Loss of first paragraph on fill and reply

Loss of first paragraph on fill and reply

[Emanuel Berg]

I have two problems that I suspect are related. Quite
often, but not always, when I use
`gnus-article-fill-cited-article' or
`gnus-article-followup-with-original' it works *except
for the first paragraph*.

This letter (found on gnu.emacs.sources) is now filled,
but let's pretend it isn't (and not cited either), as
is the form in which I first see it.

> A site to download Emacs custom configuration file
> with codes and extensions contributed by Emacs user
> around the world.

> http://sachin.pythonanywhere.com/

Filling won't fill the first paragraph, and followingup
will produce this result:

isachin <...@gmail.com> writes:

> http://sachin.pythonanywhere.com/
>
> --
> Sachin
>

I.e., the first paragraph is again ignored.

Ideas?

-- 
Emanuel Berg, programmer-for-rent. CV, projects, etc at uXu
underground experts united:  http://user.it.uu.se/~embe8573

Re: Loss of first paragraph on fill and reply

[Dmitrii Kashin]

[-- Attachment #1.1: Type: text/plain, Size: 349 bytes --]

Emanuel Berg <embe8573@student.uu.se> writes:

> I have two problems that I suspect are related. Quite
> often, but not always, when I use
> `gnus-article-fill-cited-article' or
> `gnus-article-followup-with-original' it works *except
> for the first paragraph*.

Cool. Really.

Does your problem exist if you start gnus with a clean configuration?

[-- Attachment #1.2: Type: application/pgp-signature, Size: 835 bytes --]

[-- Attachment #2: Type: text/plain, Size: 162 bytes --]

_______________________________________________
info-gnus-english mailing list
info-gnus-english@gnu.org
https://lists.gnu.org/mailman/listinfo/info-gnus-english

Re: Loss of first paragraph on fill and reply

[Emanuel Berg]

Dmitrii Kashin <freehck@freehck.ru> writes:

>> I have two problems that I suspect are related. Quite
>> often, but not always, when I use
>> `gnus-article-fill-cited-article' or
>> `gnus-article-followup-with-original' it works
>> *except for the first paragraph*.
>
> Cool. Really.
>
> Does your problem exist if you start gnus with a clean
> configuration?

How do I test that? If I run Emacs with -Q, and then run
Gnus, I'm not getting connected to the server.

If you have a good eye for what configuration might
screw it up (headers perhaps?), have a look at

http://user.it.uu.se/~embe8573/emacs.html

and especially the Gnus part.

-- 
Emanuel Berg, programmer-for-rent. CV, projects, etc at uXu
underground experts united:  http://user.it.uu.se/~embe8573

Re: Loss of first paragraph on fill and reply

[Dmitrii Kashin]

[-- Attachment #1.1: Type: text/plain, Size: 990 bytes --]

Emanuel Berg <embe8573@student.uu.se> writes:

> Dmitrii Kashin <freehck@freehck.ru> writes:
>
>>> I have two problems that I suspect are related. Quite
>>> often, but not always, when I use
>>> `gnus-article-fill-cited-article' or
>>> `gnus-article-followup-with-original' it works
>>> *except for the first paragraph*.
>>
>> Cool. Really.
>>
>> Does your problem exist if you start gnus with a clean
>> configuration?
>
> How do I test that? If I run Emacs with -Q, and then run
> Gnus, I'm not getting connected to the server.

Easily. Move your gnus-rc file (~/.gnus) somewhere, start emacs with -q
(not -Q) flag. Run `gnus` function. Then use B key [1] to browse any
nntp server you want. And then try to answer somebody.

[1] (info "(gnus)Browse Foreign Server")

> If you have a good eye for what configuration might
> screw it up (headers perhaps?), have a look at

Maybe later. Let's firstly detect the root of problems.
I just suggested it was some very smart hook you had added.

[-- Attachment #1.2: Type: application/pgp-signature, Size: 835 bytes --]

[-- Attachment #2: Type: text/plain, Size: 162 bytes --]

_______________________________________________
info-gnus-english mailing list
info-gnus-english@gnu.org
https://lists.gnu.org/mailman/listinfo/info-gnus-english

Re: Loss of first paragraph on fill and reply

[Emanuel Berg]

Dmitrii Kashin <freehck@freehck.ru> writes:

> Easily. Move your gnus-rc file (~/.gnus) somewhere,
> start emacs with -q (not -Q) flag. Run `gnus`
> function. Then use B key [1] to browse any nntp server
> you want. And then try to answer somebody.

