1M - any advantage for browsing?

People tend to say that 1M offers little advantage over 512K for web
browsing. Does everyone agree with that? Has anyone seen a
quantified performance comparison?

For ISPs that charge more for 1M, is an upgrade worthwhile if you
mostly browse?

--

Mark W. Lewis, North Somerset, UK

"Mark Lewis" <(E-Mail Removed)> wrote in
news:4139bee2$0$20243$(E-Mail Removed):

> People tend to say that 1M offers little advantage over 512K for web
> browsing. Does everyone agree with that? Has anyone seen a
> quantified performance comparison?

I'd say it's one of those "yes and no" ones.. it stands to reason that most
websites will load quicker on a faster connection, but if you tend to look
at sites where once the page loads you spend ten minutes reading text on
it, then it might not be so noticable as if you tend to click between pages
on highly graphical sites with loads of pictures..

regards
Paul
--
paul <at> spamcop.net

> For ISPs that charge more for 1M, is an upgrade worthwhile if you mostly
> browse?

Hi Mark,

If it's mostly web browsing then I would say no, it's not worth upgrading to
1Meg.

On Sat, 4 Sep 2004 14:09:20 +0100, "Mark Lewis"
<(E-Mail Removed)> wrote:

>People tend to say that 1M offers little advantage over 512K for web
>browsing. Does everyone agree with that? Has anyone seen a
>quantified performance comparison?

There's not much difference between browsing at 1mb compared to 512kb.
Certainly nothing like the difference between a 56k dialup and a 512kb
broadband connection.

Not really noticeable when surfing but it is noticeable when
downloading large files - doing a Windows Update on a clean-installed
machine for example, or downloading ISOs / MP3s / DivX:-)s. Then it
halves the time.

--
Remove .lartsspammers to reply by email.
(E-Mail Removed)

Paul Hutchings <(E-Mail Removed)> wrote:
> "Mark Lewis" <(E-Mail Removed)> wrote in
> news:4139bee2$0$20243$(E-Mail Removed):
>
>> People tend to say that 1M offers little advantage over 512K for web
>> browsing. Does everyone agree with that? Has anyone seen a
>> quantified performance comparison?
>
> I'd say it's one of those "yes and no" ones.. it stands to reason that most
> websites will load quicker on a faster connection, but if you tend to look
> at sites where once the page loads you spend ten minutes reading text on
> it, then it might not be so noticable as if you tend to click between pages
> on highly graphical sites with loads of pictures..

However, consider ebay.
A fairly typical example, or at least a commonly accessed site.

The main page (http://www.ebay.co.uk/) was 207K in total the last time
I looked.
So, it should load in 2s vs 4s?
Well, no.
The page is composed of 57 different files (javascript, button images,
....) and it doesn't finish rendering until all have downloaded.
Most of these are quite small, and the majority of the time was setting
up the connection for each file.

On 04 Sep 2004 16:07:25 GMT, Ian Stirling <(E-Mail Removed)>
wrote:

>Most of these are quite small, and the majority of the time was setting
>up the connection for each file.

IE defaults to 4 connections per site using HTTP 1.0 and 2 for HTTP
1.1 (Mozilla has similar limitations).

For faster d/l of such pages, one has to adjust these values.
--
Cheers,

Guy

** HTML email should be treated in the same manner as sexual acts between
** consenting adults. Only done in private places where willing parties, who
** have agreed beforehand, will see it!

Dr Teeth <(E-Mail Removed)> wrote:
> On 04 Sep 2004 16:07:25 GMT, Ian Stirling <(E-Mail Removed)>
> wrote:
>
>>Most of these are quite small, and the majority of the time was setting
>>up the connection for each file.
>
> IE defaults to 4 connections per site using HTTP 1.0 and 2 for HTTP
> 1.1 (Mozilla has similar limitations).
>
> For faster d/l of such pages, one has to adjust these values.

In some cases, this doesn't help.
For example, if a file is loaded by javascript, in a frame, after a javascript
browser detection routine, then it's going to be at least 20 round-trip times
before it's loaded, nomatter the speed of the link.
And then there is the speed of the link on the other end.
Many sites do not have anything better than 512K upstream.

On 06 Sep 2004 in uk.telecom.broadband, Ian Stirling wrote:

>Many sites do not have anything better than 512K upstream.

have you any figures to back that up ? Just that seeing comments from some
of the US hosting firms, they might take exception... If grc.com has some
T1 or T3 links then I would expect anyone in business to *host sites* for
end users, and hundreds of them, to cough up for higher speeds. Peter M.

poster <us-(E-Mail Removed)> wrote:
> On 06 Sep 2004 in uk.telecom.broadband, Ian Stirling wrote:
>
>>Many sites do not have anything better than 512K upstream.
>
> have you any figures to back that up ? Just that seeing comments from some
> of the US hosting firms, they might take exception... If grc.com has some
> T1 or T3 links then I would expect anyone in business to *host sites* for
> end users, and hundreds of them, to cough up for higher speeds. Peter M.

Small buisnesses.
I'd expect any professional web hosting company to have much more bandwidth
available of course.
It can often work out cheaper to host web pages locally, for some sorts
of site.