Bug#932: Pine over-encodes files and auto-requests document
Terry Gray writes ("Re: Bug#932: Pine over-encodes files and auto-requests document"):
> Dale,
> This issue has been discussed numerous times on comp.mail.mime and
> comp.mail.misc, but here's the executive summary:
>
> o It is true that Pine encodes all attachments, even text.
> o The intended way to send text files that (a) don't need to arrive
> exactly intact or (b) are going to non-MIME-capable destinations
> is indeed to use ^R rather than MIME-attach.
What if I want to send a text file and don't know whether the
destination supports MIME (or if I'm a clueless user who has never
heard of MIME) ?
The correct solution is to use a MIME multipart message with a
Content-Transfer-Encoding of 7bit. That way a user of a non-MIME-
capable MUA can cut-and-paste the text, and MIME users get the
benefits of the MIME features. Furthermore, the message is not
unnecessarily expanded by 33%.
However, Pine
(a) doesn't allow me to do this and
(b) makes the `default' that users tend to use be something that
doesn't work for non-MIME-capable destinations, without warning
them.
(b) is particularly pernicious. I have seen quite a few Pine users
confused when they receive a message like this:
From: a.correspondent@a.site.somewhere
In-Reply-To: ... <Pine....@host.pine.users.site> ...
You write:
> This is a MIME message. The first part is readable text,
> the rest will likely not be readable without MIME-aware tools.
Yes, you're absolutely right, I can't make head or tail of this
guff you've sent me. Can't you get a proper mailer, or put the file
up for FTP ?
Respectfully yours,
A. Correspondant.
> 9831kjjdkjq223874q09 [...]
They can usually be heard to exclaim `but I didn't write that!'.
> o There are excellent reasons for this behavior: we want Pine to be
> known for dependability, and attaching a file in Pine should be
> just as reliable as transferring via FTP. Since the Internet message
> infrastructure is not completely 7bit safe, this goal requires
> encoding even 7bit files.
Pine is rapidly becoming known not for dependability, but for
arrogance and disregard for standard net practice. A dependable
mailer would do what was most likely to be the right thing.
Furthermore, your observation that there are mailers that mangle even
7bit messages is true but irrelevant. There are practically no
gateways that mangle messages and which also have MIME-capable MUA's
behind them, so that sending MIME just produces a message that can't
be read *at all* at the far end, rather than one that can be
reconstructed (as would often be the case if it were sent 7bit or
uuencoded).
> As you can see, we strongly disagree with the characterization that Pine
> is broken at all in this regard, much less "severely" broken.
If I may be potentially offensive, this kind of attitude is one of the
things that is making Pine very unpopular amongst net-aware people.
It seems to many of us that the Pine developers refuse to listen to
the experience of those who have been around longer than they have and
who actually know what they're talking about. Instead they take a
very dogmatic attitude.
What makes matters worse is that most of Pine's brain damaged features
work fine with other copies of Pine, and take effect only when some
poor sod somewhere who doesn't like Pine has to deal with one of its
messages. This may be good marketing, but it's very poor practice.
Ian.
Reply to: