Re: What's next for the Debian Wiki? (was Re: wiki.d.o on a git-backed engine)

Maytham Alsudany <[email protected]> Thu, 30 Jan 2025 17:47:14 +0800
Newsgroups	gmane.linux.debian.devel.www
Message-ID	<[email protected]>
Hi,

On Sun, 2025-01-19 at 23:31 +0100, Serafeim (Serafi) Zanikolas wrote:
> second, having looked at the freebsd and archlinux wikis, I'm starting to think
> that:
> 
> - high quality content is about boring process (doing small, well scoped
>   changes, much like one would do with code; having people actively review them;
>   and roll back changes when they don't follow certain clearly-communicated
>   guidelines)

IMO the main issues with the current wiki, regardless of content, is:
- Pages are unorganised and fragmented.
  Related pages should be grouped together rather than all being on the
  top level. There are plenty of examples of this in the wiki, one
  example would be all the different packaging tutorials in the wiki.
- Lack of guidelines.
  I agree with the fact that we need better, enforceable guidelines on
  content that can be included on the wiki, especially with redirecting
  certain types of content elsewhere. And on that point:
- Duplication of content.
  The wiki repeats a lot of things that are already better documented
  elsewhere. For instance, I don't see the point of an install guide on
  the wiki when there's already an existing official one.
 
> - the choice of engine has little to do with the quality of the content (as much
>   as I struggle to bring myself to use a web-based editor, I would not be
>   surprised if I'm in the minority in wanting a git interface)

Agreed, though we still should really move to something like MediaWiki.
MoinMoin v1 is outdated, v2 isn't coming out anytime soon, and
continuing to run buster on wilder.d.o isn't sustainable.

> On Sun Jan 19, 2025 at 11:54 AM CET, Jonathan Dowland wrote:
[...]
> > And so on to my 2p's worth: A lot of the discussion I've seen focussed 
> > on (a) narrow list of problems (moin 1.9.x, Python 2), (b) advocates 
> > solutions (e.g. mediawiki) and (c) makes assumptions (we must migrate 
> > all the content in perfect fidelity if we move wiki engine) which might 
> > not hold. I feel we should spend some time thinking hard about what we 
> > want from the Debian Wiki, i.e., a good old-fashioned requirements 
> > analysis, before we look at (b) or (c).
> 
> I completely agree on (b) being premature, but I think of (c) as a requirement
> (or non-requirement if we agree that fidelity is optional)
> 
> my 2c on (c): many parts of wiki.d.o are really outdated and important parts of
> the wiki are in dire need of consolidation and cleanup (e.g. see
> /DebianWiki/NextGen discussion of contributors' guidance in the Debian vs other
> wikis) that aiming for high fidelity to me seems entirely secondary (if
> desirable at all)

Are there any objections on doing major changes on the wiki? There's not
really any clear standard on what editors are allowed and not allowed to
do. Personally, I'd like to move a bunch of pages around and perhaps
delete some pages that aren't needed, but am unsure whether that's
acceptable or not.

--
Maytham
DM
signature.asc (application/pgp-signature, 833 B)