Recommended reading: How to Write a Git Commit Message
This directory contains three files, which show what can be achieved when pasting the raw HTML into the web page editor of the DVB website's CMS:
featuretest.xhtml
is a sample document trying to exercise the basic layout features for static documentsfeaturetest.pdf
shows howfeaturetest.xhtml
is rendered in my local browser (Opera)featuretest-results.pdf
shows what the DVB website CMS makes offeaturetest.xhtml
It seems that:
- normal flow text and outline levels works as expected
- but line breaks are significant (this means that lengthy text
content, such as e.g. in a
<p>
or<td>
element, must not contain any line breaks, but the entire element must be on a single line) - bold, italics, and underline tags are stripped
<code>
tags are stripped- links to fragments on the same page may not work (but further investigation needed)
- flat lists work as expected
- nested un-numbered lists work as expected
- nested lists involving numbered lists don't work
- having several
<p>
inside a list item does not work (due to a specialp:first-child
CSS rule) - simple tables with and without captions work ok
- left/centre/right aligning text inside table cells works
- table header rows and header columns work
- table borders don't work
- table alignment does not work (tables are always stretched to 100% width)
- merged/split cells in tables don't work
<div>
doesn't work<section>
elements are removed, too
This is the static page that will be hosted somewhere at
https://dvb.org/standards/. Make sure you took the above listed
go/no-go feature list above into account before pushing updates. It
describes the contents of the files in the metadata
tree.
This directory holds the tree of XML and JSON files which will be accessible at https://dvb.org/metadata/.
The top level catalog.xml
, and its siblings in the subdirectories
index all XML files in the tree (but not JSON), and are generated
using
XML Catalog Builder. Do
not edit these files.
The subdirectories hold the current and historical versions of the XML
and JSON metadata definitions. Where redundancy occurs (e.g. a file
was not updated with a new release of the specification), symlinks are
used as appropriate. Note well: this implies that checking out a
copy of this repository on a client that does not support symlinks
will not work. If you push updates to the metadata
tree from
such a client, you will mess up the repository. So for all good,
make sure you know what you're doing. ;-)
The catalog builder script expects a certain directory structure, where XML files are only recognised at the third level of subdirectories. Let's consider an example:
metadata/
+-- catalog.xml
+-- foo/
+-- catalog.xml
+-- first-schema.xsd // BAD! Don't Do this.
+-- bar/
+-- second-schema.xsd
+-- baz/
+-- third-schema.xsd
Here, the second-schema.xsd
and third-schema.xsd
would be picked
up by the XML catalog builder script, but not first-schema.xsd
.