Well after several days I give. TreeCacheAOP is just impossible for me to get 
set up given the information available so I'll have to overlook it for now. 
Hopefully the feedback below will help update the documentation for the next 
guy:

1) Documentation does a LOT of theory up front and never takes a simple 
non-bean case as an isolated case of making this architecture work. This makes 
it impossible to know what all the pieces are and whether or not the process is 
working. There needs to be a clear, clean - 
   a) here is a class
   b) here is a jboss-aop.xml
   c) here is what you need to put in a build script if you on JDK1.4 or 1.5, 
etc.
   d) here is a simple TreeCacheAop instantiation with a simple treecache.xml 
that uses local replication
   e) here is what you can do to troubleshoot if it doesn't work (for example I 
know mine is likely related to JbossAOP since its saying that non of my methods 
match the signature, but damned if I know what to do about it)

The docs commit the ultimate sin - no simple HelloWorld level example so you 
can make sure everything compiles and runs. Without this, you just have a level 
of complexity from multiple vendor products interacting. The regular TreeCache 
docs are fine. Page 3 (or section 4) puts up a simple case and explains what's 
going on in isolation. I can take that code, compile it, and run it and see 
what's going on.


2) Don't pay simple lip service to annotation, precompilation configuration. 
You'll have to learn JBossAOP and configuration to even begin to understand 
TreeCacheAOP and how it operates when a simple explanation of how the ant tasks 
work is sufficient.

3) If I need a special classpath put that in the configuration section as one 
ot the steps. 

4) The whole documentation should evolve as a series of lessons, not a 
description of features in action. One shouldn't be at the back of the 
documentation before you actually get something that at least resembles a 
lesson/setup

5) If you're going to use magic locations (META-INF/jboss-aop.xml) then you 
need to log when you didn't find stuff in those locations or if you did. My 
first pass through the docs say "put the jboss-aop.xml file in this location". 
Cool, okay its not working. Is it finding the file? Is it not liking the file? 
Is it not liking my OSX environment? What's wrong with it - what should I 
expect to see if its actually working.

I'd love to write a How-To guide to getting this thing set-up. I've done it 
before for JBoss (I wrote chunks of Jboss docs years ago) and would do it 
again, but unless you actually understand the system and how it works - the 
documentation here is just not useful unless you're modifying the toy examples 
that the documentation and source tree provide. If I have to go digging through 
unit tests to figure out how stuff works - something is wrong.

After going through TreeCacheAop for a few days I have a clear understanding of 
how it works, and actually understand how I could implement it myself if needed 
by breaking up the object hierarchy into treepaths in a TreeCache and 
intercepting the get/set field access and mapping it to methods which update 
the cache. Its actually a very cool approach - if you can get it to work.

View the original post : 
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3885523#3885523

Reply to the post : 
http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=3885523


-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
JBoss-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jboss-user

Reply via email to