Linked by David Adams on Tue 25th May 2010 04:07 UTC
PDAs, Cellphones, Wireless Over at Daringfireball this past weekend, John Gruber put words to what many people are thinking about after Google's rush of Android announcements and not-subtle Apple-bashing at this week's I/O conference: "all-out war." I agree with Gruber that a good old-fashioned bitter rivalry could be a great thing for the computing world, and for smartphone/handheld fans in particular.
Thread beginning with comment 426447
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE: Really?
by WaltFrench on Tue 25th May 2010 15:32 UTC in reply to "Really?"
WaltFrench
Member since:
2010-05-25

Amen.

No part of the original post more clearly proves the maxim, “in war, the first casualty is truth.”

I'm a fan of strong competition and Google has made some incredible progress in making a *nix-based smartphone the mainstream approach. Incredible.

But the presentation was embarrassingly sophomoric, and deeply dishonest.

Google has claimed that they did the Android thing “to prevent one company, one man” from controlling the internet. Who, in 2005, when they bought Android, would that have been? Google has a clear and legitimate business purpose in increasing the ability of consumers to see their ads; why lie about it?

The speed comparisons were downright juvenile. Android seems to be a fine platform now that their Just-In-Time runtime takes away the sharp disadvantage they had faced with their current development tools. But the only real advantage over the iPhone 3GS is the clock speed of the chip in their most expensive phones. Earlier phones are now only up to competitive, and if Apple indeed announces their HD model in less than 2 weeks with a 1GHz chip, the crown will be Google's for an embarrassingly short time. I didn't see a direct lie, but the demo was deeply misleading, and I wouldn't have wanted to have made it.

My biggest laugh, however, was their use of the “Linpack” benchmark to show how much faster they were. I actually have coded and use this “Gaussian elimination with partial pivoting” routine in my statistical investment programs, but I can't imagine a *single* smartphone app that would have that routine in it. (I'd love to hear that advanced stat stuff is in use in cellphones, but can't imagine how 0.01% of the population cares.) Linpack does math that's *like* JPEG and MP3 manipulation, but decent performance on smartphones requires those routines to be very carefully hand-coded into the system or hardware. The Linpack benchmark doesn't even address the stuff like it that *does* matter.

Again, Google has some neat directions and advantages but they show their immaturity and desperation by flouting irrelevant comparisons.

Reply Parent Score: 2