james-server-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Noel J. Bergman" <n...@devtech.com>
Subject RE: ID-mail
Date Fri, 07 Mar 2003 04:45:34 GMT
Serge,

I know that Sun is aware of the issue.  It is actually an FAQ:

  http://java.sun.com/products/javamail/FAQ.html#msgid

and I know that you raised the issue in the past.  We have had this issue
come up more than once in even the past month on the James list, and I just
want Sun to be aware of the fact that this is a continuing issue.

As I said, we can implement subclassing, as discussed in the FAQ and our own
mailing list, but Sun might want to reconsider the subject.  The comment I
found in the archives from Bill Shannon was: "it guards against [accidental
usage]", and I suggest that there ought to be a way within the standard code
body to control the behavior.

One reason why I think it should be addressed in the base class rather than
left to subclasses is that service providers may provide their own
MimeMessage subclasses, and they won't necessarily provide the desired
behavior.  When updating a message received from Folder.getMessage(), I
don't know what updateHeaders semantic is implemented by that service
provider's MimeMessage subclass.

With respect to the client vs server usage argument, I could have client
code that scans messages in the background for spam or virii, and adds a
header.  I expect that at some point to see a Mailet pipeline in a mail
client.  The FetchPOP and newer FetchMail service for James basically
demonstrate that people are already using James almost as a server-side
client, so why not embed a Mailet pipeline directly in a client?

	--- Noel


---------------------------------------------------------------------
To unsubscribe, e-mail: james-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: james-dev-help@jakarta.apache.org


Mime
View raw message