You are absolutely right, neither of these issues are
there when I start Emacs and Gnus the way you suggest,
i.e., without my configuration/extension.

-- 
Emanuel Berg, programmer-for-rent. CV, projects, etc at uXu
underground experts united:  http://user.it.uu.se/~embe8573

Re: Loss of first paragraph on fill and reply

[Dmitrii Kashin]

[-- Attachment #1.1: Type: text/plain, Size: 927 bytes --]

Emanuel Berg <embe8573@student.uu.se> writes:

> Dmitrii Kashin <freehck@freehck.ru> writes:
>
>> Easily. Move your gnus-rc file (~/.gnus) somewhere,
>> start emacs with -q (not -Q) flag. Run `gnus`
>> function. Then use B key [1] to browse any nntp server
>> you want. And then try to answer somebody.
>
> You are absolutely right, neither of these issues are
> there when I start Emacs and Gnus the way you suggest,
> i.e., without my configuration/extension.

Well, this is the good news. Thus, a problem exists somewhere inside
your configuration.

You have a lot of configuration files. Try now to load some of them in
order to detect which one does contain problem code.

I suppose it could be .emacs-message or .emacs-gnus.

PS: Of course, the most sane way to detect problem is to watch some kind
of backtrace when you are in the buffer for message composing. But to be
fair I'm still unable to debug such a right way.

[-- Attachment #1.2: Type: application/pgp-signature, Size: 835 bytes --]

[-- Attachment #2: Type: text/plain, Size: 162 bytes --]

_______________________________________________
info-gnus-english mailing list
info-gnus-english@gnu.org
https://lists.gnu.org/mailman/listinfo/info-gnus-english

Re: Loss of first paragraph on fill and reply

[Emanuel Berg]

Dmitrii Kashin <freehck@freehck.ru> writes:

>> You are absolutely right, neither of these issues
>> are there when I start Emacs and Gnus the way you
>> suggest, i.e., without my configuration/extension.
>
> Well, this is the good news. Thus, a problem exists
> somewhere inside your configuration.

Wow - for all my stupidity in this thread, I have to
say my intuition is amazing! Yes, it was the
headers. Specifically, this line:

;; don't show any headers
(setq gnus-visible-headers "^XXX") 

So the, ehm, *followup* question is: what is the
correct way to hide all headers, naturally without
breaking the fill and "followup with original"
functions? (And possibly other functions as well.)

But this is so much better. I care how it looks, but I
care *more* it works. Spasibo bolshoe!

-- 
Emanuel Berg, programmer-for-rent. CV, projects, etc at uXu
underground experts united:  http://user.it.uu.se/~embe8573

Re: Loss of first paragraph on fill and reply

[Emanuel Berg]

Emanuel Berg <embe8573@student.uu.se> writes:

> Wow - for all my stupidity in this thread, I have to
> say my intuition is amazing! Yes, it was the
> headers. Specifically, this line:
>
> ;; don't show any headers
> (setq gnus-visible-headers "^XXX") 
>
> So the, ehm, *followup* question is: what is the
> correct way to hide all headers, naturally without
> breaking the fill and "followup with original"
> functions? (And possibly other functions as well.)

This attempt has the same problem:

;; (setq gnus-visible-headers nil)
;; (setq gnus-ignored-headers '("^.*:"))

-- 
Emanuel Berg, programmer-for-rent. CV, projects, etc at uXu
underground experts united:  http://user.it.uu.se/~embe8573

Re: Loss of first paragraph on fill and reply

[Dmitrii Kashin]

[-- Attachment #1.1: Type: text/plain, Size: 893 bytes --]

Emanuel Berg <embe8573@student.uu.se> writes:

> Dmitrii Kashin <freehck@freehck.ru> writes:
>
>>> You are absolutely right, neither of these issues
>>> are there when I start Emacs and Gnus the way you
>>> suggest, i.e., without my configuration/extension.
>>
>> Well, this is the good news. Thus, a problem exists
>> somewhere inside your configuration.
>
> Wow - for all my stupidity in this thread, I have to
> say my intuition is amazing! Yes, it was the
> headers.

