[h-e-w] Re: Gnus, anyone?

[richardk]

>>>>> "JR" == Jason Rumney <address@hidden> writes:
    JR> 
    JR> Paul Kinnucan <address@hidden> writes:
    >> I'm a bit further down this path than you. I tried
    >> the gnus for news and email route and hated it almost
    >> instantly.
    JR> 
    JR> Almost instantly is not sufficiently long to
    JR> accustom yourself to Gnus :-)
    JR> 
    >> There are a few tricks to using vm on Windows,
    >> including compiling the base64 encoder/decoder
    >> program that comes with the vm distribution.
    JR> 
    JR> What is so wrong with the built in base64 support
    JR> that VM needs to reinvent it?

I believe VM does use the built-in base64-encode-region and
base64-decode-region functions.  VM does not use
base64-decode-region under GNU Emacs though as shown below.

I quote from vm-mime.el from VM 6.97 :

(defun vm-mime-base64-decode-region (start end &optional crlf)

    ...

          (cond
           ((and (featurep 'base64)
                 (fboundp 'base64-decode-region)
                 ;; W3 reportedly has a Lisp version of this, and
                 ;; there's no point running it.
                 (subrp (symbol-function 'base64-decode-region))
                 ;; The FSF Emacs version of this is unforgiving
                 ;; of errors, which is not in the spirit of the
                 ;; MIME spec, so avoid using it.
                 (not vm-fsfemacs-p))

Note the last comment which gives why GNU Emacs version of
base64-decode-region is not used.
However, base64-encode-region is used if available.

(defun vm-mime-base64-encode-region (start end &optional crlf B-encoding)
    
    ....

          (cond
           ((and (featurep 'base64)
                 (fboundp 'base64-encode-region)
                 ;; W3 reportedly has a Lisp version of this, and
                 ;; there's no point running it.
                 (subrp (symbol-function 'base64-encode-region)))