W3C home > Mailing lists > Public > spec-prod@w3.org > January to March 2016

Re: Proposed Final Design for W3C Technical Reports style in 2016

From: fantasai <fantasai.lists@inkedblade.net>
Date: Mon, 4 Jan 2016 14:45:29 -0500
To: David Carlisle <davidc@nag.co.uk>, Coralie Mercier <coralie@w3.org>, spec-prod@w3.org
Cc: chairs@w3.org, w3c-ac-forum@w3.org
Message-ID: <568ACBD9.9060006@inkedblade.net>
On 11/06/2015 09:01 AM, David Carlisle wrote:
> I have experimented with updating the (xmlspec XSLT generated) XML (and
> HTML) Entities spec [1] to the proposed style.
> [2] is the current editors draft on github and [3] is the same source,
> restyled.
> It all went reasonably well (comments on whether I could have used more
> of the new classes welcome)
> *
> I found the default styling of links (a faint underline, but no change
> to the text formatting) made them somewhat hard to spot, so I added
> an a.loc class that uses the same colour as a.property but without the
> monospace font.

Hi David,
Bert Bos also commented on the link styling here:
and I've updated the style sheet to be more in line with what the
CSSWG currently uses, here:

Please take a look and let me know if this works better for you.
It should be a little less faint. :)

One of the major considerations for link styling was that, in an
aggressively-cross-linked document, a very obvious link styling
pulls at your attention in a way that makes it harder to track
the flow of the sentence itself. In technical documents such as
the W3C specs, you really don't want to be disrupted while you're
trying to unravel a sentence, so we wanted something that would
still appear very obviously as a link when you looked at it, but
that also didn't grab your attention in a way that broke the flow
of the sentence.

(If you want to see how dense things can get in a fully cross-linked
spec paragraph, take a look for example at

> *
> Because of the narrow text width (a good thing:-) the title broke as
>   (3rd
>   edition)
> which is less of a good thing.
> Currently I've replaced spaces by nbsp between the brackets which works
> fine, but might confuse any crawlers expecting the <h1> text to exactly
> match he published title? (I could use a span with nowrap styling for
> the parenthesized text, would that be better?)

I'm only a CSS expert, not an SEO expert. ;) Maybe someone else can
answer this question.

> There are lots of tables in this spec, for the tables such as "epsilon"
> [4] I used class="complex data" which is OK, but I think I really wanted
> left aligned. For the main tables for the entity sets eg [5]
> I've currently left with no css but <table border> as before.
> It would be nice to pull these into the styling (somehow).
> So really just a general comment that table formatting could perhaps do
> with more variant classes.

Well, looking at your table, I think it's not particularly complex
data (in terms of the cells relationships), so you should probably
just use class="data". This looks a bit cleaner imho -- the extra
lines are really only necessary when you have complicated spanning

The second thing I'd suggest is improving the table markup by adding
some <colgroup> markup, e.g. the last three columns seem to belong
to a single group, and encoding this in the markup will make the
table styling highlight this relationship as well. (You might also
consider, e.g. using <th> on the entity name's cell, if you consider
that to be the header of the row.)

Alignment for tabular data is a very specific thing. The CSS specs
have a variety of different types of tables, and although the default
alignments work well for most of them, sometimes I do need to tweak
individual cases.

I'll switch the W3C default to left alignment as you suggest, since
that's closest to "do nothing". In any case, text alignment in tables
is, I think, something that often needs tweaking depending on the
content, and something spec editors should feel free to tweak on top
of the W3C-wide styles.

> This is a fairly simple spec using a more or less unmodified xmlspec
> stylesheet, so if anyone else wants to try an xmlspec based publication,
> feel free to fork [6]

Thank you for your experimentation and your comments! Sorry for the
extraordinarily late-sent reply, it's been sitting in my drafts box
for the last month. :|

Received on Monday, 4 January 2016 19:46:07 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:55:21 UTC