[Arm-netbook] policy for wiki and mailing list

[David Given]

On 19/05/12 01:41, lkcl luke wrote:
[...]
> 1) you put information that is useful to other people - no matter how
> patchy, scant, incorrect, incomplete, in the wrong place, and no
> matter how much you are "afraid it might be criticised" - on the wiki.

I utterly disagree with this. A wiki is totally the wrong place for this
kind of information. A mailing list is.

Posts to a mailing list are timestamped, attributed, are easily
searchable with any common search engine, and occur in context with the
discussion that prompted the post.

Information in a wiki is none of the above. It's just a random
collection of stuff with no intrinsic order and no context.

I've dealt a lot with the NSLU2 project. They have a big wiki, and
encourage people to add to it whenever the solve a particular problem.
In theory, fine. In practice, the wiki is a vast pile of fragmented
gibberish. It's impossible to find anything and what you can find is
mostly hideously outdated and frequently simply *wrong*.

Wikis are great for documentation. They do not make documentation happen
by magic. Encouraging people to add every little item of trivia to the
wiki will not form documentation --- instead you'll just get a big pile
of trivia.

I agree that things need to be documented... but only once they've been
adequately discussed on the mailing list.

...

I believe what prompted your message was my patch about memory use on
the v2 kernel. This patch is a perfect example of something that does
not belong on the wiki. It's a quick and fairly unpleasant hack done
late at night because I wanted something to work. I have no idea what
side effects it might have, whether it breaks anything, whether it
conflicts with anything else, whether it has architecture issues that
will make other things more difficult, or even whether it's a good idea
at all. As such, I cannot recommend that it is used by *anyone*.

The reason I posted it was to get feedback from people who know more
about kernel hacking than I do; specifically, the kernel curator. If
it's a good idea, I would then expect it to be merged into the kernel.
If not, then I would expect it to be rejected with comments (which would
be attached to the mailing list thread containing the patch, and would
therefore be easily findable), and I would either try again or withdraw
the patch.

Putting the patch on the wiki would be worse than useless, because it
would actively add to the amount of misinformation on the wiki: by
putting it in the 'formal' documentation, it makes it look as if it's a
recommended procedure, which it is absolutely not.

-- 
┌─── ｄｇ＠ｃｏｗｌａｒｋ．ｃｏｍ ───── http://www.cowlark.com ─────
│
│ "Never attribute to malice what can be adequately explained by
│ stupidity." --- Nick Diamos (Hanlon's Razor)

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 254 bytes
Desc: OpenPGP digital signature
Url : http://lists.phcomp.co.uk/pipermail/arm-netbook/attachments/20120519/8b0c8c63/attachment.bin