And I also have tested other cases of EJB JNDI name on 1.6.0
Here is the result
1. Local EJB Client looks up @Local EJB
global jndi name is OK
mappedname is NG
2 remote EJB Client looks up @Remote EJB
global jndi name is NG
mappedname is OK
--
View this message in context:
What does mean NG?
Le 19 mai 2013 13:16, ZhongGuan guanzh.em...@gmail.com a écrit :
And I also have tested other cases of EJB JNDI name on 1.6.0
Here is the result
1. Local EJB Client looks up @Local EJB
global jndi name is OK
mappedname is NG
2 remote EJB Client looks up @Remote
No Good?
On Sun, May 19, 2013 at 7:18 AM, Romain Manni-Bucau
rmannibu...@gmail.comwrote:
What does mean NG?
Le 19 mai 2013 13:16, ZhongGuan guanzh.em...@gmail.com a écrit :
And I also have tested other cases of EJB JNDI name on 1.6.0
Here is the result
1. Local EJB Client looks up
Well, I think @Typed is another work around. It doesn't remove the object
from the archive simply says you can't reference it by anything but what's
in @Typed.
On Sun, May 19, 2013 at 6:19 AM, Mark Struberg strub...@yahoo.de wrote:
CDI-1.0 _has_ a concept of vetoing. Simply annotate Foo with
Yes, it's kind of different to ProcessAnnotatedType#veto(), but it has the same
effect - your class and your producer don't clash anymore ;)
And it comes without any expenses straight out of CDI-1.0
LieGrue,
strub
- Original Message -
From: John D. Ament john.d.am...@gmail.com
To:
Hehe, ok (i use NOk ;)
Btw mapped names are not portable IIRC so it is not an issue no?
Le 19 mai 2013 14:12, John D. Ament john.d.am...@gmail.com a écrit :
No Good?
On Sun, May 19, 2013 at 7:18 AM, Romain Manni-Bucau
rmannibu...@gmail.comwrote:
What does mean NG?
Le 19 mai 2013 13:16,
According to
http://docs.oracle.com/javaee/6/api/javax/ejb/EJB.html
lookup should be the portable. MappedName is the product specific.
On Sun, May 19, 2013 at 8:59 AM, Romain Manni-Bucau
rmannibu...@gmail.comwrote:
Hehe, ok (i use NOk ;)
Btw mapped names are not portable IIRC so it is not
Thank you guys for the excellent info!
Now that i know what to search for, theres actually some discussion going on
over this subject in internet.
Although it works for me, i have the feeling that both solutions (@Typed and
@Veto) are rather workarounds (i saw on stacktrace one more suggestion
by default your bean will be a managed bean so injectable, just add
anything to prevent it (constructor without @Inject for instance) and
you'll remove the conflict.
veto = remove it from cdi (= handled manually)
Typed() = keep it in cdi but for Object type only (as all beans)
@Any ~= i don't
+1 OP and responses!
On Sun, May 19, 2013 at 1:24 PM, Romain Manni-Bucau
rmannibu...@gmail.comwrote:
by default your bean will be a managed bean so injectable, just add
anything to prevent it (constructor without @Inject for instance) and
you'll remove the conflict.
veto = remove it from
There are a few ways to solve this riddle, but I personally prefer the ones who
are readable.
Some containers fail if you don't have a default ct - not a perfect solution.
I also don't really like @Alternative, because it is really confusing. A later
maintainer of the code will not easily see
11 matches
Mail list logo
