>> Sure plastics have been around for ages, but
>> today we take for granted there are plastics that you can leave in
>> direct sunlight for decades without fading or going brittle.
>
> Really? Where are they?

If you buy a 10 year old BMW you don't expect the dashboard to crack if 
you lean on it or for it to have gone yellow. Or most bumpers on cars 
now are plastic that can withstand 2 tons forcing them into another car 
at a few mph without damage (IIRC that's an EU requirement now). That 
sort of performance from plastics has been developed gradually after 
they were initially invented. The first LCD displays were around the 
70's, but it's not like nothing has changed since then, the "invention" 
is only the first step, after that Engineering and development takes over.

> That's kinda my point. When the Internet first started, the limiting
> factor with surfing the web was how damned long it took to load the HTML
> and all the images. That has long since stopped being a problem. So
> until the next bandwidth-heavy thing comes along, there's no real reason
> to increase.

I remember when an mp3 took hours to download, today it's an mp4 and 
software that takes hours. Until everyone has 500ppi screens with all 
content at that resolution downloadable within seconds then I see no 
reason for the continual increase in speeds to stop.

> BTW, I just saw on the news that Peterborough is getting gigabit-speed
> Internet access. (Quite how that's physically plausible I'm not sure,
> but presumably they know what they're on about.)

Is that not just fibre to the house? The obvious next step after fibre 
to the cabinet. My village is looking forward to FTTC in a few months 
and Peterborough is just up the road!

> isn't that fast! The politician was standing there enthusing about how
> this is going to "super-charge local businesses", but I can't think of
> too many businesses where this extra speed will be of any use...

I would imagine this would benefit medium-size companies (around 50-200 
people) that currently cannot justify the cost of a really fast 
connection but due to the number of employees would make good use of 
more bandwidth.

> I guess the problem is that you're generating a lot of light, and then
> trying to selectively absorb the colours you don't want. If you could
> somehow do it the other way around - only generate the optic power you
> actually want in the first place - it could be a lot more efficient.

That's exactly the problem, and also that the light source itself is 
very efficient at converting electricity to heat :-) People have looked 
at using lasers but it doesn't look like that technology has worked out yet.

Post a reply to this message