develooper Front page | perl.fwp | Postings from April 2002

Re: Self extracting Perl ARchive?

Thread Previous | Thread Next
Bernie Cosell
April 1, 2002 09:10
Re: Self extracting Perl ARchive?
Message ID:
On 1 Apr 2002, at 17:04, Philip Newton wrote:

> On Mon, 1 Apr 2002 07:56:10 -0500, (Bernie
> Cosell) wrote:
> > As another small 'twist', if in addition to everything else you did 
> > something akin yEncoding the __DATA__, then you'd have an email-
> > able/usenet-postable archive format.
> Um, yEnc is not really email-able since it produces 8bit data and email
> is AIUI not 8-bit clean everywhere, while news is supposed to be 8-bit
> clean. I suppose you could try it (preferable with proper MIME headers
> such as "Content-Transfer-Encoding: 8bit").

Well, *dejure*, neither news nor email is supposed to be 8-bit clean.  
Email *can* be 8-bit clean if the server supports the 8BITMIME option, 
but there's no such option for NNTP.

But *defacto*, because of the widespread use of ISO-Latin [both because 
of European use AND the fact that it is the charset for HTML], virtually 
all servers pass through 8-bit stuff pretty well these days.  But only 
'text-like' 8-bit stuff... basically, you need to avoid screwing around 
with NULLs or CRs or LFs and you need to stick in a CR/LF every now and 
then.  But other than that, you can do pretty well if you stick to the 
252 or so chars that that leaves you.

No need to do really 'yenc', with all its fanciness and such, but I think 
it'd be neat to just to do some encoding that is more server-friendly 
than brute-force-binary and more efficient than base64 encoded.


Bernie Cosell                     Fantasy Farm Fibers     Pearisburg, VA
    -->  Too many people, too few sheep  <--          

Thread Previous | Thread Next Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at | Group listing | About