[sakai2-tcc] Annotation based injection within Sakai components

[Matthew Jones]

I'm 100% on your side, I don't think making any changes that will break
anything is a good idea unless there was a massive upside that I'm not
seeing. I wasn't the one that proposed this. Kernel 2.0 would essentially
mean tools which are no longer compatible with Kernel 1.0.

 On Wed, May 16, 2012 at 3:03 PM, Charles Severance <csev at umich.edu> wrote:

> Matt - How many developers are waiting in the wings to work on the CLE
> where the ease of learning is the reason they are not coding?  Zero.  Lets
> not solve a problem we don't have and break all the contrib tools.
>

I honestly don't know, I think it's potentially greater than zero though.
Last year's bootcamp had an attendance of at least 50. On the sakai-dev
list (if anyone's reading it, mostly it's Swinsburg) has many people that
appear to be struggling with simple problems like getting a Sakai instance
running. Locally, Longsight has institutions where many people want to
contribute but are stuck on problems that have no documented answer.
Possibly because javadocs are missing, hard to find or the *only* guide we
have. Possibly because the tool no longer has an active developer and
nobody knows the answer anymore. Some of this goes back to what Aaron was
saying last year at the TCC about confluence being mostly obsolete
information.

I agree with your other points. I'd say Objective C is a PITA the learn too
but there's serious money and visibly to be made publishing on iOS. Though
nowadays most people aren't going to write something in OBJC, they'll start
with an easier to use framework. Only 5 years later, the days of needing to
learn OBJC and get something fast on the App Store are gone.

The problem with Sakai CLE and OAE is really unless you're employed at an
institution that's using this as your primary LMS (or an SCA) there's no
strong incentive to develop a tool *specifically* for Sakai, unless you're
going to also make money off it (like Kaltura, Turnitin, etc) Have to grow
the community to get new stuff created, no matter how easy it is. And you
want that to easily run on as many versions of the software as possible.

LTI lets you "write once run many". We just need LTI to be able to do a
little more than simple outcomes.

LTI is easy.   But not all-powerful.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://collab.sakaiproject.org/pipermail/sakai2-tcc/attachments/20120516/3060444a/attachment.html