0x0a converted to 0x0d when saving attachments

0x0a converted to 0x0d when saving attachments

[Mike Fabian]

Pterodactyl Gnus v0.97 still converts  all 0x0a into 0x0d when  saving
binary attachments (eg. *.tar.gz files).

Shenghou ZHU did send a patch which fixed this for pgnus 0.95.

Has this patch  been forgotten and not been  applied to pgnus 0.96 and
pgnus 0.97?

Mike

-- 
Mike Fabian   <mike.fabian@gmx.de>   
In der Neckarhelle 81, D-69118 Heidelberg-Ziegelhausen
Telephone: +49(0)6221/809222

Re: 0x0a converted to 0x0d when saving attachments

[Shenghuo ZHU]

>>>>> "Mike" == Mike Fabian <mike.fabian@gmx.de> writes:

Mike> Pterodactyl Gnus v0.97 still converts  all 0x0a into 0x0d when  saving
Mike> binary attachments (eg. *.tar.gz files).

Mike> Shenghou ZHU did send a patch which fixed this for pgnus 0.95.

Mike> Has this patch been forgotten and not been applied to pgnus 0.96
Mike> and pgnus 0.97?

[...]

The patch has been applied to 0.96.  (95-zsh6? Please help me check
the patch number.)

But there is still a bug.  I have not located where it is.  When I
attach a file with 9 '0x0d's, the Gcc copy got 9, but the receiver got
only 7.

-- 
Shenghuo ZHU

Re: 0x0a converted to 0x0d when saving attachments

[Mike Fabian]

Shenghuo ZHU <zsh@cs.rochester.edu> writes:

> >>>>> "Mike" == Mike Fabian <mike.fabian@gmx.de> writes:
> 
> Mike> Pterodactyl Gnus v0.97 still converts  all 0x0a into 0x0d when  saving
> Mike> binary attachments (eg. *.tar.gz files).
> 
> Mike> Shenghou ZHU did send a patch which fixed this for pgnus 0.95.
> 
> Mike> Has this patch been forgotten and not been applied to pgnus 0.96
> Mike> and pgnus 0.97?
> 
> [...]
> 
> The patch has been applied to 0.96.  (95-zsh6? Please help me check
> the patch number.)

Yes, your patch which worked well for 0.95 was 95-zsh6.diff.

> But there is still a bug.  I have not located where it is.  When I
> attach a file with 9 '0x0d's, the Gcc copy got 9, but the receiver got
> only 7.

I tried to send a 400kilobyte tar.gz file. The Gcc copy and the copy
which the receiver got were identical:

    mike@nozomi ~$ diff tcl8.1-info.tar.gz-gcc tcl8.1-info.tar.gz-receiver

but differed from the original. The only differences are that 0x0d
has been replaced by 0x0a:

    mike@nozomi ~$ cmp -l tcl8.1-info.tar.gz-gcc tcl8.1-info.tar.gz > cmp.out

    mike@nozomi ~$ head cmp.out
        66  12  15
       552  12  15
       694  12  15
       748  12  15
      1133  12  15
      1293  12  15
      1386  12  15
      1514  12  15
      2276  12  15
      2360  12  15

Mike

-- 
Mike Fabian   <mike.fabian@gmx.de>   
In der Neckarhelle 81, D-69118 Heidelberg-Ziegelhausen
Telephone: +49(0)6221/809222

Re: 0x0a converted to 0x0d when saving attachments

[Shenghuo ZHU]

>>>>> "Mike" == Mike Fabian <mike.fabian@gmx.de> writes:

[...]


Mike> I tried to send a 400kilobyte tar.gz file. The Gcc copy and the copy
Mike> which the receiver got were identical:

Mike>     mike@nozomi ~$ diff tcl8.1-info.tar.gz-gcc tcl8.1-info.tar.gz-receiver

Mike> but differed from the original. The only differences are that 0x0d
Mike> has been replaced by 0x0a:


[...]

I can not reproduce the bug in Gnu Emacs.  Please try 
        (setq mm-binary-coding-system 'binary) 
to see whether it still happens.

