Re: [css3-values] new editor's draft (and [css3-box])

I'm sorry, this was laying around in my drafts folder quite some time.

Håkon Wium Lie:
> Christoph Päper wrote:
>>    um   micrometers (µm); 1um is equal to 1/1000 millimeter.
>>    dd   didot; 1dd is equal to 3/8 millimeter.
>>    cc   cicero; 1cc is equal to 12 didots.
>>    q    kyu; 1q is equal to 1/4 millimeter.
> Cheap to implement, yes. But it will be a painful transition period
> when some browsers support them and others don't.

I don't expect these to be used all that much on the Web (and  
therefore with browsers), but rather with other CSS-enabled software,  
like Prince.

> Do we really need these?

Technically we don't. Technically all we ever needed was a single
physical unit of length.

> When would you use um?

I actually would have used it as the preferred unit in Media Queries'  
resolution-related features (=> 'pixel-size' etc.). You wouldn't have  
needed to introduce reciprocal, ill-named 'dpi' and 'dpcm' then.

Some people prefer integer values, or at least to have more places  
before the decimal separator than after it. With micrometers one can  
do that, for example for line thickness.

Some implementations may choose it as their internal physical base  
unit, although one could also argue for the introduction of 'twip'  
then (0.05pt, "twentieth of a point").

> Who uses dd on anything computer-related?

Where _can_ someone use Didot's point (and cicero) in computer  
programs? Some DTP software probably supports it.
Who actually uses the pica by the way?

Several people were used to the unit and would prefer it over 'pt'  
anytime, not just because it better relates to the metric system (when  
defined as 3/8 mm exactly).

> And kyu?

Same story as for 'dd', but (at least allegedly) for the Japanese  
rather than the continental Europeans.

>>    ln   the line height of the element (...)
> Use case?

Imitate a typewriter-like document, where paragraphs are separated by  
one line gaps, which is not really the same as 1em. Weak, I know.

>>    pel  device pixel
> Use case?

Other people have commented on that already.

>> Do angles need minutes ('min') and seconds (...), do times need  
>> minutes more?
> As in one degree is equal to 60 minutes, and one minute is equal to  
> 60 seconds?

Sure. I didn't think this one through at all. Angles in decimal  
degrees are quite common, unlike decimal time above the second. So if  
minutes would be introduced, they should be used for durations only.

>> | Parentheses, commas, whitespace characters, single quotes (') and
>> | double quotes (") appearing in a URI must be escaped with a
>> | backslash:
>> If I remember the RFCs correctly, of these only commas and perhaps  
>> parentheses may appear in URIs.
> Anyone else knows?

 From RFC 3986:

|  reserved    = gen-delims / sub-delims
|  gen-delims  = ":" / "/" / "?" / "#" / "[" / "]" / "@"
|  sub-delims  = "!" / "$" / "&" / "'" / "(" / ")"
|              / "*" / "+" / "," / ";" / "="

So at least whitespace and double quote should be left to the URI  
layer for escaping.

Received on Thursday, 19 February 2009 07:10:58 UTC