I hereby decree, in my usual authoritarian and bossy manner, that today shall
forever be known as Magnus Manske day. Wikipedians of the distant future will
marvel at the day when the new software era dawned upon us.
Tonight at dinner, every Wikipedian should say a toast to Magnus and his many
On Wednesday 04 September 2002 10:38 am, Helga wrote:
> Hello, I am a little swamped with all the wiki list reading material and it
> seems my limited email is getting overloaded.
You might want to create an email filter to sort any emails with the string
"Helga" in the subject into a special folder (just use the help menu of
whatever email program you use and look up "filter").
Otherwise you may miss some emails that concern you.
-- Daniel Mayer (aka mav)
On Saturday 24 August 2002 12:01 pm, Karen wrote:
> Something I wondered - how do you know who the new users to greet them?
> Do you just look for user names you haven't seen before or is there some
> way to identify them? I'd be happy to do the meet-and-greet but I don't
> know how to do it.
Well - I guess I do it the hard way and scan each edit in all Recent Changes
for a 24 hour period looking for edit link user names (a dead give-a-away)
and for user names I don't remember seeing before. This works for me since I
pretty good reading comprehension and memory.
What would be most useful is a listing of new users that can be accessed from
http://www.wikipedia.org/wiki/Special:Listusers. That way this job would be
BTW we really /do not/ have 3498 real users -- a good many of these "users"
logged in only to abuse our upload utility or for other nefarious or
non-contributing reasons (I don't greet any user who hasn't contributed at
all). Is there a way to get rid of many of these no-longer used user accounts
Lee (just the ones that have been inactive for months and whose user pages
are still edit links)?
-- Daniel Mayer (aka mav)
On Monday 19 August 2002 03:41 pm, you wrote:
> Can still be done later. The problem is the lack of time. If you wait to
> long there are to many links to the new location of the english
> wikipedia that can not be broken. If there is no fundamental objection
> to put the English wikipedia at en.wikipedia.org then that must be done.
> What to do whit www.wikipedia.org can wait (a littel.)
This is just silly -- we are building an encyclopedia here not an
organization. There is nothing at all wrong with having the English wikipedia
at wikipedia.org and have all the pages that are about the English language
project be in the wikipedia namespace (or in the other languages project
namespaces). As each language figures out what to call their wikipedia we can
buy them domain names for that and make sure the xx.wikipedia.com domain
names still work.
Other than being a one-page portal to all the different language wikipedias
(which the Main Page already does -- as do most of the other language main
pages) I don't see any logic in using wikipedia.org for anything other than
the English language wikipedia.
-- Daniel Mayer (aka mav)
I don't know anything about Max and so I don't know whether he
deserves an article, but the underlying question is important and
interesting: should anything that somebody bothered to write up be
accepted into Wikipedia, assuming it is NPOV?
I say no. The item has to be of relevance in some way. It would be
fruitless to try to define "of relevance" on many policy pages. It is
implicitly defined by the overriding goal of writing an
*encyclopedia*. Reasonable people will usually agree, and if not, it
can be resolved the wiki way.
Lists of phone numbers, a description of my dream of last Tuesday, a
biography of my father are all irrelevant and don't belong in
Wikipedia. Imagine Encyclopedia Britannica was completely freed of
their space and money constraints: they still wouldn't include a
biography of my father.
Now, Simpson characters on the other hand are relevant: a very popular
show in the largest consumer market of the world. Simpson shows are
being analyzed in literature departments, and no doubt EB would
include an extensive treatment if they could.
Axel makes a good argument that the new code's simplified layout
of the Recent Changes page (see
really is an important loss of functionality. Since his comment, I
added an article history link, but I think his comment still applies
in general. I think it is of critical importance that this feature
be optimized both for function and for speed. I don't want to put
the old code back because it's a dog, but if I can find a way to
implement the needed features in a faster way, I'm all for it.
