[sakai2-tcc] Annotation based injection within Sakai components

Charles Severance csev at umich.edu
Thu May 17 04:52:27 PDT 2012


On May 17, 2012, at 12:42 AM, John Bush wrote:

> The skinny is, I agree being more stateless is a big step forward in
> that direction, but this problem may prove to be hard to resolve.
> There may be other ways to attack this problem now.  Its worth
> exploration, for the reasons you've outlined.

While stateless is a great idea - actually a small enough state and having only serializable data in state is a great goal - but not achievable without heavy handedness.

I carefully said "reduce state" where there is low-hanging fruit.   Just as an example - JSF abuses state and we can't control its behavior.

But with some of the Velocity changes I made in building the LTI Admin tool, it makes it really easy to move simple data items that our current velocity items put into state onto the URL. 

This not only reduces state but makes things like the back button less dangerous and the ability to do things in multiple windows more safely - so while there is a small but nice performance gain there is a more significant usability gain in redoing these Velocity tools.

And I was not saying this "Velocity State Reduction" is "high priority" - just a better investment than a non-upwards compatible "Kernel 2.0".

/Chuck


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://collab.sakaiproject.org/pipermail/sakai2-tcc/attachments/20120517/3c35a5af/attachment.html 


More information about the sakai2-tcc mailing list