I think that it would be useful to create a section of PlanetMath where everybody can post just any content (not only Encylopaedia entries), just like on wikibooks. The difference from wikibooks is that one can take advantage of LaTeX. There are a lot of (mathematical) courses and books that aleady exist, and they are written in LaTeX.

In fact, I am proposing this because I am personally interested in putting a course on PM. This course is in Russian and I think it is worth translating into English. I have the LaTeX source and I try to translate it. However, it would take a lot of time to translate it if I work alone. Another problem I am not a native English speaker and I do not have many around to ask them to proof-read the translation, moreover, native english speakers I can get in contact with are either busy or they are not mathematicians. Putting it on PM would give such advantages:

• people would report errors
• eventually I can find somebody who can help me translate the course and the translation would be accomplished faster
• eventually someone can try to translate it in other languages

The Exposition section only lets you to post links to books and courses situated elsewhere and does not allow for collaborative creation of content. On the other hand Encycopaedia is not the right place to post courses. I think that PM has all the technical tools necesarry to run Expostion-like section capable of Encyclapadia-like editing. Why doesn't someone just create it ? :)

--dmitri83

Initial reaction

I completely and whole-heartedly agree that this would be a good thing to do. See also the PlanetMath Goals page, where I mentioned (in the goal tree) that it would be good for PM to move to support the creation of free math works in general. You've very nicely spelled out some of the benefits of doing this. I personally think that the book and exposition sections should be completely revamped to only support the creation of free content, but some people do find the links to off-site non-free resources valuable – perhaps these links could be moved to another section of the site. Also see the Grant planning and bidi system pages, where there is a little more discussion of translation in general. Anyway, you have my vote that this would be a worthwhile thing to do. I think probably most people will agree that this is a good thing to do, and any delays will mainly be due to lack of programming time… --jcorneli Fri Mar 25 16:56:20 2005 UTC

Is it possible to just create an (empty) copy of Encyclopaedia and name it something like "Free courses" ? by the way, those users that want to leave links to non-free courses, can always create a course entry and leave a link. The wiki-style engine is more flexible in my opinion, since can do with it anything that you already can do in Books and Expositions section and something else --dmitri83

Yes, sure, you can put references to non-free stuff in the Encyclopedia too. I'm certainly not saying that PM should forbid mention of non-free works, just that it should focus on creating more free works! It would be ultra-super-cool if PM became a sort of collaborative version of Arxiv, so, people would be using it to create research papers collaboratively, but then they might get some corrections, create a revised version of the paper adding the new contributor to the list of authors, and so on. Doing this all really well would seem to require some new tools. For example, I'm not entirely sure linking between separate resources should be done. Maybe site users should have the option to have linking turned on, thus, when reading a research paper or book, they could either see only internal links that were put there by the original author, or they could see additional links to other resources on the PM site that were put there by the Noosphere engine. Also, for creating books and papers it would probably be good to have outline views of articles. Maybe even the possibility of turning each book or paper into a "micro" noosphere instance, with each important object in the paper (definitions and theorems and such) being a noosphere Object. This is beginning to sound pretty heavy, but could be worth considering. Then these objects could be adopted into the encyclopedia, or other papers, too (and we use a bidi system of some sort to keep them in sync…). Yow! --jcorneli Fri Mar 25 19:06:34 2005 UTC

I like this idea too. We should extend PlanetMath into a place people can collaborate on any type of mathematical text. I should point out that there are "collaborations" right now, but they don't really have corrections workflow. We should plan out how some of this would work. It is big and important enough that we should also think about including it in grant proposals. --akrowne Fri Mar 25 17:49:48 UTC 2005

Use wikibooks?

Guys, would it make sense to just use the wikibooks software (I don't know much about it)? --akrowne Fri Mar 25 21:50:53 UTC 2005

We should consider it (discussion of these tools in and of themselves should take place on the Wikipedia or a new "Wikibooks" page), but it seems pretty essential that LaTeX be available, and (as I indicated in the comment above) it would be good to integrate the books/etc. with the PM encyclopedia. Perhaps Wikibooks could be used as a stopgap measure? It probably wouldn't hurt to have links for them on the site. But I have doubts about the long-term usefulness. --jcorneli Fri Mar 25 22:43:09 2005 UTC

Comparison to an existing Oddwiki feature

Let me note that in addition to the idea of making "an empty copy of Encyclopaedia and name it something like 'Free courses'", we should also consider this idea: making "an empty copy" for each book. This would result in a system very much like the Oddwiki system that we are currently using (see link on HomePage). Oddwiki makes a new namespace, new recentchanges page, etc., for each separate childwiki. But in fact there is only one Oddmuse engine running the whole system. It is really quite sophisticated.

An implementation like this would easily enable us to add AsteroidMeta as a "book". And (phenomenally) it would just as easily enable us to add things like PlanetPhysics or PlanetComputing directly to the active Noosphere implementation on PlanetMath. Wow!

Of course, some people wouldn't want or need the full power of a "Noosphere system" for their project. Some people might just want one single TeX file, and this should be supported too. Other might want something more like what you get with out-of-the-box latex2html. (This would really just be a variant on the single TeX file.)

This reminds me of the stuff that logan was talking about on the [[Object_Adaptation?]] page. If we can consider messages and corrections as protocols, we can certainly consider "book" or "mini-noosphere-instance" as protocols, too. We could also (presumably) define some protocols (or maybe I mean procedures) that would transform one of these other objects into an object of the other class. (Maybe we can even get the FEM code involved, so that we could transparently produce nice PDF output for any one of the objects.)

Other extensions (e.g. my favorite, literate programming, but also a bidi system) would make the ensemble really, really powerful… --jcorneli Thu Mar 31 02:58:01 2005 UTC

Note also that if mini-instances could also add further child-instances, we could resolve the issue with naming that drini mentioned on the AsteroidMeta page today. --jcorneli Thu Mar 31 03:02:20 2005 UTC