While I agree with you about AdMob, his point isn't that the numbers are bad, it's that the metric is bad. Smartphone sales are not nearly as meaningful a metric as a browser-usage distribution based on measuring page or ad impressions.
Isn't the bigger point to be made is that Internet/Web applications should be platform _independent_?
That is, is there any merit for targeting a single platform -- even if that platform happens to currently enjoy a large market share? I wonder how many people regret targeting IE, exclusively, during the early 2000s when it seemed MS 'owned' the browser market at that time.
There's a difference between IE and webkit though. Webkit thrive on respecting and implementing open and free web standards. Also, it doesn't have any proprietary specific "extensions" and bugs are fixed ASAP. So developing for the iPhone (and therefore webkit) is developing for the open web which is a good thing and allows taking advantage of HTML5 which is the best way to offer great features without ruining your battery (ie instead of using flash).
Note that Symbian and Android both use Webkit, granted with different versions but eventually what works on Android or the iPhone works on Symbian too.
However IE was all the contrary of that: no respect for standards, proprietary extensions, long standing bugs (didn't they just fix a 10 year old one?), security issues, you name it. So let's not compare apples and oranges.
This is a different point, but I'm really glad that Webkit is getting this popular. I loved it back when it was almost just KHTML/Konqueror. Happy it got the recognition it deserved.
This is a long-standing argument of mine. Some of our users maintain that they are only interested in the functionality rather than the quality of the code/design/integration etc.
Then, six months later, when they need an enhancement and find that it is going to take twice as long/cost twice as much to build something on top of the quick & hacky solution that they originally insisted on, they still make unhappy noises.
Exactly, and while it would be great if things were standardized to make this job easier, we have to live in reality and do what is necessary for our users.
One thing that I struggle with to always keep in mind is that our users are not stupid people. I keep the image of my father in my head -- an incredibly brilliant man who taught me everything I know about arithmetic by the time I was 5, algebra by the time I was 8, then stepped back as I voraciously consumed all of the math books he bought for me, thereby basically ensuring my current success in my current field -- and he can't remember how to send links to websites to his friends. Actually, it's not that he can't remember, it's that he can't decide on the best way, between IMing, or posting on Facebook, or emailing (and if he emails, will he email the link or the HTML itself, and then will it be just the HTML, or a web archive, and will the web archive work for his brother who has always ran Apple hardware, etc. etc.).
Our users aren't dumb, for the most part. It's absolutely our failure as an industry to hold each other to a standard of professionalism (and I mean professionalism in the sense of Engineers, not MBAs) to ensure that things are actually usable and adhering to standards. We were cursed by our early adoption by counter culturalists.
From those numbers it appears that the iPhone accounts for merely about half of North American traffic. The point of the original article wasn't that sites shouldn't run well on the iPhone, it's that they shouldn't ONLY be locked into the iPhone.
Those stats seem to back that up. If your going to toss half of the potential user base over the side just to support the iPhone... well that would be a huge mistake from what I can see.