Wow. My intuition works another way. I do not see a correlation between
headers and first paragraph of citation. I suppose it is abnormal and I
would speak to somebody in gnus developer list about it. Here the
contacts are.

Newsgroup: gmane.emacs.gnus.general
List: ding@gnus.org

BTW, I noticed a strange thing. Your "References:" field does not
contain my message id, when you reply to me. It seems to be another bug.

[-- Attachment #1.2: Type: application/pgp-signature, Size: 835 bytes --]

[-- Attachment #2: Type: text/plain, Size: 162 bytes --]

_______________________________________________
info-gnus-english mailing list
info-gnus-english@gnu.org
https://lists.gnu.org/mailman/listinfo/info-gnus-english

Re: Loss of first paragraph on fill and reply

[Emanuel Berg]

Dmitrii Kashin <freehck@freehck.ru> writes:

> Wow. My intuition works another way. I do not see a
> correlation between headers and first paragraph of
> citation.

Me neither, but remember the problem was also that the
first paragraph couldn't be *filled*. So my thinking
was: if there is no headers, the fill (and reply)
thinks the first paragraph *is* a header, so it doesn't
do anything on it. I don't know if this really is the
case, but it seems likely.

> I suppose it is abnormal and I would speak to
> somebody in gnus developer list about it. Here the
> contacts are.
>
> Newsgroup: gmane.emacs.gnus.general List:
> ding@gnus.org

Yes, but I know those guys read this group as
well. Don't you think they will examine it anyway? Or
do you insist I should write a mail? If so, I will do
it, of course.

> BTW, I noticed a strange thing. Your "References:"
> field does not contain my message id, when you reply
> to me. It seems to be another bug.

Yeah? Is that so for every message? Does that somehow
disturb your reading of my messages (sorting or the
like)? Do you know how to fix it?

-- 
Emanuel Berg, programmer-for-rent. CV, projects, etc at uXu
underground experts united:  http://user.it.uu.se/~embe8573

Re: Loss of first paragraph on fill and reply

[Dmitrii Kashin]

[-- Attachment #1.1: Type: text/plain, Size: 1229 bytes --]

Emanuel Berg <embe8573@student.uu.se> writes:

> Dmitrii Kashin <freehck@freehck.ru> writes:
>
>> I suppose it is abnormal and I would speak to somebody in gnus
>> developer list about it. Here the contacts are.
>>
>> Newsgroup: gmane.emacs.gnus.general
>> List: ding@gnus.org
>
> Yes, but I know those guys read this group as well. Don't you think
> they will examine it anyway? Or do you insist I should write a mail?
> If so, I will do it, of course.

As you want. I would write. 

>> BTW, I noticed a strange thing. Your "References:" field does not
>> contain my message id, when you reply to me. It seems to be another
>> bug.
>
> Yeah? Is that so for every message?

As I see.

> Does that somehow disturb your reading of my messages (sorting or the
> like)?

In such an easy case with a linear thread this is not a problem for
thread to be constructed. But if more leaves existed between your first
letter and your next answer, the answer could be connected to wrong
leaf. And in any case it breaks scoring, because gnus cannot know that
your message is answer to me.

> Do you know how to fix it?

To be fair, I certainly can load your config and play with it... But I
feel lazy. =)

[-- Attachment #1.2: Type: application/pgp-signature, Size: 835 bytes --]

[-- Attachment #2: Type: text/plain, Size: 162 bytes --]

_______________________________________________
info-gnus-english mailing list
info-gnus-english@gnu.org
https://lists.gnu.org/mailman/listinfo/info-gnus-english

Re: Loss of first paragraph on fill and reply

[Emanuel Berg]

Dmitrii Kashin <freehck@freehck.ru> writes:

> In such an easy case with a linear thread this is not
> a problem for thread to be constructed. But if more
> leaves existed between your first letter and your
> next answer, the answer could be connected to wrong
> leaf. And in any case it breaks scoring, because gnus
> cannot know that your message is answer to me.

You are right, it should be fixed. Well, if anyone
reading this has a clue why this happened feel free to
speak :)

-- 
Emanuel Berg, programmer-for-rent. CV, projects, etc at uXu
underground experts united:  http://user.it.uu.se/~embe8573