Worth a listen -
http://www.knownet.com/writing/weblogs/Graham_Attwell/entries/1149667974
I agree with Graham's thoughts that we need to do some serious thinking before we launch into creating PLE applications, which may just limit and put walls around a learner in the same way that LMS systems do.
Some more thoughts on the future of LMSs
Following on from my previous post, Scott Wilson's latest presentation has some good ideas about the future of learning environments.
http://www.cetis.ac.uk/members/scott/blogview?entry=20060408223522
The bulk of this is from a posting I made to a discussion over at http://icommunities.org/direct/LearnDes
where some learning designers where discussing the possible lifetime of existing LMSs:
"While students pay for courses my guess is LMS's will remain at the
core….eg. you cant get in until you pay your fees.ANy other
technology - even like Blogs, are likely to be treated by institutions
(wait for the rules to come out!!) as a kind of LMS in terms of
ownership." Lyn Smith
I think it is the whole paradigm of tertiary education (or education in general) that needs to change rather than the just the LMSs. If tertiary institutions see themselves as the owners of knowledge that students have to pay to receive, why would they want to invest in technology that took a more distributed, social, vygotskian, constructivist approach 
From a personal point of view, and in my field of application development, why would I possibly want to pay money to enrol in a tertiary course, and to then be forced to log into an lms a couple of times a week (because I know the lecturer will be checking the LMS stats as a measure of my participation and learning!), and read through possibly out of date course notes, and take part in contrived discussions , and submit meaningless assignments, just so I can get a piece of paper to say I have done it?
Without the help of a tertiary institution, or a 'learning management system' I can 'manage' my professional learning much more effectively with existing social software tools, email, blogs, wikis, discussion forums, etc. and can learn up-to-the-minute information about my interest domain, share ideas with real people working in the same domain, find answers to real questions at point of need, and apply that knowledge to solving real world problems.
Abviously there is a bit of generalisation here, and I am making these statements as a basis for some possible discussion, rather than as "I know the whole problem and all the answers" The key thing I think is needed, rather than looking at making LMS's behave more like the social software tools, is looking at how Institutions can take part and use the existing social networking/learning tools, and ways of even providing some qualification pathways for people learning in this way. The likes of the Ultraversity degree programme is a good move in this direction.
Have just spent a bit of time with a fellow developer grappling with the issue of maintaining multilingual documentation for an open source software project.
The initial thought was wikis are the answer, put up a basic framework and then let the users take over the development and maintenance of it, which all looked swimmingly good until we looked at the issue of the other language versions (the application interface is already translated into several languages so the documentation base needs to support these users).
Currenlty with the standard wiki architecture there is no easy way to keep different language versions in sync, and to allow fallback to a default language if certain sections are not translated. There is also no way to take a snapshot of the current default language as a basis of your translation as there is no way to edit the link text for each node, you basically have to recreate the documentation from scratch, copying as closely as you can/want the default language set, and adding in the inter-language links as you go.
This problem is solved by your usual application interface language strings setups, as there is a unique key for each string set by the default language, and so a translation can start of by falling back to the default set of strings, and then you translate as many of the strings as are needed - and any new strings added to the default langauge can be automatically pushed out to the translations, so people using the translation at least get the default langauge version until such time as it is translated.
Having thought about it, it appears that the only way of replicating this in some sort of wiki style would be to build an imperialist wiki, in which only the default langauge version was allowed to add new content and structure, and the different language versions could only translate the content and structure put in place by the default language people.
Anyway, we will keep thinking this one over, would be interested in thoughts of others on it …
