Why the web is often woeful29 December 2004
This time last year I made an early resolution for 2004 and announced that it would be the year I break my addiction to Google and improve the quality of my searching.
I did not manage it.
Even though I am now using Firefox as my main web browser and can tell its neat little search box to use any of the major sites, it is still set to search Google.
If there has been a change, it is that I no longer settle for Google's results.
Unless I am looking for a straightforward piece of factual information - yesterday Max wanted to know what the five tastes were and could only remember sweet, salt, sour and bitter - I always search on two or three sites.
And I am making a lot more use of specific searches on places like Wikipedia and subscription database services.
But it has proven harder than I expected it would be to stop going straight to Google when I need to find something out, and I think it is time to admit defeat.
Directory enquiries
Google is, as sites like Searchenginewatch have been saying for some time, the best search engine.
The trouble is, it is still not very good.
For example, it indexes pages on the basis of the text they contain instead of using tags or metadata, so it cannot tell the difference between 'die' and 'die', even though the second one is the German word for 'the'.
Without that crucial piece of metadata, there is no difference. But a human reader would spot that 'blue die' and 'Die Welt' are not the same even without any help.
This is not entirely Google's fault: it is a problem that all the major search engines share because they are all trying to index a world wide web that has little or no useful metadata to start with.
When was the last time you built a web page with comprehensive metatags to help searchers? I know, because I am as guilty as everyone else, it is simply too much hassle with today's tools on today's web.
On the Flickr photo sharing site everyone who uploads pictures adds some data, including date, title and a classification, and that makes searching possible.
But on the majority of web pages there is nothing to help a poor search engine find its way around.
The semantic web, if it ever comes to anything, will resolve the problem by helping website creators provide a context for the material they publish, turning the web into a database that can be efficiently searched rather than today's massive collection of linked documents.
Unfortunately its development is proceeding painfully slowly, and the chances of tens of millions of existing sites being upgraded are clearly slim.
Ancient ways
The web and search are not the only examples of technologies that could be so much better than they are now.
Far too much of the internet is still using IPv4, a 20-year old networking technology, instead of the far superior IPv6.
And of course we are still using a keyboard layout that was designed to spread out commonly used keys and stop the hammers of a typewriter hitting each other when two letters were typed together.
The Sholes, or Qwerty, keyboard, solved a problem that no longer exists,
The problem lies in the installed base, the people who are already using the programs, websites, keyboards or whatever other technology.
They have found ways of working, invested in hardware or software, and built systems, working practices and businesses around tools that work in particular ways.
And they are not going to change.
Microsoft gets into trouble if it releases a new version of any program that is not "backwards compatible".
Apple gets complaints if old software will not run on new versions of Mac OS. And any change to core web protocols that forced people to upgrade their networking software would be resisted.
Change for good
In the UK we cannot even agree when we will turn off analogue TV broadcasting in favour of digital because there might be people who will not have digital-capable sets. So how are we ever going to make major changes to network infrastructure?
Change is possible, but rare.
When Google launched late in 1998 many searchers - me included - abandoned Alta Vista, Lycos and Yahoo without a second thought. But that was only a matter of changing website and did not even involve installing a new program.
The developers of the Firefox browser claim that more than 12 million people have downloaded it in a matter of months.
This may indeed show that lots of web users are unhappy with the poor interface and security problems found in Microsoft's Internet Explorer, but Firefox does only represent 1.5% of the net's 800 million or so users.
They just happen to be an outspoken and wealthy 1.5%, able to afford a two-page ad in the New York Times.
Too much of today's wired world is based on old computer science and technical compromises that are no longer necessary, as we can see in the core network protocols that underpin the web.
But if we cannot find a way to make revolutionary changes, to put our current practices to one side and rethink our relationship to our computers, then we will be stuck with systems that are simply incapable of doing the things we want them to.
Source: BBC News
All trademarks and copyrighted information contained herein are the property of their respective owners.
Related Computer Hardware Articles
|
