[WEB4LIB] RE:

[Thomas Dowling]

Keith Jenkins wrote:
> This URL should be:
>   http://surveymonkey.com/s.asp?u=28432945321
> 
> I've noticed that URLs in listserv messages occasionally have certain
> characters like [equals] and [space] converted to "=3D" and "=20",
> respectively.  Does anyone know at what point in the e-mail process
> this occurs?
> 

It happens when a mail client encodes a message as quoted-printable
(q-p), as per RFC 2045 section 6.7.  For example:

  Content-Type: text/plain; charset=iso-8859-1
  Content-Transfer-Encoding: quoted-printable

Printable US ASCII characters other than the equals sign appear as
themselves; everything else is encoded as an equals sign followed by the
character's two-digit hex number (just like URL encoding, except with an
equals sign instead of a percent sign).  Lines longer than 76 characters
are wrapped with a equals sign by itself to indicate a soft line break
that can be reflowed by a clueful mail client.  Spaces at the ends of
lines are explicitly encoded as =20 to make sure they don't get lost.

Q-P encoding was created to allow 8-bit characters and messages with
long lines to travel through e-mail systems that only handled 7-bit
characters and 80-character line lengths.

The listproc software at the Berkeley sunsite is so creaky that it
doesn't relay q-p messages with the appropriate header, so mail clients
treat them as plain text.


-- 
Thomas Dowling
tdowling at ohiolink.edu