Han-Teng Liao recently inquired about the effects of the unblocking of the Chinese Wikipedia on the traffic volume directed to zh.wikipedia.org. He may be as amazed as I am that the effect in terms of number of page requests has been pretty much none whatsoever.
The following three charts each show the number of page requests to the Chinese Wikipedia over the course of months, each at a different level of aggregation. Looking at these charts I can’t see anything that signifies at which exact date the custodians of orderly synchronized opinion forming opened the gates to the world at large, a world where expressions of misalignment and self-righteousness are a constant danger.
Note: monthly figures have been normalized to 30 days for better comparison: figure for January is 30/31 of actual value, for February 30/29, etc.
One question leads to another: in the above chart with hourly page requests a few points stand out (marked with larger symbols). I decided to look in depth to one of them: the hourly stats for 22 February.
I expected this sudden increase in page requests was proof of a sudden rise in interest in Wikipedia, possibly caused by a recent event that motivated lots of Chinese speakers to get briefed on one particular subject, something like the surge in visitors after the nomination of Sarah Palin as running mate for the presidential elections (see previous post). But nothing like it was to be found: the pages that attracted a meaningful number of visitors were no more popular than on other hours. I took me a while to realize and verify that many pages (well over 250,000) were requested exactly once that very day, and mostly between 9 and 11 GMT. Mystery solved: someone had downloaded the whole Chinese Wikipedia, page by page, using some kind of program. How exciting news it that? I would say: pretty much none whatsoever. But since it took me a while to find out, I did not want to omit this finding here.
Even though there was a clear rise in requests close before and after 31 July the absolute numbers are not particularly spectacular I would say. I can imagine all those requests were issued by reporters in Bejing who wanted to verify rumours about the unblocking. It makes me wonder whether the unblocking was really universal and for all of China.
I want to emphasize one point: we can count the number of page requests received, in other words the relevant data packages that reach our servers. We have no idea how many people tried to visit but were redirected to another Chinese site, because the url had been found on a governmental black list (this seems to be normal practice in China).
It is even possible that some pages were handed out by our servers but still did not reach the user. In this scenario the page had been scanned for sensitive content after it had been sent by our servers, and only at that stage redirection had taken place. Both methods of redirection reputedly are (or were) common practice in China.