- From: James Aylett <dj@insigma.com>
- Date: Tue, 8 Sep 1998 04:57:43 -0400 (EDT)
- To: Murray Macdonald <murray@mha.ca>
- cc: "www-html@w3.org" <www-html@w3.org>
On Mon, 7 Sep 1998, Murray Macdonald wrote: > > the imagemap spec is complete. > > WRONG. The spec does not specify legal ranges for "coordinates" nor > does it specify how many "coordinates" a polygon may have. > > To refresh your memory here is what the HTML4.0 spec says for the <AREA> > tag: [snip] > coords = coordinates [CN] > > Coordinates are relative to the top, left corner of the object. All > values are lengths. All values are separated by commas. > > If you follow up the definition of "lengths" it tells you there are > pixel or percentage definitions of size. Again this is ambiguous. I would always take a length to mean the magnitude of a displacement (meaning it can't be negative). However I agree that it isn't defined here (and would probably argue that it should be, for convenience). > Presumably I can create circles centered on the map that have a radius > larger than the map. What if the center of a circle area was located at > 1,1 and the radius was 100. Even though some of this circle's area was > off the page, you would expect this to work, right? What if the > circle's center was moved 10 pixels to the left? Then the circle would > be centered at a pixel location of -9,1. Are you suggesting that at > this point it should stop being a circle and be replotted as a polygon? > This seems silly and the html 4.0 specification does not state this. It > is simply ambiguous. That's a misleading and confused statement (the sentence starting "Are you suggesting ..."). The circle will always be plotted as a polygon - plotting is the job of the UA. The definition of the circle, to guarantee that it works (given the ambiguity of the spec at present) would have to change to be defined as a polygon at this point. I *know*, before anyone starts shouting, that I'm splitting hairs about English usage here, but in a discussion where people seem to be misunderstanding each other, it seems appropriate to at least attempt to avoid this sort of thing. (And I've gone with what I see as sensible word usages, in an attempt to avoid ambiguity, again.) Personally, I would say that: you can't assume that negative numbers will work across all UAs; however that negative numbers probably should be specifically allowed in a future HTML spec. A sensible bound on the magnitude of the coordinates would seem to be the size of the image concerned - that *should* work on all browsers which can display the images (and if they can't, they have ALT text, so the issue doesn't really apply). A question which is perhaps relevant here is: what are the bounds on the sizes of images using the current popular file formats? J -- /--------------------------------------------------------------------------\ James Aylett, dj@insigma.com Insigma Technologies Ltd Tel: +44 (0)1285 643100 Norcote Barn Norcote Fax: +44 (0)1285 643600 Cirencester GL7 5RH
Received on Tuesday, 8 September 1998 04:59:48 UTC