Transcript The Mobile Web

The Mobile Web
Reasons why it sucks
Pin-Chin Huang
Columbia University
COMS E6125 WHIM
Introduction

The Mobile Web
The World Wide Web as accessed from mobile devices such
as cell phones and PDAs

Reports in 2007
More than 60% of U.S. wireless users own an Internetenabled mobile device, but only 5% of them actually use the
mobile Internet

Why?
Mobile devices, wireless carriers, and the Web itself.
WHIM Spring 2008
Mobile Devices: Problems

Screen Size
Users want a smaller device, but a larger screen

Processor & Memory
Need beefy hardware to render web pages: CSS stylesheets,
DOM manipulation, multimedia, client-side scripting, etc.

Input Capabilities
Numerical keypad not enough for E-mail or IM applications
WHIM Spring 2008
Mobile Devices: Problems

Browser Behaviors
Different browsers (Opera, Nokia, Openwave, Pocket IE, NetFront,
Minimo, Safari) have different support for XHTML, CSS, JavaScript,
Flash, frames, background images, SSR, max page size, etc.
CNN.com
MSDN.com

Cost
Smart phones/PDAs typically cost $400 – $800
WHIM Spring 2008
Mobile Devices: Possible Solutions

Rollable Display

Handheld Projector
Scheduled to commercialize in 2008

Boost processor speed & memory
But what about heat dissipation?

QWERTY keypad

Wait for someone to unify standards…
WHIM Spring 2008
Wireless Networks: Problems

Slow

WAP = Wait And Pay; GPRS/EDGE = 10 to 20 KB/s at best

3G: theoretical rate 1Mbps, but often falls short of expectations

Scott Karp describes his 3G experience:
“It would have been faster, literally, to call my wife at home and ask her to look it
up on a real broadband connection.”

Expensive
Pay per kilobyte  but people are used to unlimited surfing
Monthly Charge
Data Limit
Additional Charge per KB
Charge for d/ling a 4MB mp3
AT&T Wireless
$34.99
20 MB
$0.004
$16
AT&T Wireless
$60
Unlimited
$0
Free
Time Warner
$29.95
Unlimited
$0
Free
WHIM Spring 2008
Wireless Networks: Possible Solutions


WiMAX 802.16

Practical rate 10 Mbps  2x faster than cable, 6x faster than DSL

Sprint Nextel deploying nationwide network in 2008

Intel’s 5th generation Centrino and Nokia’s new handsets to support WiMAX in
2008 as well
3GPP Long Term Evolution

Goals: 172.8 Mbps downlink, 86.4 Mbps uplink
First testing successful in November 2007 (by Nokia)
AT&T and Verizon Wireless planning to migrate toward LTE

Initial deployment planned in 2010



Cost?
Depends on the wireless carriers…
WHIM Spring 2008
The Web: Problems



Web Presentation

Sites not formatted for small screens  endless scrolling & zooming

“So, great, you can jam the entire NYtimes.com onto the screen – got a
magnifying glass?”
Device Fragmentation

H/W diversity: screen size, processor, keypad, stylus, etc.

S/W diversity: OS platforms, browsers, APIs
User Context

Most websites are not context/location-aware

“When I visit Costco on a mobile device, what are the chances I’m going to
shop? I want store locations, hours, and phone numbers, and I don’t want to
stumble through four slow-loading screens to get that info.”
WHIM Spring 2008
W3C One Web Principle




“Making the same information and services available to users
irrespective of the device they are using.”
Do Nothing

Pray SSR renders well

Rely on HTML transcoders
Use Raw HTML

Remove styling and images

IYHY.com provide raw rendering service
Develop Handheld CSS Stylesheets

nytimes.com viewed with IYHY
develop a CSS (Cascading Style Sheet) for each targeted screen size
WHIM Spring 2008
Multi-Web Practice


“Create associated web pages that will give good
user experiences in several of the most popular
devices.”
.mobi Domain
Top-level domain for mobile-friendly sites
 Backed by Google, Microsoft, Nokia, Samsung, Vodafone


Alternative URIs
Users are redirected to a URI appropriate to their device
 Requires browser sniffing

WHIM Spring 2008
I Would Choose…

Multi-Web Practice
Web presentation problem is eliminated
 Device fragmentation can be addressed
 Specific needs of mobile users can be met (user context)


Alternative URIs
.mobi breaks URI address space  unnecessary
 Not every company can get their .mobi domain name
 E.g. nytimes.mobi is taken by someone else

WHIM Spring 2008
Conclusion

Facts
Mobile devices becoming faster
 New display technologies coming to market
 4G networks being deployed


To make m-web “less sucky” we should…
Unify mobile web standards
 Lower price for accessing wireless data networks
 Adopt multi-web practice
 Avoid breaking address space by adopting alternative URIs

WHIM Spring 2008
Conclusion

Mobile devices outnumbers PCs
 20:1 by worldwide number

Mobile web poised to replicate the success of
Internet
WHIM Spring 2008