Moodle, Sakai, etc.
A public answer to a question about my experience with/thoughts on Sakai or Moodle …
I have not used Moodle or Sakai except for playing around with demos, which does not really qualify me as an expert. But that aside, in each case I came away feeling something like Michael Feldstein; that all LMS’s are pretty good/bad. On my superficial look, however, I would give the nod to Moodle for three reasons:
- Usability: notwithstanding absolutely atrocious default design aesthetics, Moodle seems to be pretty clean and intuitive.
- Hackability: Moodle is built on PHP, not Java. While I really dislike PHP, there’s no denying that it’s widely used, and easy to find people that can work on it.
- Community: there seem more people contributing to Moodle’s development; I suspect this goes back to my ‘hackability’ point. OTOH, when I read comments like this from people I respect, I have to wonder about Sakai. Projects without strong communities tend not to do well.
So I’m not exactly supremely impressed with Moodle or Sakai, though I think either are at least as good as Blackboard.
In my ideal world, however, I’d really like a seamless melding of the traditional course-centric LMS, with the more free-flowing learner-centric model enabled by social networking applications like Elgg. This is the vision behind the in-progress Sakai 3 effort (try the demo; promising, but not as nice as Elgg). It is also surely what will be at the core of the nascent Pinax-LMS effort (Pinax is a new generic social networking and site development framework, so the LMS features would just be added as modular applications). I would expect that the best way to achieve that now is some sort of integration of Elgg and Moodle, though I am unsure of how seamless that integration can be technically.
Anyone with more direct experience with any of this have feedback?
Not related to this post, but I thought this beta release might interest you:
Biber is a bibtex replacement for biblatex written in Perl (in case you had forgotten…).
See http://biblatex-biber.sourceforge.net
Besides the tarball of a CPAN-like distribution, I also provide a PAR archives for linux and a standalone executable for Windows, so you can experiment without having to install anything.
For now its primary aim is doing the job of bibtex+biblatex.bst as accurately as possible. Thus running “bibtex foo.aux” and “biber foo.aux” should output two functionally identical foo.bbl files. I think this objective has been reached, minor bugs excepted.
The bibtex data is handled by the C library btparse via the Perl module Text::BibTeX, so it is fast. If this is unavailable (as I believe it is impossible to install it in Windows except under Cygwin), there is also a pure Perl BibTeX parser built-in that can be used as a fall-back.
I have also developed a rather sophisticated XML format for bibliographic data which defines all the fields required by biblatex, and much more. The current RelaxNG schema already makes it possible to account for multiple languages and scripts, translations, transliteration, etc. That should eventually make biblatex a truly multilingual tool. I am really serious on this latter point, as I personally have a real need for it in my daily scholarly work. For the time being, I have plainly named this XML format “BibLaTeXML”, but I would be really grateful if someone could suggest a more exciting name!
For bibliographic management, I believe BibLaTeXML offers a richer data model than MODS. It is also much more flexible and accessible than the traditional BibTeX data format. You can experiment with it by converting your bib files with the script bib2biblatexml (LaTeX encoding is thereby automatically converted to UTF-8). For each data file with an xml suffix declared in the aux file (or given at the command line with the option -d), biber will automatically parse the xml data instead of a bib file. This parsing is done by LibXML, which means it is really fast. I have also successfully experimented with 8000 entries of biblatexml data in a BerkeleyDB XML container. This possibility should be of interest to users or groups handling large or huge collections of bibliographic data in XML format.