Talk:All iDevices/Archive

From WikiEducator
Jump to: navigation, search

OK - simple text iDevices are easier to work with with. Not sure how eXe would impement this - but at least its possible to identify the beginning and end of an idevice.

Things for us to think about:

  • Do we focus (as a starting point) on importing a wiki page? - ie in eXe, you point to a page - it scrapes all the text plus iDevices puts the iDevice content into the relevant iDevice in eXe and assigns the page to a node within the tree structure of eXe.
    • I think so - but I also think that this is really a job for the Moz version of eXe. One of the 'problems ' w/ this new eXe is that it shouldn't just replicate the existing eXe - there's no point in that. One way that it can differentiate itself is to be this tool, for contextualising open content resources, starting out w/ resources on MediaWikis. I'll chat with Matthew about making this the focus. He has quite a bit done on it already - there is a working version available that imports a single iDevice and saves/loads an XML file format. -- BrentSimpson
      • I agree that this should a job for the Moz version & that eXe 2 will need to differentiate itself from eXe 1. For example how would eXe differentiate itself if Wiki technology were to build capability for IMS/SCORM exports - shouldn't be too difficult for a wiki. I think eXe could add value (and differientiation) by providing offline capability for recontextualising content - its a powerful "niche" --Mackiwg 07:33, 17 July 2006 (CEST)
  • Given the we are dealing with dynamic content - with an eXe package containing wiki content might want to update. Ideally we would want to provide for a variety of update alternatives, eg.
    • Update a specified page (including all idevices on that page)
    • Option to update a specific idevice on a page
    • Option to update all pages in the eXe package that have for example Wikieducator content
      • This is a bit of a can of worms really and would take some thought. If i've contextualised my content I don't want the Wiki content to then overwrite my changes. Perhaps something towards these type of questions would be to hack MediaWiki to create an RSS or Atom feed of a particular set of pages. So, for example if I can designate that the "Intro to the Amazon" is my 'title' page if you will that all pages all that are heirarchically part of this content create an Atom feed. This way eXe could check on particular content(s) that it's working with and notify the user that such and such a page has changed. The we could probably do something like a 'diff' on eXe content vs the wiki content and display the changes to the user. -- BrentSimpson
      • Thinking about this a little more - I think that you're right about the can of worms. I like the Atom feed idea - that's smart. Also a key strength of the FLOSS movement lies in its experience with versioning in distributed developments - perhaps there are a few innovations associated with versioning and what we're trying to do --Mackiwg 07:33, 17 July 2006 (CEST)
  • For simple text based iDevices - it should be possible to read the lable used for the Idevice in the Mediawiki template and use this as a default for the iDevice label and scrape the icon association. That is - do we allow folk to associate different titles and icons to the ones used in a standard eXe distro?
    • Not sure, perhaps a question for Matthew to look into. -- BrentSimpson
    • This is something we should look into. What happens if a user creates a "new" idevice that's not in eXe. We could agree on a set of defacto requirements for how this is done. --Mackiwg 07:33, 17 July 2006 (CEST)

Lots more questions to add - but running out of time. I think this has real potential and if we think smart - can really do something interesting --Mackiwg 23:14, 15 July 2006 (CEST)

iDevices with a lot of content

Our Short experiance hare at th e Mauritius BC has revealed that some of the iDevices might need to be rather long, and contain complex markup like tables. This makes wikimedia's template mechanisim inadequate as they are not designed to deal with very large (or complex) paramter values.

My idea which I discussed with Wayne is to use hidden DIVs as markers see Future_Templates for details.

Btw just to let some more worms out of the can, how would one deal with interwiki etc like the templte i created for wikipdia (see Help:Tips) which the content editors here seem to have liked?!

Just another 2c: Wikieducator has a feature Special:Export which gives you the raw wikitext which would be easier to diff; but you would need a standalone wikitext rendering engine!!

--Phsi 07:31, 23 August 2006 (CEST)

There are no threads on this page yet.