ietf-smtp
[Top] [All Lists]

Re: [idn] Re: 7 bits forever!

2002-04-03 12:18:39

If we are saying that the ACE <=> IDN(UTF8) so why is the two headers not
in
sync?

To paraphrase somebody (Was it Vernon Schreier?):

Either the two fields are semantically identical, in which case it is
superfluous to send one of them, or they are not, in which case there is
breakage if you use the one with less semantic content.

His comment was regarding text/plain and text/html, but is equally true
here.

If the headers are one of the X- headers which is vendor specific, even
existing mail will have X- headers that can only be read by certain MUA, and
simply ignore by most MTA, so what breakage are you talking about?? Unless
the existing implementation of some MTA and MUA is not correct and raise
problem when parsing the headers that has X-...

also if new MUA always use the new headers and only old existing MUA wont
be
able to read the new headers will use the old ACE headers, so why MUA
wont
know which one is more recent?

You also have the case where you're using a new MUA, but the mail could
have
passed through and old MTA that didn't know about the new headers, and thus
did a conversion on the old headers but not the new ones.

How is the MTA going to do a conversion, if it is an old MTA that wont be
able to read this new headers, why and how would the MTA all of a sudden
converting a line that it cannot read, also the old header contains the ACE,
so why is the old MTA trying to change the ACE address, will existing MTA
change your valdis(_dot_)kletnieks(_at_)vt(_dot_)edu to something else?!

Those of us who lived throught the %-hack ages know ALL about this - you'd
get a mail that had a replyable 'From:' because an MTA had %-hacked it
for you, but a 'reply all' would fail miserably because the MTA had failed
to %-hack the CC: list.
So that's why I never said putting the UTF8 into existing headers where
existing MUA or MTA may % escaped it, what I said is USE EXISTING HEADERS
WITH ACE, AND ADD NEW VENDOR SPECIFY HEADERS LIKE X-IDN-FROM, X-IDN-TO, etc
FOR THE UTF8, SO THAT NEW MUA,MTA WILL BE ABLE TO PICK UP THOSE HEADERS AND
USE THEM, AND OLD EXISTING MUA, MTA WILL JUST USE THE ACE IN THE OLD HEADER
LINES. If MUA or MTA that touches the lines X- which it doesnt even
recognize, then I think that should be considered a bad or incorrect
implementation.



<Prev in Thread] Current Thread [Next in Thread>