ietf-822
[Top] [All Lists]

Re: gzip vs. deflate (was RE: gzip-8bit)

2003-02-27 10:34:33

Dan Kohn writes:
If we do gzip, the FNAME parameter will definitely be a SHOULD NOT set, MUST ignore, since Content-Disposition with RFC 2231 can do the same thing with i18n.

You also need some similar rules for MTIME vs. modification-date-parm. There may be more, although I can't think of anything right now.

The main advantage of gzip that I can see is the CRC32

And the main disadvantage is an ugly paragraph of SHOULD NOT, MUST, SHOULD NOT, MUST. And the attendant extra code in careful readers.

and the ease of piping things through the gzip program.

Isn't that the argument for shar over tar?

Sadly, I feel sure that the programmers that are swayed by the ease of system("gzip") vs. calling zlib are the programmers who don't implement new RFCs (correctly) anyway.

--Arnt

<Prev in Thread] Current Thread [Next in Thread>