nmh-workers
[Top] [All Lists]

Re: nodashmunging changed to nobitstuffing

1998-05-20 10:03:11
    Date:        Tue, 19 May 1998 23:15:40 -0400
    From:        Richard Coleman <coleman(_at_)math(_dot_)gatech(_dot_)edu>
    Message-ID:  
<199805200315(_dot_)XAA09666(_at_)cypress(_dot_)math(_dot_)gatech(_dot_)edu>

  | The BIG question is whether I should leave the default the same, or
  | change it so that bitstuffing is not done by default.  Since it is
  | uncommon to burst RFC934 encapsulated messages these days, I'm
  | tempted to turn the bitstuffing off by default.

If you get a bounce message from munnari.oz.au you get it in 934 format,
precisely because it is so easy to burst (either with something which
just does that, or just with a text editor).   Many digests also still
come in 934 format.  Mime digesting seems comparatively rarely used, so
much so in fact, that when I had a message from a listserv, telling me
to forward back the message to confirm my list subscription, and I did that
by sending back a "content-type: message/rfc822" the listserv bitched, and
said it didn't know how to handle that content type...   I had to repeat
the thing, this time using 934 forwarding.

The reason the option was undocumented in mh was because in normal use no-one
should ever be using the thing (it also isn't in any sense "bitstuffing" so
I don't think your new name is quite right, "dashstuffing" or even 
"bytestuffing" would be a better name change if "dashmunging" isn't what you
want - though changing the name without a good reason is just another
unnecessary incompatability with MH).   To go from "no-one should ever
normally use this" to "this is now the default" is a pretty big change.

Not that I much care of course, as I'm not an nmh user, and don't plan to
be, except that every incpmpatible change is going to make exmh harder to
support.

kre



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