Comments on: Optimized Sites vs Optimizing Browsers

By: Juhani Polkko

Juhani Polkko — Wed, 08 Feb 2006 16:28:03 +0000

Most mobile application developers are concentrating on advanced platforms as the S60 – Nokia has done a great job promoting it and has awesome developer programs.

But in the U.S. at least, S60 penetrations is something like 0.2%. If you want to reach the masses today (and maybe next 5 years), whether it’s an application or a mobile web site, you have no other choice than to get the crappiest “$0″ phone out there and make your thing work on that.

My bet will be on using just one address and automatically adapting that to the what ever browser or phone you are using. If you go to http://www.ibm.com or http://www.weather.com with Firefox, you get the full HTML site – if you go there with your phone you get the WML site. It doesn’t always work as planned but the point is that the consumer shouldn’t be confused with new URLs when there’s another ways to work it around.

Juhani

By: Martijn Wuite

Martijn Wuite — Tue, 25 Oct 2005 09:50:00 +0000

Exactly, and because there are so much browserstypes out there plus zillions of websites, wouldn’t it be great to have some kind of middleware that would transform any webcontent into the right mobile format using useragent profiling? Well, there is, it’s called Filtering Integration Technology (FIT) and has mobilized services like mobile banking for Deutsche Postbank and Wikipedia. Could this be the solution for the everlasting mobile discussion? I think so…

By: Eugenia

Eugenia — Mon, 01 Aug 2005 05:47:31 +0000

Thing is, the existing browsers must support this feature. And there is another problem with this too: the browser has to first download the normal page and then redirect. This can’t work because many mobile browsers crap out on more than 16 KBs of web pages (images inclujded).

This should have been an http extension instead, and it should have already happened. Now, it’s too late. There are 100 mobile browsers out there that they will never be updated for it.

By: Martin Geddes

Martin Geddes — Sun, 31 Jul 2005 17:46:31 +0000

Can it really be so hard to do some auto-discovery of the mobile URL, a-la RSS feeds? Yes, you might have to pull down a whole home page; but that’s what proxies are for, yes?

By: Eugenia

Eugenia — Sat, 30 Jul 2005 07:07:35 +0000

There is no way this situation can be fixed right now. This was something that should have been prevented by W3C years ago with the inclusion of a mandatory useragent-alike information that the device should send to the client where the resolution of the device and 2-3 more important other things are included in it (no, WURFL is no good as it is today). And then, based on this information the web masters would easily extract it and serve the right pages.

Instead, the way it has to be done today is the web master himself must literally HUNT for mobile user agents and then serve the right mobile pages to these browsers.

At OSNews I spent a few months to find all the major keywords that would prove if a browser is a mobile one or not. Right now, I have support for 120 such browsers. And based on this information, I either serve the mobile html page, the WAP page or the desktop one. All automatically.

However, this was quite some work to not only gather these keywords, but to also test with most of them. There are no emulators for all devices, I had to actually buy a lot of the devices to test my sites on. But it is a work that has been paid off as OSNews serves more about 4,000 mobile pages per day, out of the 275,000 ones it does overall daily.