Hi Thomas,
On Sat Feb 8, 2025 at 8:12 PM GMT, Thomas Lange wrote:
Maybe people can mark the pages they think are of good quality.
That's a good idea, there's precedent in Wikipedia and similar
communities, and it's something we can codify Today.
Why do most people say they look at the archlinux wiki when searching
for information?
I honestly don't have a solid answer for that question but I think it
deserves to be explored thoroughly. In my case, the value I've got from
pages on the Arch wiki have tended to be relatively long pages and
"opinionated". Looking at my browser history, for example, there's
<https://wiki.archlinux.org/title/Xterm>, which:
* doesn't just repeat content from xterm's own documentation
* is somewhat specialised to the audience
("On Arch Linux, xterm sends ^H key when backspace"…)
* tackles "modern" concerns like UTF-8, modern expected behaviours for
this class of app (visible scrollbars, alt and backspace etc behaving
as expected);
* is technically precise (e.g. "xterm's default font is the bitmap font
named by the X Logical Font Description alias fixed, often resolving
to"…: doesn't just name whatever the most likely static font is,
links off to a page describing Logical Font Description in more
detail, specifically details that 'fixed' is an alias…)
Other good examples from my history are
<https://wiki.archlinux.org/title/Systemd/User> and
<https://wiki.archlinux.org/title/Xorg/Keyboard_configuration>
Visible on all of these are meta-notes and discussions (admonitions for
sections e.g. "This article or section needs language, wiki syntax or
style improvements. See Help:Style for reference."; links to individual
Discussion ("Talk") pages for debates on particular articles).
Crucially, I think, there's a community and consensus operating *on* the
wiki. Our style has been to push meta-discussions away from the Debian
Wiki and to debian-www@ (see e.g. the stern warnings on
<https://wiki.debian.org/DiscussionTemplate>); I've never been in favour
of this and frankly I think time has shown that this hasn't worked.
The ratio of good to bad pages on must be improved
can I think this can only be done by deleting the bad ones. OR better
start from scratch.
If we do no risk to delete pages, we will never get a better wiki.
On this particular point, I think there's room to "agree to disagree",
but only because, whether one is "deletionist" or not, I hope we can
agree that removing bad content *alone* is not sufficient: we need to
identify (or write) good content, too.
You may be happy to know that this year a few editors (inc. me) have
instigated some large-scale deletions. I didn't note the page count
before/after starting but a cursory glance at
https://wiki.debian.org/RecentChanges should show LOTS of *deleted*.
(not sure who "zeha" is but, thank you, whoever you are!)
--
Please do not CC me for listmail.
👱🏻 Jonathan Dowland
✎ j...@debian.org
🔗 https://jmtd.net
