One of the defining features of PlanetMath is that it is based around a commons. Currently, this commons takes the form of several sections — encyclopedia, forum, papers, expositions, books. Over time, several drawbacks of this scheme have come to light. The current arrangement is somewhat rigid, in that there is little way to shift content between these areas, share content, or link between them. Rather frequently, there arise disputes over what belongs in the encyclopaedia section.
In a recent discussion on this topic of what does or does not belong in the encylopaedia section, Chi Woo proposed that there be a place for material which may not belong in the encyclopaedia. Combining this suggestion with thoughts about how to redo PM using the scholium system and various ideas and suggestions which have been raised over the years, I came up with the following proposal for how to re-organize PM.
In this scheme, there would be two main components, the commons and projects organized about the commons. The commons would be a rather undifferentiated area to which any user could submit content subject to some minimal restrictions:
However, there would not be restrictions placed based upon appropriateness of subject matter.
On top of this commons, there would be a number of projects. These projects would take materials from the commons, organize it in various ways, and present it to the general public. An example of such a project might be the encyclopaedia. This project would be owned by some group, the governing body of which Chi spoke, who would make decide what belongs in the encyclopaedia and what does not. Based on their rules, there would be an encyclopaedia section which contains material from the commons which they consider fit for inclusion in an encyclopaedia.
In this scheme, there would be no one official encyclopaedia. Rather, any member or collection of members would be free to build a project on top of the commons provided that they are willing to put in the effort required to maintain that project. The PM official organization and staff would take a hands-off, common carrier attitude and let members sort out matters of how to run their projects between themnselves as long as they do so in a civil fashion. For instance, if someone had a serious disagreement with the governing body of the encyclopaedia but, despite repeated discussion, it became clear that that person and the governing body could not reach an agreement, then that person would be free to make an alternative encyclopaedia section which would include content selected according to differrent criteria. Both projects would draw upon material from the same commons, but would make differrent selections and display it differrently.
In addition to encyclopaedias, there could be all sorts of content projects. For instance, there could be textbook projects, handbooks, frequently asked questions, problem books, and whatever other genres clever people think up. --rspuzio
A pretty good description of how the scholium system would work in practice, I think! Of your three conditions, (1) is too vague, and (2) is subsumed within (3). --jcorneli
I don't think that 1 is really much more vague than Chi's guidelines about not insulting other users or posting offensive content. Of course, in actual practise, any such guideline will need to be made more specific, just as we are now doing with Chi's guidelines. As for 2 and 3, I reworded 3 to make clear what I was getting at.
The fundamental point I was trying to make with this is that there is really no good reason to assume that there has to be just one encyclopaedia or one rating system and no reason to fight about what belongs there but that there could be multiple such entities sharing the same commons. For instance, rather than squabbling endlessly over whether Smarandache stuff belongs in the encyclopaedia or what to do with the entry on 42, maybe a saner approach would be to trifurcate the encyclopaedia. One branch might be a straight and narrow mathematical encyclopaedia which only includes material which is cited in established textbooks, another might be an encyclopedia of mathematics according to Smarandache, and yet another would be a popular encyclopaedia of mathematics which leaves out the abstruse, ivory tower stuff, but includes material at which professional mathematicians would thumb their noses and consider irrelevant. However, all three of these projects would build upon the same commons, so if a particular entry is deemed suitable for inclusion in more than one of these projects, one doesn't have to rewrite it in order to include it in one's project.
I also think that friendly forking could be used in other potentially contentious places. In particular, I have in mind the issue of people not being able to come to agreement about a correction. Rather than having this be settled by one side forcing ownership away from the other, a more amiable appproach might be to bifurcate the entry; in addition to the original version, there would be the version with the contentious changes made. In the case where the entry were used in the encyclopaedia, what would happen is that the changed version would be included rather than the original version, but the origianal version would stay in the commons where anybody could access it or reuse it as they saw fit. To me this approach seems most consistent with the fact that people retain copyright to their contributions but permit reuse.
--rspuzio
I would suggest (based on my experiences with this subject!) to be careful about possibly confusing blending of layers – there are some abstract principles here ("we could put things in different collections"); and there are some specific proposals ("modified entries would go into the encyclopedia").
The first proposed guideline is an example of one such specific proposal. I am happy to discuss these sorts of considerations, but I do want to be clear that they are a rather different kind of "axiom" than the ones that describe the more abstract systems they refer to.
Specifically, with regard to this potential social axiom: I don't really give a shit about being non-offensive! I envision the scholium system being used for all kinds of things, not just the ones that people find decorous. One of the nicest things about the scholium system is that people can decide pretty easily what sorts things they want to see, and then avoid everything else most of the time. Thus, the design of the scholium system should obviate the implementor from trying to enforce (or even require) impossible puritanical standards like thou shalt not insult another person or group. That's what I was getting at when I said that the condition was "too vague".
I would like to be able to connect the discussion to fundamental issues like this as much as we can. For example, how can we make sense of the requirement that submitted content be FAIF? What are "suitable terms"? At what level is the ambiguity in your choice of phrasing intentional? How is the social principle of FAIF reflected in the technology itself?
--jcorneli