-- 
Shenghuo ZHU

Re: 0x0a converted to 0x0d when saving attachments

[Mike Fabian]

Shenghuo ZHU <zsh@cs.rochester.edu> writes:

> >>>>> "Mike" == Mike Fabian <mike.fabian@gmx.de> writes:
> 
> [...]
> 
> 
> Mike> I tried to send a 400kilobyte tar.gz file. The Gcc copy and
> Mike> the copy which the receiver got were identical but differed
> Mike> from the original. The only differences are that 0x0d has been
> Mike> replaced by 0x0a:
> 
> [...]
> 
> I can not reproduce the bug in Gnu Emacs.  Please try 
>         (setq mm-binary-coding-system 'binary) 
> to see whether it still happens.

This makes the problem disappear in XEmacs. Thank you very much!

The previous value was `no-conversion'.

What does that mean? What is the difference between `no-conversion'
and `binary'?

Mike

-- 
Mike Fabian   <mike.fabian@gmx.de>   
In der Neckarhelle 81, D-69118 Heidelberg-Ziegelhausen
Telephone: +49(0)6221/809222

Re: 0x0a converted to 0x0d when saving attachments

[Shenghuo ZHU]

>>>>> "Mike" == Mike Fabian <mike.fabian@gmx.de> writes:

Mike> This makes the problem disappear in XEmacs. Thank you very much!

Mike> The previous value was `no-conversion'.

Mike> What does that mean? What is the difference between
Mike> `no-conversion' and `binary'?

That's it. I guess the difference is the type of EOL. That's why 0x0a
is converted to 0x0d.

I've committed the change to the CVS several days ago.

-- 
Shenghuo ZHU

Re: 0x0a converted to 0x0d when saving attachments

[Stainless Steel Rat]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

* Mike Fabian <mike.fabian@gmx.de>  on Sat, 02 Oct 1999
| What does that mean? What is the difference between `no-conversion'
| and `binary'?

`binary' means that the buffer should be treated as "raw" data.
`no-conversion' does EOLN conversion, I think, contradicting the idea of
performing no conversion.  Both FSF Emacs and XEmacs are rock-stupid and
broken about this.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.0 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE392pBgl+vIlSVSNkRAmfUAJ0WCyAQ4s0qleCLPWWE5/5RBxiMYgCeP00o
NeyJiRY8baPDbziPMj+HiGU=
=6b8M
-----END PGP SIGNATURE-----

-- 
Rat <ratinox@peorth.gweep.net>    \ Happy Fun Ball may stick to certain types
Minion of Nathan - Nathan says Hi! \ of skin.
PGP Key: at a key server near you!  \ 

Re: 0x0a converted to 0x0d when saving attachments

[Jan Vroonhof]

Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

> `binary' means that the buffer should be treated as "raw" data.
> `no-conversion' does EOLN conversion, I think, contradicting the idea of
> performing no conversion.

no-conversion means no _characterset conversion_ (or at least in
XEmacs it does). However I agree raw-text is a much better name.

> Both FSF Emacs and XEmacs are rock-stupid and broken about this.

In fact in FSF Emacs no-conversion == binary. IIRC correctly in XEmacs
the meaning on no-conversion ping-ponged around in order to stay
compatible with the then unreleased "New Mule API". Now it has the
current meaning because that is what the 20.x XEmacs series had.

no-conversion should really be banished to obscurity where it belongs
to give way to the more meaningful terms 'binary and 'raw-text.
However the fact that 'no-conversion seems to be the preferred name
for binary on FSF Emacs and the fact that not all previous versions of
the Emacsen had all three doesn't help.

Jan

Re: 0x0a converted to 0x0d when saving attachments

[Stainless Steel Rat]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

* Jan Vroonhof <vroonhof@math.ethz.ch>  on Sun, 03 Oct 1999
| no-conversion means no _characterset conversion_ (or at least in
| XEmacs it does). However I agree raw-text is a much better name.

"raw" is a better name.  Data and text are not necessarilly the same thing, 
and frequently are not.

| In fact in FSF Emacs no-conversion == binary.

