Probs with background color

[Bill Crosbie]

At 04:18 PM 12/21/95 -0800, Joe Schallan wrote:
>A few users have reported that when they view my library's
>web pages, they see the background as a weird sort
>of "tartan" pattern that renders my text virtually unreadable.
>I duplicated the effect on their equipment (Windows PC,
>Color Monitor, Netscape 1.1).  The background color
>is _supposed_  to be #E9C2A6, a nice, easy-on-the-eye
>solid.  When I view the pages on my own equipment
>(Windows PC, Color Monitor, Netscape 1.2N), it
>looks like it is intended to look.  Same with _most_
>other users.

I'll take a stab at this, although I must confess to just having a 'best
guess' to what is happening.

PCs are problematic in that they can have any of a different number of video
cards in them, and it seems they each have their quirks in mapping color.
Suppose a user has the card configured for 256 colors.  Also suppose that
you have 6 graphics of 50 unique colors to display on a page.  This means
that there are 300 unique colors that must be mapped into 256 available
colors.  

It is my guess that the browser begins handing the color information to MS
Windows and windows hands this info to the video card.  Some where along the
way the card 'sees' more colors than it can display and begins the process
of optimizing the pallete for this display.  This means choosing color
substitutions in such a way that it can produce patterns with the chosen
colors to emulate the missing colors.  In a small graphic this might be ok.
The dithering will hardly be noticable.  But on the broad expanse of your
page background the patterns are distracting and 'tartan-like'.

You might want to check out the HTML Sourcebook, by Ian S. Graham for a
little more information.  While there is not much more on color, it is still
an excellent resource.