ietf-822
[Top] [All Lists]

Re: interoperablity

1994-09-16 07:37:21
Keith Moore writes:

I believe it is an engineering exercise. That is we should look
how well the different approaches work for the different cases.
The criteria is: what works best for the users? 

There's a bit more to it than that.  First of all, "the users" are quite a
diverse group, with lots of different capabilities, which are difficult to
characterize in any single statement.  It's very hard to say that one single
solution works best for "the users", when one solution works better for some
users and another solution works better for others.  Even if you can make a
general statement for "most users", it might hold true a year from now, and a
solution biased towards the needs of "most users" might so limit the
capability of the other users that they would go off and develop their own
(incompatible) solution.

Agree, that was what I meant with "different cases" - including
different users.

in the case QP vs MNEM, I would say that for each QP encoding, 
I can find a better MNEM encoding, in the sense that a reader would
more easily understand it. QED.

Disagree, for the following reasons:

a) mnemonic is not fully general -- it's really works well only for latin
text.  Even then, it's not clear how favorably it would compare with
quoted-printable.  Most users would still complain anyway if their 8bit
characters went away and were replaced with mnemonic during a transition to
mnemonic.

Well, they would not complain more than that their character was
replaced by QP.

Mnemonic also works better than QP, for cyrillic,greek, arabic, kana,
and others.

b) it does not provide the same capability as quoted-printable
for error-free data transfer over the mail network.

I dont see why. I would say it provides more error-free transmission
of data than QP, as you can eg get all your CP850 data thru.
If you want binary info thru, you should use base64 instead.

c) mnemonic doesn't fit as well into the MIME scheme, where content
is independent of encoding.  This scheme makes it much easier to
add new content-types to MIME, than would otherwise be the case.

Mnemonic is designed for text type only. Use of it would not hinder
other and new types in MIME.

Being able to understand the encoded form of a letter is not the only
criterion to be used in declaring mnemonic "better" than quoted-printable. 

Well, QP is simpler todo, but then it is just a matter of some tables.
We should look into the criterias and weight them.

And even if mnemonic were widely accepted, there would still be a need for
quoted-printable.

For what?

I do not see the problems with scientific texts nor program texts
with respect to MNEM. Examples?

Both scientific/mathmetical text and computer programming language text use
more special characters than normal text.  It would seem to be harder to
distinguish mnemonic's special characters from those used by the surrounding
text, if the surrounding text were a computer program, than if it were
ordinary text.

Computer programs would normally just use ASCII. If they use
more, then it would normally be letters, to improve the readability.
(this is at least my experience).

scientific/math text cannot do much with QP, -there is simply not
character sets available for their use (except 10646).
Mnemonic also have mnemonics for math/scientific, much in the
style of troff, or other text processors. Should not be a problem
to those writers.

Quoted-printable looked like a reasonable compromise when it was 
proposed, but didn't look as good when it started cropping up 
everywhere.  I'm not sure that either mnemonic or iso-2022-int-* 
would fare any better when subjected to the same scrutiny.  Do 
your users like mnemnonic even if they were already
accustomed to seeing "the real thing" on their screens?  Didn't they
complain when their real characters went away?

Users want "the real thing". We did not make "the real thing"
go away, if we could still provide it. [...] For the sites that 
we were MX-ing for, we recorded their preferred charset and converted
back-and-forth [...]


This is not a fair basis on which to judge the likely acceptance of
mnemonic (especially when compared with quoted-printable) in the 
worldwide Internet.  

Well, this was not the base of users I was talking about.
I was talking about our user base *not* set up for these services,
and *non-customers* - they got raw ASCII mnemonic. I would say maybe
100 small to large organisations were involved in this way.

+ In the third place, some of your users complained anyway because their
correspondents were having trouble!

Well, this was a separate problems, mostly due to 7/8 bit problems,
that would be equivalent in both QP and mnemonic.

All of this seems to indicate that mnemonic would generate the same kinds of
complaints, that we're now hearing about quoted-printable.  It might well
generate fewer complaints, but would the difference be enough to justify
changing direction now (even for netnews)?  I doubt it very seriously.

There were complaints, but not the same kind. It interoperated.
IMHO it meant the difference between people communicating (with MNEM),
or not communicating (which QP gave reason to in most cases).

keld

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