Dang. This is important stuff. I guess I better just get that WP port thing working then. The bot certainly can unload to a file, but it really is set up to to this simulateous log on. It may be something really stupid in my configuration, but I have not yet isolated the problem- it may be a little more subtle because the common problem they were addressing in the pywikipedia code mades some assumptions having to do with automation of the moving of data from the various WPs to Commons.

But it may well have been an error I made.

~ Phlox 07:48, 21 October 2007 (UTC)

Noted with interest but little comprehension. It actually suggests that it's more to do with the other item I mentioned, lists of counties, where a relevant guideline would be Forum:County Pages Idea. But no rush. Robin Patterson 13:11, 21 October 2007 (UTC)

Renewed interest[]

Comments moved from main page. Maybe it needs to be discussed at forum, since talk pages are hardly ever refered to?
The year pages are indeed a mess, so much so that I have been unable to design a strategy for AWB to edit them.
I've been putting {{info year}} on year pages, which could be used as a vehicle for standardisation. rtol 16:04, November 5, 2009 (UTC)
I strongly agree that the wikipedia style year and month-day articles represent important anchor point articles that orient the reader towards the larger historical contexts. 1967 is missing a LOT of content from WP, don't we want to update from WP, and in fact have an automated way of refreshing our content from WP? Actually, it is a trivial matter to import missing ones (or reimporting existing ones) from Wikipedia and automatically create linkbacks to wikipedia. What is tough is preserving any material that has been added to existing articles, identifying person articles (links we want to keep internal) from article that we don't cover- like what existentialism is. It is also tough to separate out portions of wikipedia articles that are germane, since they often are often encoded in a nonuniform way, with section names different and so on. Doing a quick survey on WP, it appears they are much more uniform than when I was importing them in 2007.
Regarding {{info year}}. Whatever we do needs to take into consideration the volume of events when hundreds of thousands, then millions of individuals have events encoded that can be retrieved from SMW. Consider info year at the bottom of the page on 1931. How does this scale? Do we want to see tens of thousands of people listed on this page? If not, then how shall it be filtered? It is possible to list information via a query without actually creating a distinct page. The Image pages do this for people depicted. For example on page File:Stanley dunham with daughter ann and wife Madelyn.jpg, {{Media facts}} creates a list of see also images for all persons depicted.
If someone comes up with a systematized approach for importing year pages and create a standard layout like what Robin was proposing in 1967, I'll take a stab at creating an import bot for them.
The harder problem is the {{info year}} problem with how we scale up and provide useful information and navigation aids for the user. Info year is fine for distant events, like year 910 where the user will likely not be swamped with huge lists they won't read. One approach would be to generate dynamic queries that narrowed the field- narrowing by placename. Of course, we would need standardized names for that to work, so there is a dependency on importing placenames for filtering to work well. ~ Phlox 17:18, November 5, 2009 (UTC)