On 12/27/09 11:04 AM, Felix Miata wrote:
> On 2009/12/27 18:06 (GMT) MEM composed:
>
>> Sorry to bother again with this issue...
>
>> If we stay with 1em, all fonts size will be equal. However, if we
>> do less then 1em, (no matter what value I suppose) we just get
>> inconsistent results. I know the browsers should not look exactly
>> equal, that's the own nature of browsers but, at I would like them
>> to be, at least, similar. Right now, however, the results are
>> quite different.
>
>> Any advice regarding this issue?
>> http://www.chequedejeuner.nuvemk.com/educa.infantil/associar
>
> Look at http://fm.no-ip.com/auth/Font/font-rounding.html in each of
> your installed browsers and you may see that more than two
> significant digits to em font-sizes are an invitation for
> inconsistent results. If you stick to two significant digits and
> take what you learn from that URL into account you might get
> acceptable results cross-browser.
>
[...]

Further to Felix's good advice to stick to 2 decimal places in
"em" sizes (so no fractions of 1%)...

Some time ago, I ran a series of automated tests on font scaling,
testing on all the browsers I had at the time. As you discovered, each
browser engine has its own way of rounding fractions. Webkit, for
example, simply ignores any fraction of 1%, while Gecko not only rounds
fractions, but also seems to carry fractions forward, rounding
accumulated values.

I find I get the most consistent results from using these percentages,
which scale quite well. For example, if font-size of 82% gives 13px,
then 5em will very likely give 65px in all browsers.

8pt brevier             11px    69%
9pt bourgeois           12px    75%
10pt long primer        13px    82%
11pt small pica         15px    94%
12pt pica               16px    100%
14pt english            19px    119%
16pt columbian          21px    132%
18pt great primer       24px    150%

(The pixel sizes will vary according to computer settings.)

Hope this helps.
Cordially,
David
--

______________________________________________________________________
css-discuss [[email protected]]
http://www.css-discuss.org/mailman/listinfo/css-d
List wiki/FAQ -- http://css-discuss.incutio.com/
List policies -- http://css-discuss.org/policies.html
Supported by evolt.org -- http://www.evolt.org/help_support_evolt/

Reply via email to