Wikipedia Time Machine (Part 1)
/So I'm at a Timbers match earlier this month, and a friend says to me:
I wish I could use the Wikipedia during the World Cup, but I watch the games delayed, and I'm afraid I will see the results. Couldn't someone write something that show the pages as of the beginning of the game?
I remember this a couple of days later and think, hmm, she's right. AAMOF, someone must have already done that, right? Off to google:
- "wikipedia page as of the day",
- "wikipedia topic as of timestamp",
- "wikipedia topic as of datestamp",
- "wikipedia time machine",
- "wikipedia tardis",
- etc.
Nothing. Well, that's surprising. But maybe I can write one? I know that wikis keep previous versions. Maybe I can specify the version using a timestamp? So let's google "wikipedia page by version". Ah, this looks interesting: Linking to a specific version of a page .
OK, I can specify the version of a page in the URL, as shown here.
To link to a particular version of a page: http://en.wikipedia.org/w/index.php?title=Train&oldid=1374279 (where oldid is the revision number), or use Template:Oldid. A permanent link to the present version of the page can be accessed by clicking "Permanent link" under "toolbox" on the left side of the page. The version ID is unique across all pages; the title parameter here has no effect, and can be omitted. Special:PermanentLink/1374279 is a wikilink to a page version.
However, there is a warning back on the previous page:
However the permalink does not guarantee to display the coherent historical version of any page, except for very simple pages; this is because most pages include images, templates (transcluded text and images), or time-based variables (such as CURRENTTIME). All these elements are not retained using a "permanent link" and they will be rendered according to the current state of the image/template, or the current time, producing a different result than was rendered at the time the old version was saved. Only the main text of the article is retained correctly, that is, the wikitext that one sees when clicking Edit.
So I can craft a wikipedia URL that will return the page as of a particular version, but included text and images may be newer. Probably OK, as a new image, say of your favorite player scoring a goal in the game, would have a different URL than the images on the previous version of the page, so it wouldn't show up. And I don't think people would be adding scores and the like as included text.
Unfortunately, I find that I can't specify the page version as a timestamp. I'll need a way to translate a timestamp into a version. We'll try to figure that out in Part 2.
