On Sat, Dec 5, 2009 at 12:29 PM, Carl (CBM) <cbm.wikipedia(a)gmail.com> wrote:
[snip]
For example, imagine a well-meaning newbie who sees
that our article
"Logic" starts with "Logic is the study of reasoning." This newbie
might change that to "Logic is the art and science of correct
deduction", which is a priori reasonable. They would not know that
people have argued over the first sentence in detail and that the
present wording is a compromise between the many definitions of
"logic" available in reliable sources.
[snip]
We sometimes use HTML comments in the wikitext to note traps like
this.
Perhaps that is a good practice which should be encouraged?
On Sun, Dec 6, 2009 at 10:28 AM, David Gerard <dgerard(a)gmail.com> wrote:
[snip]
Creating forests of redirects is useful to the reader.
Certainly to me
as a reader.
Agreed.
We can't disclose the raw search request feed because people sometimes
plug confidential information into the search box (e.g. accidentally
pasting a password or a whole email there) but we could probably
release some kind of aggregate. For example, release search strings
which came from at least X distinct subnets during the day/week/month.
When the (insufficiently anonymized) AOL search data was released
I took the top query terms where there were no wikipedia articles
and went about making redirects:
http://en.wikipedia.org/wiki/User:Gmaxwell/seo
I'd like to think it helped...
There are probably more from that list which could stand to be created.