually deliver. 

Video is knowledge encapsulated in a moving visual form. That knowledge could 
be flawed and that is the disadvantage, since the user community cannot 
annotate or correct. However, it is powerful and plays a big part in the 
future. The knowledge management plan must speak to this and set up a 
framework whereby we:
	1. Identify the video that it even exists. There are thousands. It's too 
much for even two people to maintain a list. it needs to be crowdsourced. 
	2. attribute the content of the video by audience and subject matter and 
blender version and whether it has audio and if so in which language, and 
whether it is free or if it costs or if it is packaged on a DVD somewhere and 
where it can be found or ordered. 
	2a. Allow the community to either assign attributes (distribute the work) 
and grade it
	3. Provide a search engine so people can find knowledge, video 
specifically, by attribute, within this knowledge domain. "beginner spanish 
animation video". The search engine could be something hosted locally, or be a 
BoF kind of google/aggregator. for example, right now we have YouTube. doing 
the above query prefixed by "Blender tutorial" on youtube returned about 10 
results and referrals to 100 more. Video Aulas este es muy `util. 

> 
>    1. Videos would dominate life in future, and overtake the written
>    instructions (wiki/pdf).
>    So we better plan for this medium.
> 
>    It can easily be hosted on Youtube etc., by having a *Blender channel*,
>    (bidirectional links between Blender wiki and Youtube).
> 
Blender Foundation has a YouTube channel, www.youtube.com/BenderFoundation. It 
is awesome and growing by the day. So does David Ward, so do I. 

The wiki fully supports hosted video. Mindrones or maybe it was martin wrote 
us a nice little plugin. See my mocap tutorial as example of how the wiki can 
embed video quite nicely. Almost any other content engine can as well. 

>    2. About translations, my point was, we cannot plan for translations
> with just 3 people.
>    However, in wiki, we would not be blocking anyone from translating an
>    existing page.

We can and do and must plan for foreign language translations, and facilitate 
them in a non-blocking manner. 

>    5. I do not see how the "wiki" and "sandbox" concepts mix.

Wiki authors should use Sandbox as a temporary pre-release place for them to 
complete a logical unit of work, before moving it into production. We should 
not put half-baked garbage/outlines into the wiki, but as an author, I need a 
place to put my notes and outlines so that I can work on it from time to time.

Our tools must support this concept that writing or documenting some knowledge 
may not be able to be accomplished all in one sitting by one person. The tools 
must support a person working over time, building and refining knowledge until 
it is ready for prime time. There must be at least a  "development" and a 
"production" space for knowledge authors. It would be nice if it the tool 
supported a "testing" status, where the author wrote what they thought, but 
would like someone else to check their work. When you write a book, you have a 
technical editor that checks your work. They catch stuff before the book goes 
final. It's a good process, and our tool should support this kind of "test" 
space or connotation. An end user can use this knowledge at their peril, but 
should have comfort in knowing that "production" knowledge is good and 
accurate for the version it documents. 

regarding ideas and coordinating multiple people; I put in place the wiki user 
manual To Do list. I even documented the process and used the status 
templates. See the wiki for how it is supposed to work; basically a manual 
ticket system. Anyone can add items to be documented, anyone can grab an item 
and fix it directly, or make an edit and ask someone to check their work. It 
never caught on as a work management system; perhaps because it was manual or 
required you to be a wiki author or whatever. But, imho, Basically, 
identifying bugs and defects and missing knowledge should be crowdsourced into 
a bug-tracking system. We should allow the community to identify what 
knowledge it needs through some tool or system. 

Then the community of authors should be able to grab items they want to work 
on, and then close the bug when they commit the change...just like software. 
Right now, the blender code base is not managed that way for very good reason, 
but we do not need the same types of controls over knowledge content.....or do 
we? The plan needs to address the type and level of control over knowledge, 
and the knowledge acquistion process, and the maintenance process and workflow 
and specific tool use to support the process.