From: Solomon Peachy via rockbox-dev <rockbox-dev_at_cool.haxx.se>
Date: Sat, 27 Jun 2020 23:27:16 -0400

On Sat, Jun 27, 2020 at 09:43:39PM -0400, Jason Arthur Taylor wrote:
> No, IMO they are not the same at all.

The editing experience is different, sure.

But what no matter what solution is chosen, it will require a roughly
similar amount of effort to set up, migrate, and integrate. [1]

So what I am wondering is if the historical goals fulfilled by the wiki
are still relevant for the rockbox today, because different goals lend
themselves to different approaches.

> after a while you've written a real masterpiece. The instant you
> cannot get that live feedback and see a "your edit is awaiting
> approval" (or the equivalent awaiting an unknown amount of time for an
> auto-update) it's a no-go feeling like "That's the last time I ever
> help those aholes out."

In the past four years, there haven't even been 100 non-administrative
changes made to the wiki. About a third were made in the past year, and
of those, all but one were made by folks who already have commit rights
to the main repo.

It's worth mentioning that new user registrations have been disabled on
Rockbox's wiki since November 2017, because spammers and other bots made
a right mess of things -- and if the server logs (and the DoSes that led
me to start this thread) are to believed, are still actively trying.

So the way things are today, drive-by contributions are literally
impossible. One has to contact me to get an account, and even if you
have one, the user experience is still quite poor due to abysmal
performance, baroque native foswiki synax, and a wysiwyg editor that had
to be disabled because it was silently losing edits (on top of making
performance even more abysmal..)

These days one sees far more "helpful" activity and knowledge recording
on the forums.

So yes, I question if a live-editable wiki is the right approach going
forward. It may make more sense to prioritize the use cases of active
developers, as those benefits are easier to quantify than those of folks
that won't meanigfully contribute anyway.

(Given the current barriers to contributing to the exsting wiki, I
 strongly believe that no matter what approach gets taken, it will be a
 net improvement, for everyone)

I agree that a live-editable wiki can be more user-friendly, but I (and
several others) find working with the current wiki to be a pretty awful
experience, especially for longer-form documents. I think this is even
more true for less-experienced folks.

IMO a best-of-both-worlds would be to use a wiki that's backed by a git
repo and can synchronize in both directions -- Dokuwiki has a plugin
that should be able to do this.

> No masterpieces, only half-butted cruddy lame contributions with
> spelling errors etc. That's exactly what non-live editing is not
> asking but more begging for.

Well, if a static site generator is the approach taken, it will be
possible to competely build and locally test things. With nikola, a
single page update tyically takes less than a second, including
launching a web browser to view it. This is a lot faster than our
current wiki, even after the things I've done to try and make it suck
less.

[1] moving to foswiki 2.x would be the least amount of effort, but
    there's no guarantee it will actually resolve the problems that are
    making a migration necessary to begin with. I don't want to have to
    do this twice.

 - Solomon

-- 
Solomon Peachy			      pizza at shaftnet dot org (email&xmpp)
                                      _at_pizza:shaftnet dot org   (matrix)
High Springs, FL                      speachy (freenode)

Received on 2020-06-27