(no subject)

[noahgibbs]

This week, like most weeks, Robert X. Cringely has written a great article. This week's article is about upcoming changes in the video game market, the downloadable media market (think downloadable music and videos), and some pie-in-the-sky speculation about where Google could go next.

I figure Bob's hitting his usual accuracy of 60% to 70% - I have to agree with him on the first two, but I don't think that's the direction that Google's headed. He's right that there are some veeeery interesting options in that general direction, though...

Comments

[r-transpose-p.livejournal.com]

That "google accelerator" thing sounds like they're just going to have some app that replaces every URL you request for the version of the same URL that requests Google's cache of the page.

In that light, the author appears to be confusing "offloading processing from your computer to google" with "ammortizing costs of traffic to slow servers".

His comments on Apple and Microsoft do seem a bit less off-the-mark.

Perhaps it is a good time to be in graphics/multimedia.

[angelbob.livejournal.com]

One thing to consider about Google caching and the approach you talk about is that it *can* offload processing from your computer in the case of, say, PDF documents (which Google converts to HTML, badly, but it's still way faster than Acrobat) or, as he suggests, some Flash stuff. However, Google is more likely in that case to be taking load off the server, as you say, rather than the client.

Most of the effort on the client is spent in rendering content to screen, and while Google could conceivably accelerate that in a couple of ways, I don't think they're doing so.

Incidentally - Google *is* offloading some processing from the client, conceptually, which the client is not yet doing for itself. Specifically, Google is analyzing the sent web pages to figure out where the client is most likely to go next (what pages it will mostly likely next request), which a good web browser should do for you, but none (that I know of) do.

Granted, it's probably analyzing the web pages in a really simple way and a good proxy server might do the same thing, but still.