[mew-int 00667] Re: Questions not answered by FAQ

[Tak Ota]

Tue, 25 Dec 2001 23:14:35 +0900 (JST): Kazu Yamamoto <kazu at example.com> wrote:

> (1) If you forward a message, anotation information will be disclosed.
>     This is avoidable by resent Mew which can delete fields accoding
>     to the configurable variable. However, if you also use an old
>     version Mew or another mailer, this problem will appear.

I was thinking of adding annotation for unread messages only.  The
annotation is removed when the message is read.  Most likely a message
being forward is read beforehand.  Anyway this method is not generic
enough to record all unread/replied/forwarded status.

I sometimes attach messages as Message/Rfc822 type.  Should this also
take care of removing the annotation?

After all, is there any practical harm by having such annotation,
probably in X-Mew-xxx form, in the forwarded or attached messages?  I
understand that it does reveal some private information to recipients.
I imagine that introduction of such annotation mechanism will open up
a lot of new nice user convenience features.

> (2) If you annotate a message, the timestamp of its directory change.
>     So, the "update" feature becomes broken. Do we all UNIX users have
>     to migrate to the ".mew-touch" file?

Maybe they should.  Unintended user action such as file manipulation
without using Mew can introduce the same effect.  It'll also improve
the platform independence and code unification with minor price to
pay.

> Anyway, we have to consider this when we support IMAP because the
> protocol has this features. As you may know, Mew already implements
> NNTP and the next target is IMAP. This feature should be discussed in
> 2002 1Q.

I'm looking forward to that, so are many others I suppose.  Thanks for
the continuous improvement.

-Tak