At 3:06 AM -0700 on 2/28/00, [EMAIL PROTECTED] wrote:
>I'm going to have to continue my UI paper after hearing this. Drat. Was
>hoping to avoid it...
>
>--PLEASE tell me what you want then - you would be far more likely to get it
>because I cannot read minds.
That's what my UI paper is. It's rather long, already -- and nowhere near done.
>--Very Clever. Still does not get around the fact that the buttons will either
>have to be over sized to contain the text (not a serious problem but not
>aesthetically pleasing) or ... variable geometry... I think doing a variable
>geometry on the buttons would slow down the interface though I could be
>wrong.
No, it does get around that. You use simularly-size fonts on each
platform. If you need to use 11pt on on platform and 12pt on another,
fine.
>--In principal, I agree that a solution would be to test for the platform and
>then set the button fonts appropriately.
>--I shall experiment with your idea which I do like
>
>
>
>
>Then, let's not do our sizes in some silly unit. We'll do them in
>PostScript points (like the Mac) and do any necissary conversions. Or
>maybe we should store them in something even more universal:
>milimeters. We also have size names, which can be either "relative" or
>"absolute". Relative means that it is based on the intended use (e.g.,
>all the tv ones would be much bigger; print ones smaller) or absolute
>(it's 14pt everywhere). These could be:
>
>--Scott pretty clearly wants 14 point. I favor 12.
And I favor 12, 10, or 9, depending on the font.
The relative ones take into account the relative size of a font; Times
9 is a lot smaller than MPW 9. Courier is smaller than Monaco. Geneva
is bigger than just about anything, etc.
They can also take into account user preference -- a farsighted use may
use Geneva 24, while you(?) and I happily stare at Geneva 12 (and, in
my case, think 'dang that's huge type!').
>--Finally one potential difficulty: Some one who has access to Unix, Windows
>and Mac has to verify that the thing actually looks good when finished.
I can verify it on Linux/PPC and probably Winhell '95.
