Folding long header lines

Folding long header lines

[Lars Magne Ingebrigtsen]

To be compliant, Gnus folds long header lines (like User-Agent in this
message).  However, Gnus doesn't use RFC2047 coding on headers that
don't, er, need RFC2047 coding (i. e., only contains US-ASCII
characters).  So when displaying these, Gnus doesn't un-fold the
headers again, leading to what you see here -- a two-line User-Agent
header.

Either:

1) This is OK, since it's perfectly valid and stuff
2) It's annoying that Gnus folds lines that it has no
   way of putting back together again, so it should just stop
3) It should do extra RFC2047 coding around the folding just
   so that the recipient can put the header back together again
4) Gnus should just remove all folding when displaying headers

None of these options seem particularly attractive to me, so I must be
missing some option or other.

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

Re: Folding long header lines

[Jesper Harder]

larsi@gnus.org (Lars Magne Ingebrigtsen) writes:

> To be compliant, Gnus folds long header lines [..]
> Either:
>
> 1) This is OK, since it's perfectly valid and stuff
> 2) It's annoying that Gnus folds lines that it has no
>    way of putting back together again, so it should just stop
> 3) It should do extra RFC2047 coding around the folding just
>    so that the recipient can put the header back together again
> 4) Gnus should just remove all folding when displaying headers
>
> None of these options seem particularly attractive to me, so I must be
> missing some option or other.

I think 4) is best, at least that's what RFC 2822 says (section 2.2.3).

Re: Folding long header lines

[Lars Magne Ingebrigtsen]

Jesper Harder <harder@ifa.au.dk> writes:

>> 4) Gnus should just remove all folding when displaying headers

[...]

> I think 4) is best, at least that's what RFC 2822 says (section 2.2.3).

It doesn't seem to recommend unfolding for display purposes; only for
parsing and stuff.

Anyway, I've written a treatment function that unfolds folded headers,
and defaulted the variable to `head'.  If people complain, we can
change the default.

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

Re: Folding long header lines

[Russ Allbery]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> To be compliant, Gnus folds long header lines (like User-Agent in this
> message).  However, Gnus doesn't use RFC2047 coding on headers that
> don't, er, need RFC2047 coding (i. e., only contains US-ASCII
> characters).  So when displaying these, Gnus doesn't un-fold the headers
> again, leading to what you see here -- a two-line User-Agent header.

> Either:

> 1) This is OK, since it's perfectly valid and stuff
> 2) It's annoying that Gnus folds lines that it has no
>    way of putting back together again, so it should just stop
> 3) It should do extra RFC2047 coding around the folding just
>    so that the recipient can put the header back together again
> 4) Gnus should just remove all folding when displaying headers

> None of these options seem particularly attractive to me, so I must be
> missing some option or other.

I'm not entirely sure that I understand this message, but in case the
following information is helpful:  Folding is not, per the mail standards,
semantically part of the header.  In other words, RFC 2822 says that:

Subject: This is
 folded

is equivalent to:

Subject: This is folded

Which seems to argue in favor of either 1) or 4).  (Personally, I'd rather
see headers folded for display than see headers wrap because they're over
80 columns; folded headers are easier to read.)

I definitely don't like the sound of 3); the less RFC 2047 has to be used,
the better.

-- 
Russ Allbery (rra@stanford.edu)             <http://www.eyrie.org/~eagle/>

Re: Folding long header lines

[Lars Magne Ingebrigtsen]

Russ Allbery <rra@stanford.edu> writes:

> Which seems to argue in favor of either 1) or 4).  (Personally, I'd rather
> see headers folded for display than see headers wrap because they're over
> 80 columns; folded headers are easier to read.)

Yup.  I've made the unfolding function only unfold if the result is
less wide than the window you're using.

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