Re: nnimap error reporting

Re: nnimap error reporting

[Amos Gouaux]

>>>>> On Mon, 01 Jul 2002 16:03:01 +0200,
>>>>> Jan Rychter <jan@rychter.com> (jr) writes:

jr> nnimap produces strange errors when it can't connect to the server: I
jr> get a a message "datum not string or error symbol".

[...]

jr> I guess it should fail more gracefully?

That reminds me of something I've always been curious about: does
nnimap report server ALERT messages?  Just wondering.


-- 
Amos

Re: nnimap error reporting

[Simon Josefsson]

Amos Gouaux <amos+lists.ding@utdallas.edu> writes:

>>>>>> On Mon, 01 Jul 2002 16:03:01 +0200,
>>>>>> Jan Rychter <jan@rychter.com> (jr) writes:
>
> jr> nnimap produces strange errors when it can't connect to the server: I
> jr> get a a message "datum not string or error symbol".
>
> [...]
>
> jr> I guess it should fail more gracefully?
>
> That reminds me of something I've always been curious about: does
> nnimap report server ALERT messages?  Just wondering.

Just with `message', so it is probably easy to miss.

Re: nnimap error reporting

[Simon Josefsson]

Jan Rychter <jan@rychter.com> writes:

> nnimap produces strange errors when it can't connect to the server: I
> get a a message "datum not string or error symbol".

Exactly when does it happen?

> Backtrace:
>
> Signaling: (invalid-argument "datum not string or error symbol" nil)
>   signal(invalid-argument ("datum not string or error symbol" nil))
>   cerror(invalid-argument "datum not string or error symbol" nil)
>   apply(cerror invalid-argument ("datum not string or error symbol" nil))
>   error(invalid-argument "datum not string or error symbol" nil)
>   cerror(nil)
>   apply(cerror nil nil)
>   error(nil)
>   mail-source-fetch-imap((imap :server "server.name" :stream ssl :port 993 :user "jwr") #<compiled-function (file orig-file) "...(17)" [orig-file file nnmail-split-incoming nnml-save-mail nil nnmail-get-split-group (imap :server "server.name" :stream ssl :port 993 :user "jwr") nnml-active-number] 7>)

Unfortunately this isn't very useful, perhaps (setq debug-on-signal t)
produces a better backtrace (be prepared to press c sometimes).