>>>>> Unified EL Spec says:
>>>>>>
>>>>> If A is String: return A
>>>>> Otherwise, if A is null: return ""
MW> But...
MW> isn't that silly ?
JH> I believe this is a carry over from JSTL.
RL> Most likely. Getting it logged on the spec tracker. That way Ed/Roger
RL> can track it and bring the issue up with the EL spec folks.
Matthias, Jacob is correct. Returning "" in this case is a usability
decision. A similar one is how we return a descriptive string when you
try to look up a key from a ResourceBundle via the EL, and the key is
not present. The underlying technology, java.util.ResourceBundle,
throws a MissingResourceException, but we decided showing "key {0} not
found" to the user was better.
This is the sort of thing for which we can now specify software lifecycle
dependent behavior, thanks to [313-DevLifecycle] [1].
In that regard, can people on this list please share some ideas for how
best to make use of [313-DevLifecycle]?
Sincerely,
Ed
[1]
https://javaserverfaces-spec-public.dev.java.net/issues/show_bug.cgi?id=313
--
| 110 Business Days until JavaOne 2008
| 215 Business Days until Proposed Final Draft for JSF 2.0
| ed.burns_at_sun.com | office: 408 884 9519 OR x31640
| homepage: | http://purl.oclc.org/NET/edburns/
| aim: edburns0sunw | iim: ed.burns_at_sun.com