something I've always wanted...

something I've always wanted...

[Josh Huber]

Something I've wanted for a while now, which I think will be very hard
to implement, is some sort of visual indication that a group with new
messages has a post either
(a) authored by me
(b) is a reply to a message from me

Obviously, this may be harder (or slower?) with news, but I think it's
possible with mail.

I have messages which meet that criteria scored up automatically, so
after I enter the group I can see the messages immediately, but I
subscribe to SO many groups/lists that I forget what I've posted
messages in. (oops ;)

Any insights, or is this too hard?

ttyl,
-- 
Josh Huber

Re: something I've always wanted...

[Karl Kleinpaste]

Given what information is acquired when, yes, it's very hard.  If you
want a *Group* buffer indication, the problem is that Gnus has no idea
of group content when creating the *Group* buffer.  All Gnus knows is
the minimalist data point, "there are articles here numbered between X
and Y."

For Gnus to know more than that, you would require Gnus to enter each
group and effectively do *Summary* generation on every single group,
without displaying the summaries, but using the information learned to
modify the *Group* display.

What you want is possible, but it's bound to be alarmingly expensive
and slow, especially when (as you said) you are subscribed to oodles
of forums.

Re: something I've always wanted...

[Matthieu Moy]

Karl Kleinpaste <karl@charcoal.com> writes:

>  All Gnus knows is
> the minimalist data point, "there are articles here numbered between X
> and Y."

I think it knows a bit more : I know the number of flagged and dormant
articles from the *Group* buffer already.

-- 
Matthieu

Re: something I've always wanted...

[Derrell.Lipman]

Matthieu Moy <Matthieu.Moy@imag.fr> writes:

> Karl Kleinpaste <karl@charcoal.com> writes:
>
>>  All Gnus knows is
>> the minimalist data point, "there are articles here numbered between X
>> and Y."
>
> I think it knows a bit more : I know the number of flagged and dormant
> articles from the *Group* buffer already.

An interesting possibility would be to have a "mark" which can be set by a
user-provided function, as the articles are being split.  The group buffer
could then display the number of articles that have that mark set...

Derrell

Re: something I've always wanted...

[Karl Kleinpaste]

Matthieu Moy <Matthieu.Moy@imag.fr> writes:
> I think it knows a bit more : I know the number of flagged and dormant
> articles from the *Group* buffer already.

True, but useless in context.  There is no information about article
content, especially authors and parent/child followup relationships.
That sort of information is gained by Gnus during group entry, in
preparation for *Summary* generation.

Re: something I've always wanted...

[Josh Huber]

Karl Kleinpaste <karl@charcoal.com> writes:

> True, but useless in context.  There is no information about article
> content, especially authors and parent/child followup relationships.
> That sort of information is gained by Gnus during group entry, in
> preparation for *Summary* generation.

Well, maybe something could be done while splitting? (as has been
mentioned)

What mechanism does the "followup" scoring use to score an article up?

I could look for a particular string in my messageid, perhaps and use
gnus-eye-candy to show it perhaps?  This would of course be a hack,
but I'd like it ;)

ttyl,

-- 
Josh Huber