2008/6/12 Dave Pawson <>:
> 2008/6/12 Peter Dolding <>:
>> Location and size perfect scaled up
>> or down to page should be 100 percent acquirable without question.
> Agree, but only when constrained by a 'profile' (if that is what we
> are going to call different use cases).
I submit that this may presently be the 95% case of present usage of
e.g. an ODF text or presentation document :-)
>> In a desktop environment, where printing is used. If
>> same document is printed A4 and A5 and user wishes the A5 document
>> should only be a smaller form of the A4 document. Scalable formating.
> -1. Scaling and 'location and size' need to be described better.
> Are we talking about a printed output from an application, or screen
> positioning?
> Using printed output is far easier (though yet again it is a manual test, not
> automatable)
> Take a word processor instance. Print it out using A4 setting. Measure
> position of (some part of the content). Should be x,y mm +- z mm wrt top left.
> That's a usable metric. Is that a fair definition of 'pixel perfect' for this
> profile?
Line and character placement. The WWW Acid3 test turns this into
"pixel perfect" by making the test download given test fonts as well -
if an ODF is layout-perfect, then with the same font it *should* be
pixel-perfect (unless I've missed something).
> Perhaps users want scaling too. I don't know.
Layout-perfect should solve that.
> Now we have a profiled description of 'pixel perfect'.
> Anyone not OK with that?
Or "layout-perfect". As I say, if a standard test font (what's free
enough? I suppose a test font can be created ... Latin, CJK, symbols
...) can turn that into "pixel-perfect", then that's very good indeed.
- d.
