Re: Sent message still in drafts (lisp error: "Selecting deleted buffer")

[Felix Natter]

Katsumi Yamaoka <yamaoka@jpl.org> writes:

hi Katsumi,

> Felix Natter <fnatter@gmx.net> wrote:
>> Katsumi Yamaoka <yamaoka@jpl.org> writes:
> [...]
>>> I found some other problems on expiring sent drafts to a group.
>>> The principal one is that a draft is not like a normal article:
>>>
>>> ・There is a separator between the header and the body.
>>>   - You can't read its body in an expiry group.
>>>   - Xref herder, etc. are put at the end of the body.
>>> ・Non-ASCII text has not been encoded properly.
>>> ・Attachments do not necessarily exist when re-using it.
>
>> So the problem is that when I send a draft, its content is subject
>> to expiry, and this does not work because a draft is not like
>> a normal article (and with the "G c" below we tell Gnus to delete
>> and not expire sent drafts)?
>
> Maybe that a draft is not a normal article is not a direct cause
> of your problem.  Though I haven't yet investigated why sometimes
> it fails with the message "Selecting deleted buffer".
>
>>> It will cost to improve it.
>
>> What do you mean by "costs to improve it"?
>
> Sorry for my funny English.  I meant that it will take time (IOW,
> cost me time ;-) to improve the Lisp code so as to transform
> a draft message, that is copied or moved to a normal group, to
> a normal article.

Ah ok ;-)

> [...]
>> Yes, I am doing that:
>> (setq nnmail-expiry-target "nnml:expired")
> [...]
>>> how about setting a group parameter so that only sent drafts
>>> are deleted?
> [...]
>> If that works, then I think it's a good idea to mention this problem
>> in the Gnus documentation?
>
> I have another idea.  That is to make the behavior of expiring of
> drafts be always `delete' no matter what nnmail-expiry-target is.
> It lets you and friends have no need to have such a special group
> parameter.  Ones who want to have a local copy of a sent message
> can use Gcc, that saves it as a normal article.  WDYT?

That's a really good idea, much better than just documenting a
workaround :-)

Thanks and Best Regards,
-- 
Felix Natter