Re: html prefetching

On Fri, 2 Jun 2000, Andres Kroonmaa wrote:

>
> CacheFlow is using technique to speedup whole web page
> fetch by snooping into initially loaded .html and looking
> for tags that would instruct browser to start fetching other
> components, like gifs, java scripts, etc.
> Now instead of waiting for the browser to initiate fetch for
> each and every component, CF speculatively starts the fetch
> for these itself. Ideally, by the time browser fetches the
> next component, it is already in cache, and served at full
> client speed.
> More than that, CF initiates multiple connections to the
> origin server, to reduce delays with parallel fetches.
> In overall, user perceives web surfing as much faster.
>
> I believe that assumtion that tags found in html would be
> fetched shortly has very high probability of being true.
>
> What thoughts you have about hacking such feature into squid?

I'm skeptical.

Do you know if it actually makes pages load faster?
Or did they just tell you that it does?

It would not be a trival hack. Deciding which
objects to prefetch is somewhat complicated I think.

There's a great risk of pissing off content providers
with aggressive prefetching. I know of at least one
large site that to this day blocks requests that come from
"Microsoft Catapult" which is/was their early
prefetching proxy.

I don't know if anyone blocks requests from cacheflow.
It would really suck if people started blocking
requests from Squid.
Received on Fri Jun 02 2000 - 17:41:14 MDT