To me it looks like d) is better than c) (and e) ). I think headers
explicitly express the expected behavior and that should not be
altered by the pre-existing conf. imho the pre-existing conf is there
only for the case where explicit info is not present in the headers,
constituting a default behaviour, but not a base to be added to.
my $0.02,
Hadrian
On Aug 11, 2008, at 7:54 AM, Claus Ibsen (JIRA) wrote:
[ https://issues.apache.org/activemq/browse/CAMEL-777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=44898
#action_44898 ]
Claus Ibsen commented on CAMEL-777:
-----------------------------------
The short answer is that it will break existing behaviors in Camel
1.4 or older.
I also though that if you have any recipient informations in the
headers what should Camel do?
a) only use headers
b) only use pre configued (no - then dynamic recipients is not
possible)
c) headers append to existing pre configured
d) headers override existing pre configured
e) pre configured override headers (reverse as option e)
What we have now is option (e). What you want is option (d). But
what about option (c) isn't the the most correct and least surprised
one to have?
I would like to change the behavior and if its not 100% backwards
compatible then its okay anyway - we will write about it in the
release notes.
Any thoughts?
Allow '@' in usernames.
-----------------------
Key: CAMEL-777
URL: https://issues.apache.org/activemq/browse/CAMEL-777
Project: Apache Camel
Issue Type: Bug
Components: camel-mail
Affects Versions: 1.4.0
Reporter: Wojciech Durczynski
Assignee: Claus Ibsen
Fix For: 1.5.0
Many servers use 'username'@'servername' as a login name. Camel-
mail doesn't allow this.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
