Subject: Re: better integration of news pages for NetBSD and pkgsrc
To: None <email@example.com>
From: Hiroki Sato <hrs@NetBSD.org>
Date: 05/29/2005 23:07:03
Content-Type: Text/Plain; charset=us-ascii
[I sent this mail again because the previous one was rejected from
netbsd-doc@ due to the attachment files]
Klaus Heinz <firstname.lastname@example.org> wrote
k.> This format puts the date information into the id attribute instead of
k.> mixing it with the title string.
k.> It will not be necessary to change the id attributes of all news items
k.> in htdocs/Changes/index.xml, but we probably should do that anyway for
k.> the ones recently added (maybe up to 5).
k.> Is it ok to implement that?
k.> Does someone have a better idea?
If we do not have to stick to the current format for news items,
I suggest to switch over to separating the news items from other
normal documents which use them.
Patches for proof of this concept are attached. news2004.xml
is a file simply converted from Changes/2004.xml by using newsconv.pl,
and this is a news item database which can be used from various
documents. The format is something like this:
<entry year="2004" month="12" day="6">
<title>pkgsrc now frozen</title>
<para>As <ulink url="#pkgsrc-freeze-2004Q4">announced previously</ulink>,
the <ulink url="../Documentation/pkgsrc/">NetBSD Packages Collection</ulink>
is now frozen in preparation of the fourth stable branch
<entries> is the root element, and <entry> is a container corresponds to
a news item. <category> can be used to classify this item (empty string
is allowed), and in <description> we can use DocBook vocabulary.
news.xsl converts this database to another .xml file, and other documents
can include the .xml file directly by using XInclude (see 2004.xml in the
attached patches). So, how to use this database depends on news.xsl.
In the attached patch, selection of news items by "category" is implemented,
but other features for query and/or output formats can
also be implemented without modification to news.xml itself.
Would this approach be overkill or laborious? I think use of this
sort of database is useful to avoid scatter of duplicated data
such as news items, mailing lists, mirror sites, description of
ports, and others which can be shared in multiple documents.
[originally attached files can be found at:
| Hiroki SATO
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (FreeBSD)
-----END PGP SIGNATURE-----