Hmm... I seem to recall that.  This is one of FSF Emacs 20's stupidities:
it assumes that if no coding system is set it *must* perform a conversion,
which is just plain wrong.  So I put no-conversion coding systems into
pop3 -- which should not be required since POP3 operates on RFC 822
messages, which forbids 8-bit data.

Of course, this caused XEmacs 20 to break because on "illegal" messages
becaue no-conversion means something totally different.  What a pain.

Since I changed the coding systems to `binary' nobody has complained about
it, so I guess it works.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.0 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE394mQgl+vIlSVSNkRAmitAKDBhnSA5VV8gRqgeHREsQZ0h8MEawCglFEy
5g8y0tlhtLaJ6A7PMp4cQyI=
=DNrb
-----END PGP SIGNATURE-----

-- 
Rat <ratinox@peorth.gweep.net>    \ Do not use Happy Fun Ball on concrete.
Minion of Nathan - Nathan says Hi! \ 
PGP Key: at a key server near you!  \ 

Re: 0x0a converted to 0x0d when saving attachments

[Jan Vroonhof]

Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

> Hmm... I seem to recall that.  This is one of FSF Emacs 20's stupidities:
> it assumes that if no coding system is set it *must* perform a conversion,
> which is just plain wrong.

I think you'll find that this is not limited to FSF Emacs. I think in
itself the idea that not specifying a coding system leads to
auto-detection is sound however the autodetection should be _much_
more conservative and for _non-files_ (or at least internet sockets)
the default should be 'binary.

> Of course, this caused XEmacs 20 to break because on "illegal" messages
> becaue no-conversion means something totally different.  What a pain.

The joys of two teams, at least one of them manifestly not trying to
be compatible[1], working to implement the same manifestly
undocumented or even undefined API.

Jan

Footnotes: 
[1]  Note the careful positioning of 'not' in this sentence.

Re: 0x0a converted to 0x0d when saving attachments

[Stainless Steel Rat]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

* Jan Vroonhof <vroonhof@math.ethz.ch>  on Sun, 03 Oct 1999
| Footnotes:
| [1]  Note the careful positioning of 'not' in this sentence.

I suppose a major problem is that the Japanese MULE team has a habit of
frequently changing their own API (if API it can be called), causing all
sorts of things in FSF Emacs to break (cf much of Erik Naggum's rantings).
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.0 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE3+NIsgl+vIlSVSNkRAlCcAKDrl7BQ5Hfb79sWaBLdnbzlG9KlcACfZCu+
Fsu54Z/VDIKCEdU78D3ocds=
=V3OH
-----END PGP SIGNATURE-----

-- 
Rat <ratinox@peorth.gweep.net>    \ Warning: pregnant women, the elderly, and
Minion of Nathan - Nathan says Hi! \ children under 10 should avoid prolonged
PGP Key: at a key server near you!  \ exposure to Happy Fun Ball.

Re: 0x0a converted to 0x0d when saving attachments

[François Pinard]

[-- Attachment #1: Type: text/plain, Size: 722 bytes --]

Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

> I suppose a major problem is that the Japanese MULE team has a habit of
> frequently changing their own API (if API it can be called), causing all
> sorts of things in FSF Emacs to break (cf much of Erik Naggum's rantings).

I'm very far from knowing all the details of the story.  I got that more
than once, Richard made changes on his side, and 半田さん (Handa-san) had no
other choice than to try to repair things the best he could.  I found him to
be incredibly polite and patient on that matter.  I would have exploded and
died many times in similar situations, and I've a nice character :-).

So the quoted sentence, above, might be a bit unfair.

-- 

[-- Attachment #2: Type: text/plain, Size: 54 bytes --]

François Pinard   http://www.iro.umontreal.ca/~pinard

Re: 0x0a converted to 0x0d when saving attachments

[Kai Großjohann]

[-- Attachment #1: Type: text/plain, Size: 51 bytes --]

François Pinard <pinard@iro.umontreal.ca> writes:


[-- Attachment #2: Type: text/plain, Size: 334 bytes --]

> 半田さん (Handa-san)

I don't know any Japanese at all, and I wonder: is `-san' a suffix
used when speaking of a person informally; somewhat like the use of
first name (rather than `Mr' or `Ms' and last name) in the Western
culture?  Or is it part of his name?

