gnus-summary-highlight should use overlays?

gnus-summary-highlight should use overlays?

[Alan Shutko]

Because of the gnus-agent-downloaded-article-face overriding other
summary highlights, I was wondering if perhaps gnus-summary-highlight
should use overlays (on Emacs, at least) instead of text properties
for the face.  That way, we could have one overlay for each entry in
gnus-summary-highlight which matches, and the resultant face would be
the composition of the matching faces.  I presume XEmacs could do the
same thing, somehow.

How does that sound?

-- 
Alan Shutko <ats@acm.org> - In a variety of flavors!
In our world, we also have a Hawaii! - Dr. F

Re: gnus-summary-highlight should use overlays?

[KGreiner]

Alan Shutko <ats@acm.org> writes:

> Because of the gnus-agent-downloaded-article-face overriding other
> summary highlights, I was wondering if perhaps gnus-summary-highlight
> should use overlays (on Emacs, at least) instead of text properties
> for the face.  That way, we could have one overlay for each entry in
> gnus-summary-highlight which matches, and the resultant face would be
> the composition of the matching faces.  I presume XEmacs could do the
> same thing, somehow.
>
> How does that sound?

Just to let you know, I've already rewritten this code to address
feedback from other users.  The code is ready for check-in, just
waiting for some feedback on a couple of private questions.

The changes and the concerns that they address are:
  1) The current gnus-agent-downloaded-article-face currently displays
     when an article was fetched.  The new agent article face will be
     applied when an article has NOT been fetched.  The idea being
     that the fetched status should be fairly low priority.

  2) The gnus-agent-downloaded-article-face stepped on the
     high/normal/low score faces.  There are actually three new agent
     article faces.  They provide the same high/normal/low score
     indicator as usual with a background color indicating that the
     articles have not been fetched.

Will this help with the overriding problem that you are seeing?

Kevin

Re: gnus-summary-highlight should use overlays?

[Alan Shutko]

KGreiner@xpediantsolutions.com writes:

> Just to let you know, I've already rewritten this code to address
> feedback from other users.
[...]
> Will this help with the overriding problem that you are seeing?

It would solve the immediate problem, but it would be nice if you
didn't have to develop workarounds like that.  I think overlays would
make it easier to show multi-state things.  (So we don't have to keep
combining different ideas into single faces.)

-- 
Alan Shutko <ats@acm.org> - In a variety of flavors!
But this one act of consecration is what I ask of you - NIN

Re: gnus-summary-highlight should use overlays?

[kgreiner]

Alan Shutko <ats@acm.org> writes:

> KGreiner@xpediantsolutions.com writes:
>
>> Just to let you know, I've already rewritten this code to address
>> feedback from other users.
> [...]
>> Will this help with the overriding problem that you are seeing?
>
> It would solve the immediate problem, but it would be nice if you
> didn't have to develop workarounds like that.  I think overlays would
> make it easier to show multi-state things.  (So we don't have to keep
> combining different ideas into single faces.)
>
> -- 
> Alan Shutko <ats@acm.org> - In a variety of flavors!
> But this one act of consecration is what I ask of you - NIN

Agreed.  I believe that I was considering the same idea from the
opposite end; the gnus-summary-highlight definition.  I've been toying
with the idea that gnus-summary-highlight should specify a set of
conditions and font attributes.  Each time a condition evals to true,
the indicated attribute is added to the line's face. That way, you
don't have to redefine a half dozen faces to get a consistent scheme.
To be honest, I haven't gotten past the thinking stage.  I'll take a
look at overlays, they may make the idea straight-forward to
implement.

Kevin

Re: gnus-summary-highlight should use overlays?

[Lars Magne Ingebrigtsen]

kgreiner@xpediantsolutions.com writes:

> Agreed.  I believe that I was considering the same idea from the
> opposite end; the gnus-summary-highlight definition.  I've been toying
> with the idea that gnus-summary-highlight should specify a set of
> conditions and font attributes.  Each time a condition evals to true,
> the indicated attribute is added to the line's face. That way, you
> don't have to redefine a half dozen faces to get a consistent scheme.

That would make things a lot easier.  If both Emacs and XEmacs now
supports doing something like that, that would be very nice.
Historically, doing face stuff in Emacs has been a pain.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen