Harry Putnam <reader@newsguy.com> writes:

>> Can you M-x toggle-debug-on-quit RET and press C-g on the prompt to
>
> OK, but it doesn't happen that often, and I'm not sure what brings it
> on.
>
>> get a backtrace?  I think Gnus is supposed to guess charset
>> automatically.  Hm.  Unless you use emacs in unibyte mode.  Do you?
>
> Not afaik.  I've set nothing knowingly to do that. Is there a C-h v
> entry I can check to see?

Here is a back trace:
It came up when posting: 
  Message-ID: <m3sn3dbugl.fsf@newsguy.com>
Here.

Debugger entered--Lisp error: (quit)
  completing-read("Charset used in the article: " (("japanese-jisx0213-2") ("tibetan") ("indian-2-column") ("chinese-cns11643-7") ("chinese-cns11643-6") ("chinese-cns11643-5") ("chinese-cns11643-4") ("chinese-cns11643-3") ("ethiopic") ("mule-unicode-0100-24ff") ("mule-unicode-e000-ffff") ("mule-unicode-2500-33ff") ("tibetan-1-column") ("indian-glyph") ("indian-is13194") ("arabic-2-column") ("lao") ("ascii-right-to-left") ("arabic-1-column") ("arabic-digit") ("vietnamese-viscii-upper") ("vietnamese-viscii-lower") ("ipa") ("chinese-sisheng") ("chinese-big5-2") ("chinese-big5-1") ("japanese-jisx0213-1") ("chinese-cns11643-2") ("chinese-cns11643-1") ("japanese-jisx0212") ("korean-ksc5601") ("japanese-jisx0208") ("chinese-gb2312") ("japanese-jisx0208-1978") ("latin-iso8859-14") ("latin-iso8859-15") ("latin-iso8859-9") ("cyrillic-iso8859-5") ("latin-jisx0201") ("katakana-jisx0201") ("hebrew-iso8859-8") ("arabic-iso8859-6") ("greek-iso8859-7") ("thai-tis620") ("latin-iso8859-4") ("latin-iso8859-3") ("latin-iso8859-2") ("latin-iso8859-1") ("ascii") ("eight-bit-control") ...) nil t nil charset-history nil)
  mm-read-charset("Charset used in the article: ")
  mm-encode-body(nil)
  mml-generate-mime-1((part (type . "text/plain") (contents . "Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:\n\n> Harry Putnam <reader@newsguy.com> writes:\n>\n>> Raja R Harinath <harinath@cs.umn.edu> writes:\n>>\n>>> Actually it isn't :-)  The QP mangled form of '=' is '3D', not '3D'\n>>> which the message showed.\n>>\n>> Those two look indentical here..\n>\n> The right form is an equals sign, followed by a three, followed by a\n> capital d.  The wrong form is missing the equals sign.\n>\n> Your Gnus seems to delete the equals signs...\n\nViewing the above referenced message with C-u g doesn't help either.\nI still see '=' is '3D, not '3D'\n\nWhat might be causing this strange behavior? \n")))
  mml-generate-mime()
  message-encode-message-body()
  message-send-mail(nil)
  message-send-via-mail(nil)
  message-send(nil)
  message-send-and-exit()
  gnus-draft-send(56 "nndraft:drafts" t)
  gnus-draft-send-message(nil)
  call-interactively(gnus-draft-send-message)