W3C home > Mailing lists > Public > www-style@w3.org > October 2010

Re: [css3-values] [css3-images] no units for default 'image-resolution'

From: Shelby Moore <shelby@coolpage.com>
Date: Tue, 26 Oct 2010 18:25:37 -0400
Message-ID: <e0f878aab3f078a9fa6041e3384411ae.squirrel@sm.webmail.pair.com>
To: shelby@coolpage.com
Cc: "Brad Kemper" <brad.kemper@gmail.com>, "Tab Atkins Jr." <jackalmage@gmail.com>, "www-style@w3.org" <www-style@w3.org>
In reply to myself:
> Brad Kemper wrote:
>> On Oct 26, 2010, at 1:30 PM, "Shelby Moore" <shelby@coolpage.com> wrote:
>>> DPI is an ambiguous term and is discouraged:
>>> http://en.wikipedia.org/wiki/Dots_per_inch
>>> It is advised instead to use PPI.
>>> Thus I suggest you change 'dppx' to 'pppx'.
>> It is only ambiguous if we don't define it well, since CSS properties
>> get
>> their definitions from the module specs, not from wikipedia. That said,
>> we
>> try to use familiar and intuitive words for the property names, as much
>> as
>> we can.
> Afaics, established definitions play a larger role in what it means to be
> familiar and intuitive.
> Also, afaik CSS is designed for a broad audience, so unnecessarily
> creating new vocabulary does not meet the 'amatuer user requirement',
> which afaik CSS is subject to.
>> 'pppx' is not especially familiar and intuitive,
> Neither is dppx.
>> and expanding that out in
>> my mind to "pixels per pixel" just makes it seem weird and confusing.
> Pixels per pixel is the exact meaning we desire.
> Whereas, "dots per pixel" is not the meaning we want, because 'dots' are
> ambiguous per the historic context provided by the Wikipedia link above.
>> "Dpi", on the other hand, is familiar to designers, especially those
>> with
>> a print background like myself.
> Including me, as I wrote TurboJet back in the 1980s which was one of the
> first printer drivers for laser printers on AtariST to take advantage of
> HP's RLE capability to reduce the amount of data sent over the slow serial
> ports of that era.
> So I am very familiar with "dots per inch" and those are device dots, not
> image dots. And if you read carefully the Wikipedia article, it gets into
> the issues about halftoning and dot sizes.
> You just don't want to go there.
>> This leads to "dppx" to be pretty easily
>> intuited for designers and other authors working with CSS px.
> Not for anyone who understands what dots really were historically.
> It is sloppy.
>>> Also 'image resolution' is ambiguous:
>>> http://en.wikipedia.org/wiki/Image_resolution#Pixel_resolution
>>> I think you need to change this to 'image-spatial-resolution'.
>> Brevity of property names is another important consideration when
>> picking
>> a name for a property, not just accuracy.
> 'imaging-scale'
> The term 'resolution' is ambiguous, because it also means the pixel
> dimensions of the image.

Google "pixel versus dot"

For one thing, pixels "adjoin" each other (even if they really overlap
with a quadratic sampling), whereas dots typically have gaps. Afaik, that
is why we use "dots" to describe things that devices make and we use
pixels to describe the continuous integration of analog samples.  "Dots"
are often not even perfectly round, nor consistently shaped (e.g. on rough
Received on Tuesday, 26 October 2010 22:26:11 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:38:40 UTC