Round and round we go .. The major problem was the import of ~30,000 articles, but there is a valid point to be made that these should be counted: Many of them *are* built upon, either by people using "Random page" to find articles to work on, or by people searching for their hometown etc. Then again, there are 10- person-towns which are likely never to be edited. I suggested counting only articles with a minimum number of edits, but that wouldn't help us much here, as Rambot updates his own articles occasionally to change the wording. Neither would a specific article length. What would help us would be flagging particular users as bots and excluding them from the count. [No offense to Ram-Man: A lot of work went into the bot's design, but it is not equivalent to 30k human-written articles.] But there's the problem that we're going to cross 100k articles in a few days, and nobody is going to rewrite the code until then. And after that, drastically reducing our article count would seem silly. It's not that the counter is *deceptive*, it just doesn't properly reflect the human attention that articles have been given. So here's what we should do: Go ahead, put out the press release when we reach 100K. Work on a system to properly handle bots in the article count. Then, suggest this revised article count as a way to measure our *human* contributions and list it as an additional count in the stats. If we feel like it, we can put out another 100K press release when we hit that count, and hope that nobody remembers our first one ;-) Proof: 1) Vera Cruz appeared immediately after Lir's ban. Her edit behavior is virtually identical (mark everything as minor, don't use preview etc.). Her homepage style, listing all minor edits, is the same. She has worked on exactly the same articles and finished the work she started as Lir. 2) Brion checked the email address and it was Adam [name omitted for privacy reasons] (she has reportedly since changed it to another Hotmail account). A. [name omitted for privacy reasons] runs this blog: http://qwert.diaryland.com/ where she has written about her Wikipedia experiences as Lir. Lir has also posted to this list as "Bridget [name omitted for privacy reasons]", a name for which Google has no hits other than these archived mails. See wikitech-l discussion - let's go ahead and redesign the image code to generate thumbnails on request. You know I agree. The main problem I see is that we already *are* voting (votes for deletion, votes for NPOV, votes for www.wikipedia.org) without an ordered process to do so because, thanks to the anarchists who constantly decry voting, we are living in cognitive dissonance. This ordered process should actually be supported by the anarchists, because we could say "before any vote, we have an argument period of X days and try to collect all pro/con arguments on a wiki page; everyone who votes is encouraged to make a reasonable effort to read and understand these arguments". We can also set specific thresholds instead of the current guessing what the "not quite unanimous consensus" is. Things will only change when Jimbo decides to let us officially use voting in some cases. He said to me that he thinks that voting may be necessary in a small number of cases, so perhaps it's possible to convince him. Drop him a mail. Regards, Erik

On Sun, Jan 19, 2003 at 02:09:45PM +0100, Magnus Manske wrote: Would you accept NATO? -- jason(a)jasonandali.org.uk http://www.jasonandali.org.uk/jason/