[sakai2-tcc] [Management] [Building Sakai] UI frameworks WAS> Re: Gradebook Situation

[Clay Fenlason]

On Thu, Jul 29, 2010 at 8:46 PM, Steve Swinsburg
<steve.swinsburg at gmail.com> wrote:
>
> ... if Sakai 3 is all about the client side, might be a moot discussion.

If Sakai 3 is all about the technical choices, I've got a spare noose
in my closet.

I think GWT gets attention because (a) we have a lot of Java skill
already in house, (b) Kuali projects use it, (c) Google, and (d) it
never seems like a good idea to limit yourself to only one way to
fail. It's interesting to hear Eli's thoughts about the first point.

My attitude tends to be that I'm pretty serious about the design-led
part of all of this, and so tend to be inclined toward approaches
which can be more literally faithful to HTML prototypes, and therefore
more resistant to technical considerations fudging things around the
edges. GWT seems like it might be a little like playing the piano with
gloves on, though they may be very thin gloves indeed. That's not a
very good reason, the more so because it hangs on such a nice
metaphor.

The maintenance question is not trivial, but on balance I think we're
better off trying to position ourselves to be able to take
contributions of more than one type over the longer run, though we'll
want to find a sweet spot that diversifies approach while minimizing
chaos. Hedging our bets by thinking both server-side and client-side
seems prudent on the long view. GWT or not is another question, which
I think is Michael's point.

Over the shorter run, we need a focused team tightly coordinated to
get us to square one. We've got an approach to get us there, but that
doesn't make it the last word or the wave of the future.

~Clay

PS Thanks for starting a new thread for the topic.