Site Ratings & Reviews - Re: Critique please

Author

Re: Critique please

Toby A Inkster

2007-04-23, 6:16 am

Jim Moe wrote:

> I'm a bit unclear about the rules, if any, for date format. I thought
> that to avoid confusion with US layout that European format used either
> hyphens or dots: 20-04-2007 or 20.04.2007. (I've always preferred the
> unambiguous Apr 4, 2007 or 4-Apr-2007. suitable for English only of
> course. Or 2007-04-20 (sorts nicely).)

Not commenting on the rest, as it's template-related stuff which I'm not
particular interested in at the moment. My current template is just
a temporary measure at the moment. I'm no good at visual design, so it
takes me a long time to come up with something I really like. It will no
doubt be quite some time before I do, and it's unlikely to look much like
the current design!

Regarding date formats, both DD/MM/YYYY and DD-MM-YYYY are common in the
UK. DD.MM.YYYY would be recognised as a date, but is more common on
continental Europe. I'm not awfully bothered about the dates being seen as
ambiguous because the percentage of people who use MM/DD/YYYY is fairly
small (worldwide, well under 10% when I last checked). Those people who
are used to MM/DD/YYYY dates ought to notice at least one of:

* 'lang="en_GB"' on the <html> element;
* DC.language <meta> element is "en_GB";
* HTTP header "Content-Language: en_GB"; and
* Top level domain name is "uk.".

Dates are stored internally as UNIX timestamps. The CMS software then
has 6 basic date formats set up for displaying dates: short, medium and
long formats, each with or without the time. These date formats can be
localised by the publisher, so an American publisher might choose
"MM/DD/YYYY" as the short format without time.

With regards to making it easy to programmatically sort articles, you
should notice that the created, modified and accepted dates are shown as
<meta> elements using Dublin Core and W3C date/time format, e.g. from
http://beta.tobyinkster.co.uk/blog/...f8-validation/:

<meta scheme="W3CDTF"
content="2007-03-21 13:46:20+0000"
name="DC.date.accepted">
<meta scheme="W3CDTF"
content="2007-03-21 13:46:20+0000"
name="DC.date.created">
<meta scheme="W3CDTF"
content="2007-04-08 22:48:38+0100"
name="DC.date.modified">

> Completely trivial fluff: The code uses tabs to indent. It becomes
> rather hard to read as the text quickly disappears off the right.
> Bandwidth reduction measure (1 tab vs 2 spaces)?

Tabs are hardcoded everywhere, so it would be a lot of work to change
them. Besides which, if you've got a decent text editor, you ought to be
able to change the tab width to your liking. I have mine set so that one
tab is equivalent to 4 spaces.

--
Toby A Inkster BSc (Hons) ARCS
http://tobyinkster.co.uk/
Geek of ~ HTML/SQL/Perl/PHP/Python*/Apache/Linux

* = I'm getting there!

Jukka K. Korpela

2007-04-23, 10:15 pm

Scripsit Toby A Inkster:

> Regarding date formats, both DD/MM/YYYY and DD-MM-YYYY are common in
> the UK.

I won't argue about that statement, but I do say that both formats are
manifestly wrong in web authoring. The former gets confused with the
American (US) style, the latter with the ISO format, which was once selected
just because it was estimated that a notation with hyphens was not in use to
a significant extent, making YYYY-MM-DD unambiguous even when the year is
expressed in two digits.

If you need to use a compact, all-digit notation, then YYYY-MM-DD with a
four-digit year is the only sensible alternative in the World Wide Web
context.

But mostly you don't. Unless you tabulate dates, or something like that, you
should use month names or (in English) their three-letter abbreviations.

> I'm not awfully bothered about the dates being
> seen as ambiguous because the percentage of people who use MM/DD/YYYY
> is fairly small (worldwide, well under 10% when I last checked).

That might be only about 600,000,000 people, but I doubt whether your guess
is correct. The American notation is widely known outside the US, too. Even
the governmental statistics office in Finland uses that notation in some of
their statistics, in the Finnish language, violating Finnish standards but
following some "de-facto standards", I guess. The only cure to this mess is
to use _neither_ of the two competing systems (MM/DD/YYYY vs. DD/MM/YYYY).

> Those people who are used to MM/DD/YYYY dates ought to notice at
> least one of:
>
> * 'lang="en_GB"' on the <html> element;
> * DC.language <meta> element is "en_GB";
> * HTTP header "Content-Language: en_GB"; and
> * Top level domain name is "uk.".

Why should they notice that? In terms of HTML specifications, the decisive
declaration is the lang attribute, and your example contains a syntactically
malformed value for it. HTML specifications mandate the use of a hyphen, not
a low line, between subcodes of a language code. Besides, it is widely known
that lang attributes are very often just bogus - some authoring software
spits out lang="en" or something like that, with no regard to the language
actually used and without even informing the poor author.

Moreover, the lang attribute has no visible effect on the page, so how
should the use notice anything? It might affect a speech browser, but then
again, it might not.

> With regards to making it easy to programmatically sort articles, you
> should notice that the created, modified and accepted dates are shown
> as <meta> elements using Dublin Core and W3C date/time format,

If _you_ use software that actually utilizes metadata in Dublin Core format,
fine. But it is questionable whether we should encourage people into
spending their time with such metadata, which is basically write-only as far
as web authoring is considered. That is, people would just waste time in
souping up meta tags that will be ignored by all parties, except perhaps
specialized software - and if you are authoring for specialized software,
make sure you know what you are doing.

> <meta scheme="W3CDTF"
> content="2007-03-21 13:46:20+0000"

Last time I checked, that was an incorrect format, since the W3CDTF
prescribes the use of a version of ISO 8601 that uses "T" and not space " "
as separator between date and time; see
http://www.w3.org/TR/NOTE-datetime
And in practice, specialized software that digs up Dublin Core metadata
might well play by those old rules (rejecting the above data as malfored),
even if newer rules were issued.

--
Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/

Adrienne Boswell

2007-04-23, 10:15 pm

Gazing into my crystal ball I observed Toby A Inkster <usenet200703
@tobyinkster.co.uk> writing in news:9bbuf4-hco.ln1@ophelia.g5n.co.uk:

> I'm no good at visual design, so it
> takes me a long time to come up with something I really like.

Here, here! My therapist said the right side of my brain is about as big as
a peanut.

--
Adrienne Boswell at Home
Arbpen Web Site Design Services
http://www.cavalcade-of-coding.info
Please respond to the group so others can share

Jim Moe

2007-04-23, 10:15 pm

Toby A Inkster wrote:
>
>
> Not commenting on the rest, as it's template-related stuff which I'm not
> particular interested in at the moment. My current template is just
> a temporary measure at the moment. I'm no good at visual design, so it
> takes me a long time to come up with something I really like. It will no
> doubt be quite some time before I do, and it's unlikely to look much like
> the current design!
>
I mostly like your current layout. It's all about the content, not all
the whiz-bangs (so popular because they can be done, not that they should).
>
>
> Tabs are hardcoded everywhere, so it would be a lot of work to change
> them. Besides which, if you've got a decent text editor, you ought to be
> able to change the tab width to your liking. I have mine set so that one
> tab is equivalent to 4 spaces.
>
I normally just use the viewer provided in Firefox which has a fixed tab
width (8 spaces). My programming editor does not suffer from such
restrictions.

--
jmm (hyphen) list (at) sohnen-moe (dot) com
(Remove .AXSPAMGN for email)