"Over the past 11 months, people have created more than 9 million My Maps, encompassing a total of 40 million placemarks. That's an impressive 1 new placemark created every second! We never anticipated that people would become so interested in mapmaking, which used to be accessible only to priests, scholars, and academics."
I've made a few maps myself, none worthy of publication. Thinking about how this is probably done, Google gives up a few hundred bytes of disk space (it's not like the actual map data has to be duplicated or anything) and the user gets something that looks really impressive with relatively little effort.
But think of the bandwidth being used by someone who uses the "Random My Map" display (mentioned at the link) as a sort of alternate screen saver. Google has several things like this that can be used endlessly to throw content up on your screen whether you are there or not.
Gadgets such as this seem to say "Go ahead Microsoft, buy Yahoo, go ahead Facebook and dominate social networking (well that seemed to be true until the last month or so), but can you give users sub-second response time and an ever growing storage capacity for things that they can actually use?"
The trick to beating Google (if they don't beat themselves first) is not only capturing click-throughs, but also maintaining an infrastructure capable of keeping up with it all. "Live.Yahoo.com" is anything but lively (not to mention it will be very confusing if the merger with Microsoft goes through) so far and every day I hit popular web sites that either fail to load or load so slowly (usually waiting for a remote ad server to do its part) that I give up the wait. Do the metrics capture these failures or the frustration and resentment they may cause?
In my mainframe days we used to worry about keeping end-users response time below a certain threshold (more than a couple seconds was considered a failure). This wasn't always one company versus another, but also within a single organization, and the issue was that data-entry people and others that got paid to interact with the mainframe would not only be delayed by those pauses during which their keyboards were locked up, but would also lose focus, so that even when they could finally type again their minds would have strayed to something else.
With all our blazingly fast desktop systems does anyone even think of such a thing? Certainly not the vendors I deal with, and not the IT people I keep in touch with. They are resigned, if not happy with the the fact that our screens contain far more graphics than data and no matter how fast our desktop system may be, "painting" that next screen is going to take a while.
Google is the only company that seems to have the infrastructure to come close to the one or two second responses I used to expect, but I wonder if anyone there is keeping an eye on this, or will their be a point at which even they run out of some critical resource and subject us all with interminable "waiting".
No comments:
Post a Comment