On May 31, 1999 at 12:56, Jason Molenda wrote:
> I am an Apache module that the mozilla people wrote called
> mod_gzip_content. If the client browser can handle gzipped content, and
> if a .gz version of the requested file exists, it sends the .gz version.
> If the browser doesn't handle gzipped content, then the plain .html
> file is sent. To use this, you obviously need to keep the plain html
> and the .gz file in sync.
...
> What I'd really like to do is have mhonarc gzip all of the .html files it
> writes/updates during a -add, including the indexes. The GZIP* resources
> seem to be for writing the .gz files in lieu of the .html files, which
> would hose our non-gzip-handling clients.
>
> I don't suppose someone already come up with a patch to handle this case,
> have they?
No one has contributed one.
Is mod_gzip_content smart enough to have the server decompress the
data if the client does not support gzip?
To have mhonarc generated both the uncompressed and compressed versions
together would take more than a little patch; probably a bunch of
little patches. Anytime a file is created, or edited, a gzip copy
will have to be created. I can look into it to see if the modifications
are quick or if it will take some time.
If you are doing archive updates via cron, you can have a post-processing
process that creates .gz versions of html files. A time stamp check
can be made for pre-existing .gz versions to check if the uncompressed
versions is newer or not to minimize work.
--ewh
