Re: viewing uuencoded content

[Emmanuel <e allaud wanadoo fr>]

On 04.06.2002 11:13 Carlos Morgado wrote:
> 
> On 2002.06.04 09:10:00 +0100 Ian Leonard wrote:
>> 
>> On 2002.06.03 23:19 Carlos Morgado wrote:
>>> On Mon, Jun 03, 2002 at 08:53:08PM +0200, M. Thielker wrote:
>>> > first 50 lines of each message. Present that part as if it were a 
>>> MIME
>>> part.
>> 
>> [ snip ]
>> 
>> I understand the reasons for not wanting the completely automatic
>> approach to this. However, for simplicity I told a small lie in my
>> original posting. It's not me who is having the trouble - it's my
>> wife and daughter - so it's a question of what they can do.
> 
> eh, a problem i can relate to :)
> 
>> 
>> I wonder how much of a problem it is or what other weird formats
>> there are? Maybe look for '\nbegin [a number] ' then look for '\nend\n'.
>> There are other clues as well - most of the lines are the same length.
>> 
> it's fairly simple and you can even make it smart enough to avoid false
> positives but the computational load would be heavy compared to the
> amount of positives
> 
>> I would guess that you don't want to add an extra filtering mechanism
>> if it will hardly ever be used.
>> 
> 
> it's more like, how generic would this be ? it's actually not very hard 
> to
> do it (at least i don't think so - famous last words :)) but if we do it
> we should do it right.
> 
> i still think the proper way is getting the other end to use mime,
> uucoding is Good but has it's limited uses.
> 

We could imagine that filters could pipe the message body (and/or headers) 
to an external prog. Security concerns should be kept in mind doing this I 
guess ;-)
Bye
Manu