On Wed, Sep 23, 2009 at 12:23 AM, Erik Zachte <erikzachte(a)infodisiac.com> wrote: Not really, they are both factors and one or the other may dominate in some places. Possessing a computer isn't the criteria I'd impose, it's having reasonable access to one and there are many people who are illiterate who could be reasonably counted as having access even though they can't make much use of it. [snip] [snip] It does but that vivid painting is not a relevant one, and as such it is potentially deceptive. The reason "how we have not reached large parts of the world yet" is because access to Wikipedia is significantly influenced by things outside of Wikimedia's control and scope. I think a reasonable argument can be made that Wikimedia's actions could not produce a statistically significant improvement in the penetration vs population metric; simply because the causative factors outside of our ability to influence are so large. You sound like you've heard me making this point and you acknowledge that this way of measuring has weaknesses, but perhaps I did not state this bluntly enough: I think you're measuring something which is largely irrelevant to the activities of the WMF. If you instead took each language as it was and simply said that we'd be doing well to increase the users of each language, I'd agree. But saying that the WMF base its decision making on a measurement of "not reached large parts of the world yet" is effectively equal to measuring our performance by observing that Wikipedia is only used on 1 of ~382 planets known to exist in the universe. Failure to consider this leads to bizarre conclusions like "the lower birth rates in the developed world compared to the developing world decrease Wikimedia's success over time". Which is simply silly— we shouldn't ignore our progress just because it may be dwarfed by shifting populations. Surely someone must have a respectable count of internet users by language that we could use for comparison? That would be a much better metric for our success today; while raw literate speaker numbers would be a useful comparison for what we could start reaching with non-internet mechanisms. [snip] I fully agree with you that article count stinks as a metric in many ways— but at least it is measuring something which is pretty much fully under our control. And this is really the point I was trying to make: While article count is a bad metric because not all articles are equally helpful. Penetration vs Population isn't even bad, because it's quite possibly not measuring something which is generally useful to our decision making. (Except perhaps in the decision to chase non-internet distribution; but without literacy calibration and a comparison to a measure of online reach it's even of questionable value there) If we had a goal to double the number of articles in some reasonable period of time we could do it. If we had a goal to double the penetration, ... well, unless the WMF changes its mission nothing it could do would get us there.