My biggest reason for not upgrading so far, is that the things I depend
upon, like Spindle and third party components like tacos, t-deli and
others, are not yet stable on T4. I want to make absolutely sure they
are fully reliable before I upgrade.
I expect the upgrade to be moderately expensive, both in upgrading code
and in learning about hivemind and the new practices in Tapestry. I
expect the biggest concern of these two to be the learning curve.
Inge
Geoff Longman wrote:
I'm interested too to hear more stories from those who have ported
significantly sized T3 applications to T4.
Geoff
On 10/19/05, Eric Schneider <[EMAIL PROTECTED]> wrote:
Pat,
I think your concern is valid. I started looking at 4.0 a couple of
months back (around beta 4) to see what it would take to upgrade our
3.0.3 and 2.3 apps (all of which are large).
I wouldn't say upgrading is a re-implementation. But, if you're
dealing with a large application, I'd give yourself a couple of weeks
(minimum) for the upgrade.
This was the approach I took:
1) Read the Quick Start guide
2) Downloaded the new jars and watched the pretty red squiggles light
up everywhere in eclipse.
3) Ported my custom services (this was one of the difficult steps,
felt like a re-implementation).
4) Ported my shared library components (this could have been easier,
especially for components that extended core components in Tapestry).
5) Hunt down all of the 3.0 API that is either gone or been
deprecated (Most of the stuff that is gone can now be injected, which
is good!).
6) Take a step back, and actually learn some Hivemind basics (mostly,
cuz I had no clue how to inject the stuff that went away).
7) After learning Hivemind basics, I ported most of my singleton
application scope objects into hivemind services (this was way easy).
8) Go through every file in the application and swap out all of the
xml changes (this becomes routine and goes quickly).
I think in the long run, the upgrade is worth it. The hivemind stuff
seems great so far (and I'm still just a novice). Plus, there are
major improvements with form validation and component parameter
handling.
We deployed our first 4.0 app two weeks ago and it's been going
strong (85K+ registered users so far).
I hope this helps.
Cheers,
Eric
On Oct 19, 2005, at 2:32 PM, Patrick Casey wrote:
I've been reading these lists (and looking through the
doc and
change notes) for 4.0 for long enough to see that Tapestry 4.0 is a
*big*
change in the way a lot of things work. Lots of stuff is getting
deprecated,
some things are apparently just going away and, equally
importantly, the
preferred way to do a lot of things is being changed pretty
dramatically.
With that being said, is there a compelling case to
upgrade an
existing Tapestry 3.0.3 project to Tapestry 4.0? I'm not talking
about a
fresh project starting form line one, but rather an extensive,
large and
complex 3.0.3 application. My gut feel is that any upgrade is going
to end
up being close to a reimplementation; is that true? Has anyone
tried porting
a non-trivial 3.0.3 app to 4.0?
I'm worried that I won't be able to port to 4.0 and, in
doing
so, I'll be stuck on an orphan branch of the source tree :-(.
--- Pat
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
--
The Spindle guy. http://spindle.sf.net
Get help with Spindle:
http://lists.sourceforge.net/mailman/listinfo/spindle-user
Announcement Feed:
http://www.jroller.com/rss/glongman?catname=/Announcements
Feature Updates: http://spindle.sf.net/updates
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
