Hello Robert,
I'm not sure I understand your point(s) from your mail below.
At 17:53 05/02/16, Robert Sayre wrote:
>This is how I generate my Atom 0.3 feed, using the popular Movable Type program: > ><content type="text/html" > mode="escaped" > xml:lang="en" > xml:base="<$MTBlogURL encode_xml="1"$>"> > <$MTEntryBody encode_xml="1"$> ></content>
Not a single namespace declaration here, so it's rather unclear to me what's going on (except if this is an explicitly wrong example, in which case I don't understand why it is here, because we know people get things wrong without needing examples.
>I enter the "EntryBody" in an HTML form, by hand. Then, a Perl script runs through the template I excerpted above, and generates the feed. The template approach is neat, because it lets only slighty technical users modify their content with a lot of flexibility and power.
No problem with that in principle.
>Is this a smart way to generate XML from a purely technical perspective?
As long as it's XML and otherwise conformant, I think it's fine.
>Probably not. Do you and Julian and Anne and Henri approve?
I don't see how I would want to complain about how you generate your stuff, as long as the result is following the specs.
>Probably not. Will tools stop generating feeds this way in the forseeable future? Probably not.
>
>My homepage is valid XHTML[0], as are lots of folks'. They are going to take their PHP, Perl, Python, JavaScript, and Visual Basic, and use the basic string functions of those languages to cobble together a feed. They are probably going to go with "XHTML" for ease. They are not going to get the namespace right at first, so they are going to add a div element around the outside.
All fine. I don't have any problem if they get here. If they add a <div> to get the namespaces stuff right, great. If they figure out that they can use an explicit atom: prefix in the template, and declare the html namespace as default, also great. There are many ways to get namespaces right, even in a simple template/concatenation implementation.
>However, this is not going to reflect what's in MySQL.
Does it need to? Could anybody except for you check? Should anybody (including you) care? Is it part of conformance?
Regards, Martin.
>Robert Sayre
>
>[0] http://validator.w3.org/check?uri=http%3A%2F%2Ffranklinmint.fm
>