First, I'd like to solicit an opinion from Jan or other database
gurus about what might make the speed better. Currently, it is
driven by a single SELECT from the current article table, which is
sorted by reverse timestamp (on which there's an index), and with a
LIMIT. The old (N changes) feature also required accessing the old
versions table and an expensive GROUP BY, which I'd like to avoid if
One way I might do that is to create a new "changelog" table, updated
on each page change. This would hold the information needed for the
RC page which would avoid having to hit the article table or the old
versions table at all. It would use basically the same query (by
reverse timestamp). It might even be able to do the GROUP BY faster,
but if not, it would still have the advantage that multiple changes
to the same page would all show up in the list at their appropriate
position, which might actually be better, because that would also
show the person changing it and other stats which are now buried in
the (N changes).
Jan, would the increase non-atomicity of the article save process be
a problem? MySQL doesn't have transactions, so it's possible that
either an aritlce might be saved while recording that fact in the
changelog fails, or vice versa. I can't think of any real negative
consequences of that (each article's history list would still be
accurate, because those go straight to the article tables).
Any other ideas?
> So, could we perhaps work on being a kinder, gentler, less punitive
Amen. Along the same lines, I personally would like to see most of the sysop powers removed (deleting and locking pages, baning IPs, etc) from the general Wikipedia population. I just don't think it's necessary in most cases to take (or threaten) such actions, even in the case of vandalism. We had vandals before everyone had these powers, and they were easily delt with: a dozen Wikipedians undoing every act of vandalism within 30 seconds is hard to get around. Of course, there are some instances when banning an IP is necessary, such a continuous attack over the course of a week. But I am in fabour of leaving these powers in Jimbo's hands and letting himmake the call. It just seems too tempting to use these powers, with good intentions, unnecessarily.
- Stephen Gilbert
Download the free Opera browser at http://www.opera.com/
Powered by Outblaze
I asked why the person was important, JHK answered and
that was good enough for me. Everything else important
has already been said.
Shall we get back to business now?
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
Another update on http://www.piclab.com/newwiki/wiki.phtml
Many, many, thanks to Axel Boldt for much testing and for
giving me a clue how the differencing engine in the current
codebase works. I transplanted it into the new code, but
changed the display to a side-by-side layout much like CVS
diffs on Sourceforge.
I also changed the history list format, and I'd like some
feedback on that. Since we don't have revision numbers anymore
(an earlier decision--should we revisit that as well?), it
made no sense for the links on the history list to have the
page name--they're all links to the same page name, so the
text of the links is now the date and time of their revision.
Each is preceded by a link to a diff page, either that revision
compared to the current, or that revision compared to the next.
I think "previous" would have been a better link here than next,
but that would require some tricky lokk-ahead code, and all the
links one might want are there in this format, just maybe a line
below where you might expect them. I'm not sure how big a usability
problem that is.
Begin semi-rant, apologies to all...
PEOPLE!!! Why are we wasting time on this? First, somebody thought
Weismann was important enough to warrant an article. Why not ask the
person who added him? Alternatively, one could look on google. The
information there pretty much shows that Weismann is a close colleague
(and maybe sycophant) of Adler. Adler IS important, whether or not one
agrees with his philosophy. He's influential -- a large part of a
generation of American pedagogy has been influenced by his ideas -- not
just the Great Books/Great Ideas thing, but also, if I'm not mistaken,
New Math. Weismann seems to be fairly influential in peddling off the
Great Ideas thing into home-schooling -- a major trend in the US these
Wikipedia isn't paper. Helga puts up insane articles about obscure
(hired famous German painter X) city council members in Hansa cities in
the 17th century, and we keep them (despite wanting to nuke most of
them)... Dan, why is this an issue?
Unless there is a huge demand to police relevancy, shouldn't people be
more worried about making sure information is first accurate, and then
NPOV? As much as I'd like to see articles weeded out (like every effing
chapter of Atlas Shrugged), I don't go all sysop and delete at will, nor
have I blocked any of the people I consider intellectual vandals or
liars....or just stupid cows... DESPITE GREAT TEMPTATION.
So, could we perhaps work on being a kinder, gentler, less punitive
wikipedia? Rules are great, but sensible application of guidelines
might be a better approach...and too much information may be better than