Re: [Springframework-developer] TSS Symposium

[Rod J. <rod...@in...>]

I'd just like to say that this was a great event. I really enjoyed it, I
felt people were really receptive to what I was saying--including Spring
JDBC, transactions, AOP CMT, bean factory concept.

I also got to meet and talk to some really interesting people like Cedric,
Floyd, Mike Cannon-Brooks and Gavin King. And even got to have a serious
conversation with John Crupi.

This was a great opportunity for the Spring community. Now my book is back
in circulation, we should also benefit from that. Sure, there are code
changes (and many enhancements) but explaining Spring over and over to
different people made both Juergen and myself aware just how simple and
consistent the overall concepts are--these haven't changed, which is one
reason that we now have a lot of code but not necessarily more complexity in
Spring apps.

Btw, it was great to meet Juergen! He did a lot of good work as an
ambassador for Spring, as well as picking up other talks. He even won a
group of people at a table, including me, a Tivo each in a trivia game!!
(Don't know if it will work in the UK yet.)

Maybe I should start a blog. Here's my brief summary of some of the content
I liked/didn't like:

- John Crupi's talk on next gen tools--not convinced Sun really "get" this
space
- Bill on JBoss AOP: not a great presentation (a bit flat), didn't really
justify their decision to use field interception or address performance and
other potential problems. Did show some nice examples of JBoss internals
implemented with AOP. I thought Bill made a good case for implementing a
J2EE server with AOP, rather than opening their AOP up to developers to lock
them into JBoss. Bill is charming and very normal, whatever impression JBoss
gives overall.
- Mike Cannon-Brooks: really good speaker (but then he would be, being a
fellow Aussie). XWork presentation was interesting--I like our way of doing
things, but XWork is elegant in its own way. (We have lots of stuff they
don't, though, such as JDBC, JTA etc.)
- MDA presentation based on Pet Store. I'm just going to say "unconvincing"
- Erik Hatcher: Excellent speaker (and nice guy), wish I could have caught
all his talks, but it was hard when I was presenting/networking. Very
pragmatic and wise regarding Struts limitations.
- Panel discussion: really interesting. Basically the "status quo" vs
"reduce complexity" schools coming into conflict. The audie- nce laughed
when Jim Knutson (from IBM) try to defend entity beans after I'd used them
as an example of design by committee before implementation not working.

I detected some themes throughout, such as:

- Entity beans suck (not just from yours truly). Vibe around Hibernate.
- Struts is sub-optimal. Even Erik Hatcher (Struts/XDoclet contributor and
guru) knows of the problems and is open to XWork/Spring. Maybe Struts really
is so bad that it can't even survive as a virtual monopoly.
- EJB is overused (my talks on this were very well received)
- AOP
- A real open source vibe. It was great to talk to Gavin, Mike & Jason
(XWork etc.). We're all doing great stuff and we're all pushing in the same
direction, if with different code bases.

It would be good to have Juergen's take, too.

Regards,
Rod

P.S. If they do another TSS symposium, try to go! It was good--way better
than JavaOne.