Mm.. Is there a reason you wouldn't just have the imap attributes in the
emit tag to stay close to standard?
E.g
<emit source="imap_mailboxes" host="imap://user:<pass[at]imapserver.com>">
&_.name; &_.recent; &_.exists<br />
</emit>
<else>
<redirect to="/wrongcredentials.htm" />
</else>
I would also expect the emit tag to throw auth error messages without the
else..
David Hunter.
On 14/01/09 9:22 PM, "Marc Dirix" <<marc[at]electronics-design.nl>> wrote:
> I'm still working on a imap module for Roxen (without internal
> pre-defined layouting).
>
> Currently I'm deciding which way to ga with the roxen module, as basic
> Client in pike stands. Maybe you can give me your thoughts on these two
> questions.
>
> 1) To container or not.
> Basically I need some kind of mechanism to connect to the server if
> the connection is not in the cache. I'm thinking I can go 2 ways. Either
> use a container or not. See below 2 rxml examples:
> example 1
> <imap user="bla" password="foo" connectionid="">
> <emit source="imap_mailboxes">
> &_.name; &_.recent; &_.exists<br />
> </emit>
> </imap>
> <else>
> <redirect to="/wrongcredentials.htm" />
> </else>
>
> example 2
>
> <imap user="bla" password="foo" connectionid="theid" />
> <else>
> <redirect to="/wrongcredentials.htm" />
> </else>
> <emit source="imap_mailboxes" connectionid="theid" >
> &_.name; &_.recent; &_.exists<br />
> </emit>
>
> So the second example leaves more freedom, but I think the first is more
> controlled? What is your opinion?
>
> Second Question, I want to use some kind of connectionid to cache the
> connection. Should the connectionid either always be giving by the RXML
> page, or should the imap module enforce an id by means of the sessionid
> (internal), or am I missing some obvious direction here?
> If the module enforces the sessionid, it is obvious only one imap
> connection can be used per browser. Using more has no current business
> case for me, but if it is easy to do then why not?
>
> Best regards,
>
> Marc Dirix
>
|