[Air-l] Web 2.0
Janna Anderson
andersj at elon.edu
Thu Feb 15 15:01:52 PST 2007
http://www.iiszone.ziffdavis.com/article/Web+20+Off+the+PC/197323_1.aspx
Web 2.0 Off the PC
DATE: 22-DEC-2006
By IIS Zone Features Staff
Web 2.0 can be a slippery subject. Not everyone agrees on exactly what it
means, leaving the term open to gratuitous marketing use. But it does mean
something, and that something is related to the richness of Web applications
that make data more accessible and useful.
One way to look at it, as a practical matter, is that it makes Web
applications more like dedicated local computer applications, rather than
having to deal with data the way Web applications traditionally did so, with
poor user interactivity and frequent round-trips for data operations.
Now with the cutting-edge of application development on the mobile device,
developers and service providers are dealing with some tough questions, such
as how well Web 2.0-style applications work on mobile devices. Some have
been arguing that simple apping of the PC style of Web 2.0 to mobile
handsets is doomed to failure, and there is much to support that argument.
One of the hallmarks of these newer, richer apps is that they push data
liberally out to the client (generally in XML formats) in order for the
client to be able to display more to the user and allow rich editing
techniques. On a handset, the memory and bandwidth this entails are
expensive, both in their scarcity relative to a PC and in the battery life
they consume.
Latency is another problem that may affect mobile applications in a more
profound way than on a PC, which typically has a much faster and cheaper
connection. Building up and tearing down connections is a frequent operation
in PC-based Web apps, even in Web 2.0. But in a mobile environment, it can
cause large bottlenecks since the connection overhead is much greater. And
of course these last two problems work at cross-purposes to each other.
The solution to the latency issue is usually in more aggressive and
intelligent pre-caching of data, which aggravates the memory and bandwidth
problem. The solutions are found in good engineering of trade-offs and in
advances, both in the handsets and the networks.
These trade-offs also put a burden on the back-end applications which, in
order to deal with different types of devices and connections optimally,
need to grade the quality of data and perhaps even the number of features in
apps in order to keep the user experience pleasant and the application
affordable.
One approach that is used by many apps is the following. Rather than use the
standard browser interfaces, create custom-client applications that manage
the data more effectively, perhaps with custom compression.
But custom-client apps are an approach well worth avoiding if possible.
Users and administrators are leery of having to manage multiple client
applications, and of the potential for security problems they present. The
added burden to developers in having to create a custom app or perhaps
several versions of a custom app for different types of devices also won't
sit well with developers.
There's no way around the mobile handset being a harder environment to write
for, especially for the data-rich applications of Web 2.0. But developers
will still flock to it because of the advantages it presents, such as the
ability to use GPS data from the device. And the more the user relies on the
application, on or off the PC, the more valuable the application is.
In the end, it's likely that the technological advances in the power of
handsets, power management, bandwidth, and the intelligence of networks will
make these problems obsolete, just as they largely obsoleted the need for PC
application developers to do their own memory and device management.
And even in the interim, as we build these networks and applications into
what we expect in the long term, they are gaining wild popularity. Nothing
speaks better to their value.
On 2/15/07 4:15 PM, "elw at stderr.org" <elw at stderr.org> wrote:
>
>
>> Uh, this will sound naive, but is Web 2.0 actually using larger system
>> implentations Internet2 (IPv6) or is it traveling on the same IPv4
>> platform as the WWW today?
>
> IPv4.
>
> I don't know of any non-experimental hosts that are reachable by IPv6.
>
> If someone anecdotally knows of a 'biggie' that is routing/advertising
> IPv6 services, I'd love to hear about it.
>
> --e
> _______________________________________________
> The air-l at listserv.aoir.org mailing list
> is provided by the Association of Internet Researchers http://aoir.org
> Subscribe, change options or unsubscribe at:
> http://listserv.aoir.org/listinfo.cgi/air-l-aoir.org
>
> Join the Association of Internet Researchers:
> http://www.aoir.org/
--
Janna Quitney Anderson
Assistant Professor of Communications
Director of Internet Projects
School of Communications
Elon University
andersj at elon.edu
(336) 278-5733 (o)
(336) 446-0486 (h)
More information about the Air-L
mailing list