Please keep discussions on the mailinglist.
On tis, 2008-07-22 at 14:12 +0700, johan firdianto wrote:
> what kind logic ?
1. Analysing how rapidshare really works..
2. From the analysis, figure out how to properly identify downloads.
3. Add code to Squid (or maybe an ICAP server) capable of performing
this automatically.
> I already succeed cached many youtube video (so far 180K video), just
> use combination redirect program and small script that ekstrak flv
> from store.log. StoreURLrewrite is good, but according my experience
> some times hit some times not.
It depends on freshness settings. You can tune this with
refresh_pattern.
> Than I extract the saved object video
> from cache_dir to flv directory. So redirect program check
> avalaibility in flv directory. So I just monitor pattern of video url
> of youtube.
That's a good idea.
> But if they change from GET method to POST, than I'm
> knocked out.
Yes.
> From your email below, it looks there's still hope to do.
> Pleasee.. let me know
As long as they don't switch to https with client certificates for
authentication there is hope to work around content provider brain
damage. The question is just how much work is required to implement the
workarounds, and what pieces needs to be touched in the process..
Regards
Henrik
This archive was generated by hypermail 2.2.0 : Wed Jul 23 2008 - 12:00:05 MDT