Re: gzip/deflate compression/encoding

On Jun 27 2005, Ned Freed wrote:
> 
> Of course this leaves open exactly what compression/encoding pairs
> need to be defined for use as CTEs. As you say, there are potentially
> five of them, but I only see three that are genuinely useful:
> 
>    binary-8bit
>    deflate-8bit
>    deflate-base64
>    deflate-binary
> 
> Of these I see little value in the last, but the others all make good sense.

There's another potential problem with CTEs that occurred to me,
namely inherent brittleness. The text based formats are largely robust
to localized corruptions such as line splitting etc., and were
obviously designed with that intent. However, small corruptions can
irrecoverably destroy compressed data streams. 

So I would be hesitant to recommend CTE usage for textual data on
principle, even if it is 8bit internationalized. As a long term
storage medium, it is much more important that the RFC 2822 format
allows graceful degradation rather than a 2x-5x typical size reduction.

-- 
Laird Breyer.