[Open-education] Import Lesson

[Raniere Silva]

> Raniere, as far as your blog post goes. I think that technically I agree
> with a lot of what you are saying but you could be in danger of over
> thinking it. I think that while a git tech approach clearly has
> advantages for forking and remerging materials, I think operating it
> would create too many barriers for most educators self-publishing
> materials.

Yes, I agree that version control would create a barrier for most educators. I
hope that build something on top of it will reduce this barrier, think of
Wikipedia.

> Rather than try to gather a movement to adopt a new standard. I would
> suggest working within the limits of what epub can deliver is a better
> approach.

EPUB was one of my inspirations. Some issues dealing with epub directly is that
I couldn't handle dependencies on it and for high education many contents can be
originally write using LaTeX, IPython Notebook, RMarkdown, ...

> Raniere, would you consider looking at coding an import epub function to
> the tool you will be using. Can you imagine some kind of workflow with
> epub that invites forking and remerging?

My plan was include a export EPUB function at the first prototype. Use epub as
dependence isn't hard to include. The reason that I didn't stay only with epub
was that I had a bad previous experience dealing with epub due the need to keep
track of all files.

> http://coursefork.org/ had promise (a wysiwyg git editor)
> Kathi Fletcher's OER ePUB editor is based on github too
> OpenStax and Connexions had forking abilities

> If you didn't try it yet try out this grab my books - tool it's pretty good.
> http://www.grabmybooks.com/

Thanks for the links.

> I think (with no evidence but gut instinct) that most reuse is linking

As others suggest, linking is bad due the nature of the web where you can't hope
that some content still be available in one or two years in the future.

> Data portability should mitigate the problem so users of the platform
> can at a minimum archive their own data and upload it somewhere else.
> And ideally it encourages reuse /remix.

The reason that I want to use an distribute version control system (e.g. git or
mercurial) is that it will solve the problem from the begin. Any venture
capitalist can create a WYSIWYG web tool to make the life of users easy but as
far as the "protocol" use an distribute version control system will be easy to
provide "data portability".

Cheers,
Raniere