Re: More namespacing woe ...

[Miguel de Icaza <miguel helixcode com>]

> Well, this is confusing -- is the request to gzip it, or to gunzip it?
> There isn't any way for the moniker to know; while you could argue
> that you'd never want to gzip an already gzip'd stream, maybe you would --
> i.e. an application that uses a moniker to get at a stream (of unknown
> data), which it then compresses using the gzip moniker before sending
> it off elsewhere. If the gzip moniker was called "gzip" for both cases,
> it would end up uncompressing the original stream, causing reading from
> the stream interface to give you uncompressed data, where you were
> expecting gzip'd data. (Yes, gzipping data twice is a losing operation,
> but if you're trying to make something general enough to not have any
> knowledge of the input stream format, you'd end up wanting to compress it
> anyway.) Note that this doesn't come up with the VFS case, since it only
> supports gunzipping the data.

Mhm, you raise an interesting point.  From the moniker perspective,
using #gzip: should compress the parent, and return data compressed on
the other end.  

So I think we should keep #gunzip: for gunziping, and #gzip for
gzipping.

Miguel.