Scott made me do it.

[Wesley Kenzie]

On Tue, Aug 18, 2009 at 8:26 PM, Andrew Lewman <andrew at torproject.org>wrote:

> [snip]
>
>  The summary of results:
>
> 1) Native polipo is 54.5% faster on average than native privoxy.  This
> could be due to polipo's caching, http 1.1 pipelining, and it can serve
> bits as fast as they come in from the network.  Privoxy needs to load
> the whole page, scan it, and then send it to the client.  Even if
> privoxy filtering is disabled, it still works the same way.
>
> 2) Polipo caching shines with Tor usage.  Common images are cached, and
> served from the memory cache in single-digit millisecond ranges.
> Privoxy needs to wait for Tor to wholly deliver the bits.  Caching is
> faster, this we know already.  However, from a user perspective, it's
> just faster to load pages.
>
> [snip]
>
> Constructive criticism is welcome.
>

Hello, Andrew.

One thing to consider is that polipo's cache maxes out at 32,000 entries.
 Once you hit that, nothing more is cached.  And there is no built-in
pruning mechanism to keep it smaller. I believe that Juliusz said he would
address this someday, but afaik it has not been done yet.

There also is the annoying fact that polipo crashes regularly, which Juliusz
knows about, but I understand that he has not had enough time in the last
year to fix it.  I believe it is non-trivial to fix.

Lately I have been experimenting with DeleGate proxy cache software, and it
has passed almost all testing so far.  It does hang for some reason when
trying to do a port 443 connection to https://login.icq.com but this is not
a hard crash like polipo - just one of its forked threads goes into a loop.

Wesley
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.torproject.org/pipermail/tor-talk/attachments/20090818/35d41c7e/attachment.htm>