Firstly, with regards to objections to using Wikipedia for your project, see http://en.wikipedia.org/wiki/Wikipedia:Replies
well, you've got to realize that they primarily aren't my objections, but objections of researchers that I've talked to in sounding out about doing this. Since they have been working in research and publishing pretty much all their lives, I thought it wise to give them credence.
I think that we can have two possible attitudes to this. One, we can say - 'no, you're wrong, look it up here'.
Two, we can say - 'hey look you've got a point. Wiki's never been used for collaborative research and publishing before. why not acknowlege this, set up a new separate forum to try it and you can give us feedback on what works and what doesn't.'
Personally, what I'm aiming for is two. If they find out that they're fears were really unfounded, we'll find that out too. But I doubt that'll be the case - just looking at mediawiki I don't see it being completely suited to the task for the reasons I mentioned in the last email, as well as those below.
#1 - Just copy and paste the text of a page you think is good enough
... and go through the headaches of taking yet another input source and merging it on an ongoing basis with the already revision addled manuscript that they have when passing it around to other collaborators? That's pretty much the last thing they want to do.
I think you assume that its easy to get text out of a proprietary document that has revisions turned on and format it for the web and submit it there via wiki. This isn't so. They have their hands full just dealing with bugs in word.
The one chance that wiki has is to make it far simpler to collaborate online than it is to collaborate via email. Its strength is to *replace* the current bug-ridden system with something new. And that's where people would listen.
#2 - you can linkto/reference old versions of a page, which are guaranteed not to change.
good point - what's the syntax for this?
What do you mean by #4?
1) no way to input data for graphs
2) no set way to see how close a page is to publication (rating/editorial system)
3) no way to specify 'questions' - topics plus hints on how the topic should be answered.
4) no simple way to guarantee that you have your own namespace and hence don't collide with other topics.
In addition, the ability to poll and collect information by form would be helpful, as well as the ability to moderate submissions.
Anyway, like I said IMO the best way to do this would be to make a new sister project, where we were free to experiment.
And the best, best thing possible would be if mediawiki was modular enough that more than one frontend for editing/publishing could be written in terms of it - that would save a lot of headaches when it comes to writing the wiki that is a word replacement and actually turns wiki content into publishing material (PDFS, etc).
Ed