Receive Window and TCP Speed

Given that TCP default receive window is 64K, would a page load much
faster if the content of every GET in the HTML was < 64K? This
started as an Apache question, but then I realized "receive window" is
client-side.

We don't seem to care anymore about file sizes because of bandwidth
increases, but pages with many, many images seem to load slow due to
congestion of some sort.

R

"rocco" <chrisgirocco@gmail.com> wrote in message
news:faf65ddf-cee3-4df9-83b9-7cbae719426a@u10g2000prn.googlegroups.com...
> Given that TCP default receive window is 64K, would a page load much
> faster if the content of every GET in the HTML was < 64K? This
> started as an Apache question, but then I realized "receive window" is
> client-side.
>
> We don't seem to care anymore about file sizes because of bandwidth
> increases, but pages with many, many images seem to load slow due to
> congestion of some sort.

If you're worried about effective throughput, at least turn on compression
(mod_gzip) for all items that aren't already compressed. Don't forget that
your script outputs should also go through the deflater.

On Mar 4, 10:26*pm, "D. Stussy" <s...@bde-arc.ampr.org> wrote:
> "rocco" <chrisgiro...@gmail.com> wrote in message
>
> news:faf65ddf-cee3-4df9-83b9-7cbae719426a@u10g2000prn.googlegroups.com...
>
> > Given that TCP default receive window is 64K, would a page load much
> > faster if the content of every GET in the HTML was < 64K? *This
> > started as an Apache question, but then I realized "receive window" is
> > client-side.
>
> > We don't seem to care anymore about file sizes because of bandwidth
> > increases, but pages with many, many images seem to load slow due to
> > congestion of some sort.
>
> If you're worried about effective throughput, at least turn on compression
> (mod_gzip) for all items that aren't already compressed. *Don't forget that
> your script outputs should also go through the deflater.

So then you could optimize the quality or clustering of emails in
conjunction with modgzip based on typical compression factors to
achieve the fastest transmission? Or is the whole window size aspect
insignificant compared to these other factors?

What is your opinion?

R