Re: Internationalization--Metric Typography

From: Asmus Freytag (
Date: Tue Jun 01 1999 - 18:39:03 EDT

Nelson is correct. The scaling on paper is universal (that'w where you
apply your ruler Markus) but the scaling on screen devices is not. This has
nothing, but absolutely nothing do to with how you mark your ruler, but
everything to do with tradeoffs between screen resolution and assumed
viewing distance.

Windows can drive most displays with several different resolutions,
allowing you the tradeoff between seeing next to nothing, but very well
defined and seeing a lot, but not very clealy...

There's nothing misconfigured about any of these choices, as long as
software plays along and lets you set a comfortable viewing scale for the
task at hand,
without affecting the printing scale.


At 03:40 PM 5/31/99 -0700, Markus Kuhn wrote:
>Nelson Minar wrote on 1999-05-30 13:22 UTC:
>> There's a Web disaster in using points in web page styles - Windows
>> assumes 96dpi screens, X and Macs assume closer to 75. So a 10 point
>> font looks fine on Windows, and tiny on X. See
>> Using millimeters wouldn't fix this, though.
>I do think that millimeters are at least a bit more likely to fix it,
>because millimeters are a *much* more accessible unit than points. Most
>people have millimeter rulers somewhere near their workstation, but very
>few have a point reference. (And those who do call themself
>"typographers" and conduct fancy magic dances after midnight, singing
>"pica-em-quad-en-dpi-didot" towards the full moon). I believe users are
>more likely to note that their screen size parameters are misconfigured
>in their GUI environment when a more accessible unit (mm) is used
>everywhere and will therefore fix it more likely. I have configured my X
>server such that the screen size displayed by xdpyinfo actually matches
>my real monitor. The calculation involved in doing this was made
>unnecessarily more complicated by the fact that X servers expect it to
>be provided in dpi. Dpi are an awful unit for two reasons: A) they are
>not metric, and B) reciprocal length (one over inch) is quite
>counter-intuitive for less numerate people, especially when unit
>conversion is involved. The screen size (or pixel size) in millimeters
>(or micrometers) would have been the obvious parameter, but as I noted
>before, when units are concerned, US software developers seem to have
>some strange traditionalistic mental block for seeing the obvious
>Computers should adobt to people, and people should not have to adopt to
>archaic lead-typesetting measurements such as the point.
>(who formats his PhD thesis in Computer Modern Roman 4.5 mm)
>- (Norm-Entwurf) DIN 16507-2, Ausgabe: 1998-04, Drucktechnik -
> Schriftgrößen - Teil 2: Digitaler Satz und verwandte Techniken
> (German draft standard on metric typography)
> <>
>- Otl Aicher: Typographie. 2. Aufl, 1989. (bilingual: German/English)
>- Friedrich L. Bauer: Der typographische Punkt. Informatik Spektrum,
> Vol. 22, No. 1, pp. 41-42, February 1999. (German)
>Markus G. Kuhn, Computer Laboratory, University of Cambridge, UK
>Email: mkuhn at, WWW: <>

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:46 EDT