You should either apply your "no bigger than" argument rigorously - and I bet
someone could find a platform somewhere with a narrower screen width than
512 pixels - or apply your "probably used to it" argument to 640 pixels...
Arguing from pixels is a poor choice in any case. Even sticking with computer
screens there is a wide variation from 60 to 100 or more pixels per inch. Other
devices such as dye sub printers have higher densities, and your 470 pixel image
is going to look pretty tiny at 300dpi.
> Or, the browser could scale down images to fit the page it has to work
> with. This is probably a good idea regardless of whether you scroll or
> page through the document.
That is certainly an option, but should be under user control - the document
author should indicate that a particular image is to be scaled to full screen
width, or height. Depending on the aspect ratio of the graphic and the browser,
either width or height might be the limiting factor.
> Inline images are described as non-essential
> decoration,
Where? The HTML2.0 specification makes no such value judgement.
http://www.w3.org/hypertext/WWW/MarkUp/html-spec/html-spec_5.html#SEC62
I suspect you are confusing this with the description of the IMG tag in
HTML 3.0, which is somewhat different as there is also a FIG tag.
Do not make the assumption that all images are trivial decoration. That
may be true in some cases - many cases, even - but is not universal. It
is dangerous to suggest that browsers can play fast and loose with image
data without warning.
> and the style guides I have read suggest using a link to a
> full-size image if the graphics are really essential.
That is one way to do it, and with current browsers is likely to result
in higher image quality when there are inadequate colours. There may be
cases however where it is required to place the graphic in association
with text, or to have selectable hotzones on the graphic, and this it must
be presented inline.
> Therefore I think
> it's perfectly fine for the browser to down-size an image to fit the
> screen.
Only if the author explicitly requests this. Certainly until browsers start
offering user preferences on image resampling. Current browsers resize by
pixel replication which is fast and ugly.
Other options should include bilinear interpolation, bicubic interpolation,
and none.
> If I'm not mistaken, the HTML 3 <IMG> tag will let you give
> attributes that define a size relative to the browser page, as well as
> specifying a size in pixels.
No, both IMG and FIG take optional WIDTH and HEIGHT attributes plus a
UNITS attribute (pixels or ens). No relation to the browser window.
http://www.hpl.hp.co.uk/people/dsr/html/img.html
http://www.hpl.hp.co.uk/people/dsr/html/figures.html
However, HTML 3.0 is a draft spec and subject to change.
-- Chris Lilley, Technical Author +-------------------------------------------------------------------+ | Manchester and North HPC Training & Education Centre | +-------------------------------------------------------------------+ | Computer Graphics Unit, Email: Chris.Lilley@mcc.ac.uk | | Manchester Computing Centre, Voice: +44 161 275 6045 | | Oxford Road, Manchester, UK. Fax: +44 161 275 6040 | | M13 9PL BioMOO: ChrisL | | URI: http://info.mcc.ac.uk/CGU/staff/lilley/lilley.html | +-------------------------------------------------------------------+ | "The first W in WWW will not wait." François Yergeau | +-------------------------------------------------------------------+