From: Philippe Verdy (verdy_p@wanadoo.fr)
Date: Thu Sep 21 2006 - 16:45:46 CDT
From: "Jon Hanna" <jon@hackcraft.net>
> Philippe Verdy wrote:
>> If you cite the case of blogs or community forums,
>
> Once we've cited that case it's worth adding another factor -
> availability of the decompressor.
>
> If we're storing something that's going to go out over a web connection,
> then storing it as DEFLATE blocks could allow us to do work on it
> without decompressing it.
>
> The "perfect" compression algorithm may not be as useful as the
> compression algorithm that happens to be built-into every modern browser.
Is it needed to have a "perfect" algorithm with such "one fits all" rule? The choice of algorithm is a local solution, and nothing prevents changing also the compression method on the fly (i.e. piping a decompression from one algorithm to a compression with another one; call it an adaptation of interface, but such adapters are common in almost all device drivers, and in fact present in every standard communication protocol). This adaptation should not have any impact on the way programs are written.
This archive was generated by hypermail 2.1.5 : Thu Sep 21 2006 - 16:50:10 CDT