On Fri, 12 Oct 2001 12:18:40 John Seago wrote:
I must apologise for this mornings message. I wrote it just after the pain killers were kicking in, and as a consequence it seems a little incoherent, let me clarify.
Like Syd I have noticed =20 at the ends of the lines, (or where the line ends were when I wrote the message). This only seems to occur on the list, (unless those I have emailed individually know otherwise). Messages between myself and Syd don't show the =20, at least at my end. It appears to me therefore, as the problem comes and goes, irrespective of my settings, that the =20 appears somewhere between Syd and I, as senders, and the list itself.
It is perfectly acceptable for mail software, including the software which stores and forwards the e-mail along its route, to convert spaces to =20 - it is a transfer coding called quoted printable.
It is also happens that mail software undoes this coding and restores the message to its original state with spaces in place of the =20 codes and this can happen en-route as well as at the final destination.
There are some requirements for this to work properly though:
1. Any peice of mail software that encodes the email using the quoted-printable system must declare the fact by adding the Content-Transfer-Encoding: quoted-printable header to the messages and subsequently any mail software that processes the message but does not decode it must preserve this header.
2. Mail software which decodes quoted-printable back to normal text must adjust the header or remove it as appropriate.
3. Mail reading software must be able to cope with transport encodings like quoted printable.
So, the problem is due to piece of software doing the encoding without adding the header, dropping the header without decoding the message, or being incapable of doing the decoding when the message is read.
This could be any piece of mails software from that used to create and send the message including everything on route and the mail reader used to read the message.
I have run some tests, and if the alug list server is sent quoted-printable messages it decodes them and sends them to the list recipients as 8bit messages (as long as the header is there to tell the software that quoted-printable was used) so I think the list server is unlikely to be the culprit.
I am sending through the 01986 (Bungay?) exchange, which in one of the exchanges between myself and BT over the quality of their service, they told me had yet to be updated. I am of the opinion that it is somewhere along the route that the =20 appears on the final display of the message.
When it comes to e-mail messages the critical information is not which pieces of network equipment the message went through (as these don't look inside the message) but which mail servers the message went through. Most mail readers have the option to turn on viewing of "All" headers and this will include some headers like this:
Received: from ns.drake.org.uk ([209.61.182.158]) by punt-1.mail.demon.net id aa1004851; 12 Oct 2001 17:56 GMT Received: from amavis by mail.drake.org.uk with scanned-ok (Exim 3.33 #3) id 15s6ea-0004Js-00 for fozzy@pelvoux.demon.co.uk; Fri, 12 Oct 2001 19:03:24 +0100
These headers trace the message's progress through the mail system with the most recent being at the top. If someone complains to you about the =20 problem try to have then send the message back to you with the headers of the message as they received it still in tact. This helps find where the message has been and where the problem may lie.
As for the BT exchanges, I am not sure what modernisation they are taling about. AFIAK the old mechanical exchanges have gone now so there should be no exchanges that can't offer a decent service. Not every exchange has ISDN or DSL capability but they should all support ordinary modems (whether or not DLE grooming is installed).
Steve.