In message <Xns9C31D0B6557C3blutofabercom@85.214.105.209> John Blutarsky
<bluto@faber.com> was claimed to have wrote:

>I'm fine with the browser vs. browser aomcparison because it makes sense.
>Saying that a comparison of apps was done is too much of a stretch.

How about searching a 2,000 entry contact list? Even loading 250
contacts at once was a bit much for one of the middle-road iPhone 2.x OS
releases, bring up the "Phone" app and touching Contacts would take
several seconds (For the record, 2.2.1 was much better)

Despite not running identical applications, I was easily able to compare
that the same contact list took the iPhone 3-5 seconds to load, and took
my Palm too-short-a-time-to-guestimate (read: instant) to load the exact
same list. Searching for a unique first name was another easy to
guestimate test, with the iPhone taking a noticeable amount of time, and
the Palm returning an instant result.

If you wanted to compare devices in the "instant" category to see which
is actually faster, you'd increase the number of contacts by an order of
magnitude and judge the results.

This type of benchmarking isn't particularly scientific, but nor are
users when they decide if a device is fast or slow, so the results are
still useful when one or more devices perform noticeably different.

The iPhone 3G S feels much faster. On a 3G running 2.2.1 the Mail app
would stall for 3-5 seconds very regularly, and sometimes as much as 30
seconds. Safari would lock up randomly for 5-10 seconds at a time. The
3G S doesn't suffer from these problems, likely from a combination of a
faster CPU and more RAM (with more RAM being the big difference) -- You
don't need a benchmark to judge the difference, just using the device is
enough.