[TUHS] who invented the email attachment?

[TUHS] who invented the email attachment?

[Mary Ann Horton]

I'd like the opinion of this August Group.

Should I make a claim to be the inventor of the email attachment?  (It 
would go on my web site, resume, the Wikipedia page, that sort of thing.)

Here's my understanding of the time line on all of this.

 1. Originally, our files were all plain text and we just included them
    in the email message body.  The ~r command in Kurt Shoen's Mail
    program was typical.  There was no name for this, we were just
    emailing files.
 2. In 1980, I wrote uuencode.  It's stated purpose was to "encode a
    binary file for transmission by email".  I didn't use the term
    "attachment".  It became part of 4.0BSD and later systems, and was
    widely used.
 3. In 1985, Lotus created cc:Mail.  It eventually included attachments,
    using a file store method.  When they added an SMTP gateway later,
    it used uuencode as the format.  I believe cc:Mail first used the
    term "attachment".
 4. Microsoft did the same thing with MS Mail somewhat later, possibly
    in the 1990s.  It also used uuencode in the SMTP gateway.
 5. In 1992, Nathaniel Borenstein and Ned Freed invented MIME.  It had a
    different (and IMHO much better) way to send attachments, and it
    became an Internet Standard sometime later, possibly in 1996.

What do you all think?

     Mary Ann


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20170318/3016713a/attachment.html>

[TUHS] who invented the email attachment?

[William Corcoran]

Hello Ms. Horton,

In reading your initial part of your question, my visceral response was to think "the email attachment was created by whoever developed uuencode."

Only, to discover you wrote uuencode.

So, who created the email attachment?  Mary Ann Horton created the email attachment.  That's who.


Truly,

Bill Corcoran


On Mar 18, 2017, at 7:46 PM, Mary Ann Horton <mah at mhorton.net<mailto:mah at mhorton.net>> wrote:


I'd like the opinion of this August Group.

Should I make a claim to be the inventor of the email attachment?  (It would go on my web site, resume, the Wikipedia page, that sort of thing.)

Here's my understanding of the time line on all of this.

 1.  Originally, our files were all plain text and we just included them in the email message body.  The ~r command in Kurt Shoen's Mail program was typical.  There was no name for this, we were just emailing files.
 2.  In 1980, I wrote uuencode.  It's stated purpose was to "encode a binary file for transmission by email".  I didn't use the term "attachment".  It became part of 4.0BSD and later systems, and was widely used.
 3.  In 1985, Lotus created cc:Mail.  It eventually included attachments, using a file store method.  When they added an SMTP gateway later, it used uuencode as the format.  I believe cc:Mail first used the term "attachment".
 4.  Microsoft did the same thing with MS Mail somewhat later, possibly in the 1990s.  It also used uuencode in the SMTP gateway.
 5.  In 1992, Nathaniel Borenstein and Ned Freed invented MIME.  It had a different (and IMHO much better) way to send attachments, and it became an Internet Standard sometime later, possibly in 1996.

What do you all think?

    Mary Ann

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20170318/4ff59df5/attachment-0001.html>

[TUHS] who invented the email attachment?

[Dave Horsfall]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 497 bytes --]

On Sat, 18 Mar 2017, Mary Ann Horton wrote:

>  1. Originally, our files were all plain text and we just included them in
>     the email message body.  The ~r command in Kurt Shoen's Mail program was
>     typical.  There was no name for this, we were just emailing files.

1.5 They started to include in-line shell scripts, then we piped them 
    into the "unshar" utility, which did basic security checks.

-- 
Dave Horsfall DTM (VK2KFU)  "Those who don't understand security will suffer."

[TUHS] who invented the email attachment?

[Nick Downing]

I can't speak to the original question since I was not around at the time,
but uuencode is really cool. When I first saw it, probably in connexion
with a BBS of some kind, I immediately went and wrote my own 6.5 bit
encoder based on some number corresponding to the number of printable
characters available which when squared gave a 13 bit number. Played around
with this and gave it to a friend for fun to exchange messages in. But the
original is obviously much more in use :)
Nick

On Mar 19, 2017 1:41 PM, "Dave Horsfall" <dave at horsfall.org> wrote:

> On Sat, 18 Mar 2017, Mary Ann Horton wrote:
>
> >  1. Originally, our files were all plain text and we just included them
> in
> >     the email message body.  The ~r command in Kurt Shoen's Mail program
> was
> >     typical.  There was no name for this, we were just emailing files.
>
> 1.5 They started to include in-line shell scripts, then we piped them
>     into the "unshar" utility, which did basic security checks.
>
> --
> Dave Horsfall DTM (VK2KFU)  "Those who don't understand security will
> suffer."
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20170319/4ec68ca7/attachment.html>

[TUHS] who invented the email attachment?

[Mary Ann Horton]

This is great feedback, and the shar mechanism would seem applicable 
here too.  But I can't seem to find a date for shar.

I'm under the impression that shar came later in the 1980s. Google's 
archive for net.sources only goes back to 1987 (unless I'm doing it 
wrong) and clearly shar was already well established by then.

Can anyone put a date on shar, or at least before/after 6/1/1980?

Thanks!

     Mary Ann


On 03/18/2017 07:41 PM, Dave Horsfall wrote:
> On Sat, 18 Mar 2017, Mary Ann Horton wrote:
>
>>   1. Originally, our files were all plain text and we just included them in
>>      the email message body.  The ~r command in Kurt Shoen's Mail program was
>>      typical.  There was no name for this, we were just emailing files.
> 1.5 They started to include in-line shell scripts, then we piped them
>      into the "unshar" utility, which did basic security checks.
>

[TUHS] who invented the email attachment?