So many things to learn,
kai
-- 
Life is hard and then you die.

Re: 0x0a converted to 0x0d when saving attachments

[Mike Fabian]

[-- Attachment #1: Type: text/plain, Size: 116 bytes --]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> François Pinard <pinard@iro.umontreal.ca> writes:
> 

[-- Attachment #2: Type: text/plain, Size: 513 bytes --]

> > 半田さん (Handa-san)
> 
> I don't know any Japanese at all, and I wonder: is `-san' a suffix
> used when speaking of a person informally; somewhat like the use of
> first name (rather than `Mr' or `Ms' and last name) in the Western
> culture?

It's comparable to using `Mr. Handa' in English. But only
approximately.

> Or is it part of his name?

No.

Mike

-- 
Dr. Mike Fabian <fabian@mpia-hd.mpg.de>
Max-Planck-Institut fuer Astronomie, Koenigsstuhl 17, 69117 Heidelberg
Telephone: +49(0)6221/528216

Re: 0x0a converted to 0x0d when saving attachments

[François Pinard]

[-- Attachment #1: Type: text/plain, Size: 113 bytes --]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) écrit:

> François Pinard <pinard@iro.umontreal.ca> writes:


[-- Attachment #2: Type: text/plain, Size: 1800 bytes --]

> > 半田さん (Handa-san)

> I don't know any Japanese at all,

Neither I. :-)

> and I wonder: is `-san' a suffix used when speaking of a person informally;
> somewhat like the use of first name (rather than `Mr' or `Ms' and last
> name) in the Western culture?  Or is it part of his name?

Here is what I learned from my Japanese correspondents.

Traditionally (that is, when Japanese do not give into "americanizing"
their name, tss, tss, tss! :-), the family name always come first, followed
by the given name.

You, _when_ you address your correspondent, exclusively use the family name.
The given name is present on business cards or signatures, or when you want
to identify someone while speaking about him/her.  But when addressing your
correspondent directly, the given name is reserved to the close family or
very intimates.  That is, nothing common with the usage we do of the first
name in Western countries.  Japanese are surely tolerant that we use it,
but this is not really the polite way.

Instead, there is the friendly `-san' suffix after the family name, that
corresponds a bit to our usage of the first name, which is the common way
and very acceptable in most circumstances.  If you want to be more formal,
and show greater respect, a bit like we use `Mr.' in English, you use the
`-sama' suffix instead of `-san'.

Notice that the name is written with Kanji characters (ideograms), but the
politeness suffix (like any other kind of grammatical suffix, by the way)
is written with Hiragana characters (phonetic).  The grammatical properties
are written phonetically after the ideograms.  Levels of politeness are
part of the grammar.  There is no correspondence in English, and only a tiny
bit in French (familiar `tu'/ polite `vous' when addressing a single person).

-- 

[-- Attachment #3: Type: text/plain, Size: 54 bytes --]

François Pinard   http://www.iro.umontreal.ca/~pinard

Re: 0x0a converted to 0x0d when saving attachments

[Michael Welsh Duggan]

[-- Attachment #1: Type: text/plain, Size: 116 bytes --]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> François Pinard <pinard@iro.umontreal.ca> writes:
> 

[-- Attachment #2: Type: text/plain, Size: 682 bytes --]

> > 半田さん (Handa-san)
> 
> I don't know any Japanese at all, and I wonder: is `-san' a suffix
> used when speaking of a person informally; somewhat like the use of
> first name (rather than `Mr' or `Ms' and last name) in the Western
> culture?  Or is it part of his name?

It is closer to being equivilent to Mr., Mrs., etc.  It is sort of a
politeness suffix applied to names.  さん(san) is a nice fairly
neutral one.  There are others indicating more familiarity with or
more humbleness in relation to the person, and of course titles can be
used.  One never (well almost never) uses these suffixes when
referring to onesself.

-- 
Michael Duggan
(md5i@cs.cmu.edu)