Home Messages Index
[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index

Re: (News) Googlebot Stops Lynxing, Starts Using Mozilla

__/ [ canadafred ] on Thursday 23 February 2006 18:09 \__

> "Roy Schestowitz" <newsgroups@xxxxxxxxxxxxxxx> wrote in message
> news:dtksb9$2v8g$1@xxxxxxxxxxxxxxxxxxxx
>> http://www.adsensebits.com/node/24
>> Google will no longer view our sites as textual fragments, but rather
>> render
>> the pages and interpret them in a richer context (including JS and CSS).
> That's cool Roy,
> I checked out the article. I have a question maybe you could answer. For us
> who are less programming savvy, can you tell us what the implications would
> be with this development? What does this mean to our web sites? How will
> this affect the average "back of the room" code cruncher?
> Thanks

I am flattered that you consider me "programming-savvy", but I am not
proficient or well-oriented in this domain which is browser renderers. I
know a fair bit about KHTML and Gecko, yet I have never used Lynx, only seen
screenshots of it.

If Google decided to render pages graphically, which requires some extra
computational labour, they could initially choose to do it with just a
subsample of popular pages. Alternatively, they could just employ Lynx and
subsequently Mozilla /only/ if the scanned page has changed based on shallow
analysis by Lynx.

So what could they do with the product (assuming that John Bokma's statement
is inaccurate and thing work as the blog suggests)? With tools that are
equivalent to http://khtml2png.sourceforge.net/ , they could make use of
pattern recognition and image analysis programs (also see
http://browsershots.org/). Interpretation of images happens to be my field
of research actually. All sorts of tests could then be run and lead to a
reward or penalty. Running a series of such test, gives a figure of merit,
which may depend on the surfer in question (use parameters like browser,
O/S, screen size, known accessibility issues, etc.). Tests I can think of

* Does the page get rendered gracefully in all browser rendering engines?

* Does the page redirect using JavaScript (like richo.de and bmw.de)?

* How 'pretty' is the page? Cosmetics are a matter of taste and a 'fluffy'
notion, so I doubt it's of any fair or valued use.

* How heavy is the page (including images) and does the size justify the
visual gains and enhancements? Is there a 'nice' combination of colours used
(discerns pricy design work from amateurish DIY)

* Is anything pornographic possibly contained in the page (use a *rough*
scoring mechanism)?

* What screen sizes are properly supported? Should the search engines deliver
different results depending on the perceived support for screen sizes and
other pre-requisites like JavaScript? Could PDA users get different results
pages altogether?

* Arguments similar to the above but in reference to the visually impaired,
the astigmatic. This can currently be done by looking at colour contrast (in
the CSS/Source) and font sizes. What about a special option for the
colour-blind, e.g. "give me no pages with yellow and green on the same

* Okay, enough for now... *smile*

To Webmasters, this also means that more bandwidth will be consumed by the
crawler in question, if it truly exists or ever *will* exist.

With kind regards,


Roy S. Schestowitz      |    Y |-(1^2)|^(1/2)+1 K
http://Schestowitz.com  |    SuSE Linux     |     PGP-Key: 0x74572E8E
  6:25pm  up 6 days  6:44,  8 users,  load average: 0.36, 0.89, 0.91
      http://iuron.com - next generation of search paradigms

[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index