|
|
Posted by Jeff on 01/29/08 15:38
Andy Dingley wrote:
> On 28 Jan, 17:56, Jeff <jeff@spam_me_not.com> wrote:
>
>> I had thought that since everything is CMS driven
>
> _Inside_ a CMS, there are strong arguments for using XHTML (or at
> least, some XML that shares the same XML schema)
>
> On publishing from a CMS, it's usually easier to serve the document to
> the web as HTML than it is as XHTML
That has been my thinking. To create it as xhtml and serve it as html.
That leaves me with either fixing the stray bits like <br /> or just
ignoring them as the browsers do anyway. I suppose I should fix them...
(usable XHTML, meeting Appendix C
> and the requirements of good web practice). This is certainly the case
> for XSLT-based output. It's hard to achieve Appendix C from XSLT -
> XSLT wants to serve it as "XML standards compliant" content, which
> isn't appropriate for IE. There's a simple switch to flip it into
> "HTML output", but no similar switch for "Appendix C XHTML".
I'll have to read through that.
>
>
>> that I can just create something like an RSS feed, that would have a bit of html in it
>> (like "strong" or "i" or br) and serve that depending on accept type.
>
> Have you read the infamous article on this, "Myth of RSS version
> comaptibility" from Dive Into Mark? You ought to.
Nope, hadn't seen that. I had noticed when I was learning about RSS that
it was a bit of a mess. That article dates to 2004. If you have a more
current resource on useability I'd like to read it.
>
>
>> Now, it's not hard to generate well formed RSS
>
> There's never anything simple about non-trivial RSS, because RSS 2.0
> doesn't have a competent specification. There's no clear way to embed
> HTML in it. Practical experience favours escaping through entity
> encoding. This is different to using CDATA sections, but similar in
> meaning. Both are a way to embed "<" safely, but both do it by
> embedding "<" as a mere character, stripping away all implication that
> it might be marking the start of a HTML tag.
That seems to be what I've done. It looks like I added a handful of
other elements (like "&"). The RSS reader reconstructs these as what
they are. Frankly, when I wrote my RSS generator I fixed the errors I
found. Perhaps something to do with the widely varying specs I saw and
not really understanding them.
>
> Your RSS reader _might_ later decide to assume that any "words" that
> are "wrapped" in angle brackets should thus be treated as HTML tags.
> This works (and it's how it's done), but it's far from robust. It has
> several drawbacks:
>
> * Such embedded HTML content can't be validated as being valid HTML,
> outside of a final RSS tool that knows about this assumption.
Now, that makes sense!
>
> * How do you publish a HTML tutorial that is marked up in plain text,
> not HTML? What does this mean:
> <item>
> <title>HTML elements Introduction Course</title>
> <description>Today we'll meet the <BR> element!</
> description>
> </item>
>
> * It's hardly rare to use this style of markup in plain text either:
> <description>Set the value of the <customer-identifier>
> field</description>
> In this case, that isn't a HTML tag at all.
>
>
>
>> Will those bits then also have to be in
>> correct xhtml? In other words: <br> or <br />?
>
> Just about the only constant for embedding (X)HTML into RSS is that
> it's not done through XML or XML namespacing. This is for two reasons:
>
> * RSS (2.0 specs) doesn't grok namespacing, as it's not defined to be
> XML (in a compliant sense).
>
> * XML namespacing requires balanced tags and closed elements. This is
> a restrictive thing to impose upon embedding HTML fragments. Consider
> this:
>
> <rss:description>
> <html:p>A honking great list, which we've truncated for display.
> <html:ul>
> <html:li>Foo</html:li>
> <html:li>Bar</html:li>
> <html:li>Bat</html:li>
> [...]
> </rss:description>
>
> Now that's a reasonable fragment to want to embed, but it's
> impractical by namespacing.
Ah, now I understand! Sort of! Fortunately I'm not doing that.
Thanks,
Jeff
[Back to original message]
|