[Clem Cole]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 670 bytes --]

On Sun, Mar 19, 2017 at 9:21 PM, Mary Ann Horton <mah at mhorton.net> wrote:

> ​I​
> 'm under the impression that shar came later in the 1980s.

​That would be my memory.   After I left UCB in the mid 1980s.​  Your work
with uuencode certain predated it.  But as Doug points out, putting sources
in email was done before that, al biet fought with error.  I used to a
couple capital X's in column 1 to and other such things to deal with mailer
issues.

You certainly automated that a bit.

Clem
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20170319/2f12a557/attachment.html>

[TUHS] who invented the email attachment?

[arnold]

Mary Ann Horton <mah at mhorton.net> wrote:

> This is great feedback, and the shar mechanism would seem applicable 
> here too.  But I can't seem to find a date for shar.
>
> I'm under the impression that shar came later in the 1980s. Google's 
> archive for net.sources only goes back to 1987 (unless I'm doing it 
> wrong) and clearly shar was already well established by then.
>
> Can anyone put a date on shar, or at least before/after 6/1/1980?
>
> Thanks!
>
>      Mary Ann

ISTR that there was a shar-style program in Kernighan & Pike's "The
UNIX Programming Environment".  That was circa 1984 but shar had been
around for a while before that.  Writing a simple shar creator
in shell was pretty trivial:

	for i in "$@"
	do
		echo echo ---- Extracting $i ----
		echo "cat << 'EOF' > $i"
		cat $i
		echo EOF
	done > file.shar

And I think many people just did that sort of thing on the fly
at first.

Over time shar creators got fancier, with the ability to make directories,
uuencode binaries, and split large files and put them back together
again. (Cf. the kind of postings in the netnews *.sources groups.)

Not sure this helps to put a date on shar, but certainly shar archives
were widely used in the 1983-1985 time frame...

Arnold

[TUHS] who invented the email attachment?

[Mary Ann Horton]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2044 bytes --]

I was at Berkeley until July 1981.  The oldest SCCS file I have is
4/1/81 (for my dissertation project) and that was clearly my first use
of it.  I wasn't using SCCS in 1980 when I wrote uuencode.  uuencode got
SCCS-ized later when they put all of 4.xBSD under SCCS.

On 2017-03-20 03:27, schily at schily.net wrote:

> Mary Ann Horton <mah at mhorton.net> wrote:
> 
>> I'm under the impression that shar came later in the 1980s. Google's 
>> archive for net.sources only goes back to 1987 (unless I'm doing it 
>> wrong) and clearly shar was already well established by then.
>> 
>> Can anyone put a date on shar, or at least before/after 6/1/1980?
> 
> BTW: do you remember why you did not check in uuencode into the SCCS?
> 
> /*--------------------------------------------------------------------------*/
> ...
> Wed Jul  6 11:06:51 1988 bostic
> * uuencode.c 5.6
> * uudecode.c 5.4
> written by Mark Horton; add Berkeley specific copyrights
> 
> Wed Feb 24 20:03:58 1988 rick
> * uuencode.c 5.5
> use library fread instead of rolling your own
> 
> Mon Dec 22 14:43:09 1986 bostic
> * uuencode.c 5.4
> bug report 4.1BSD/usr.bin/2 and 4.1BSD/usr.bin/3
> 
> Wed Apr 10 15:22:23 1985 ralph
> * uudecode.c 5.3
> more changes from rick adams.
> 
> Tue Jan 22 14:13:07 1985 ralph
> * uuencode.c 5.3
> * uudecode.c 5.2
> bug fixes and changes from Rick Adams
> 
> Mon Dec 19 15:42:38 1983 ralph
> * uuencode.c 5.2
> use a reasonable mode for encoding data piped in.
> 
> Sat Jul  2 17:57:51 1983 sam
> * uuencode.c 5.1
> date and time created 83/07/02 17:57:51 by sam
> 
> Sat Jul  2 17:57:49 1983 sam
> * uudecode.c 5.1
> date and time created 83/07/02 17:57:49 by sam
> /*--------------------------------------------------------------------------*/
> 
> In special, do you know why it has been checked in by Samuel Leffler and 
> whether it existed before July 1983?
> 
> Jörg
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20170320/4a862f64/attachment.html>

[TUHS] who invented the email attachment?

[Josh Good]

On 2017 Mar 18, 16:46, Mary Ann Horton wrote:
> 
> Should I make a claim to be the inventor of the email attachment? [...] 
> 
> 2. In 1980, I wrote uuencode. [...]

I would say yes to your question. More or less.

However, the "attachment" as a concept has more to do with
"presentation" of the email message: the "attached" part is always, even
in MIME messages, part of the body of the RFC822 message.

In that sense, it does not matter whether it is inline source code,
inline 'shar'-ed files, inline UUencoded, or inline MIME. What makes
it an "attachment", is the MUA ability to present it and to manage it
as a distinct part of the email message, different from the headers and
different from the "textual" body. 

So, according to that, the inventor of the "attachment" was whoever wrote
the first MUA capable of presenting and managing certain body parts of
a RFC822 message as such separate entity. The "attachment", therefore,
would be a user interface design decision, undoubtedly facilitated by
some underlying technology such as uuencode which defined a boundary to
mark the 'begin' and 'end' of the inlined non-textual data.

But I have been wrong before - many times.

-- 
Josh Good

[TUHS] who invented the email attachment?

[Doug McIlroy]

> Should I make a claim to be the inventor of the email attachment?

uuencode was critical to attaching arbitrary files, and I am sure
one can find emails with uuencoded bits in them that read, "please
find attached ...". But they would have said the same thing if
what was being sent was source code. So attachment in that sense
obviously predated uuencode. But to identify that kind of
attachment with what mean by the word today is like identifying
cat with tar.

Doug