What is going on with MSE's "first" question?

Question

Out of curiosity, I decided to try sorting MSE questions by age from back to front, and I noticed something odd: while the second through $48^\text{th}$ questions are all from July 20, 2010, there is a single earlier post, from March 27, 2010. It has a few comments dated March of 2010, as well.

Furthermore, the URL of this post is weird. Every MSE question or answer is assigned a unique positive integer identifier, counting up one at a time. https://math.stackexchange.com/questions/1 takes us to the first of these July 20 questions, but the sole earlier post has a URL of https://math.stackexchange.com/questions/106814, and its answer a URL of https://math.stackexchange.com/questions/106815/. Compare to https://math.stackexchange.com/questions/106813, which dates to 2012-02-07 21:37, and https://math.stackexchange.com/questions/106816, which dates to 2012-02-07 21:46. (I looked at several questions from that evening and wasn't able to find any "ghost" questions or answers in the interval between 21:37 and 21:46.)

In fact, if one looks at the profile of the user who posted this first question, as well as that of the only contemporary commenter whose account remains, neither one's account age is compatible with having used the site in March of 2010, as they both appear to have made their accounts later that fall. In fact, the apparent earliest user of MSE whose account is viewable seems to have created their account on July 20, 2010.

I checked MathOverflow and StackOverflow, and a similar phenomenon does not occur; the first question has some very low-indexed URL as expected, and was written by a sufficiently old user.

Doing some archaeology on the Internet Archive, I can find a single snapshot from 2009 (!) which displays a single question by Brian R. Bondry. Current MSE pages show the index-1 question as this one, and the user as being #53. (Frustratingly, no 2010 snapshots appear to have been taken after January of that year.)

What's going on here? Are there any MSE veterans who can shed light on these weird relics of the earliest days of the site? I'm very confused.

Answer 1

In the process of writing out the above question, I finally figured out the answer! I thought I would post this as a self-answer rather than delete the question, to save any future MSE archaeologists the effort.

While the two users involved in this "first" question did not create their MSE accounts until the fall of 2010, they had StackOverflow accounts since the earliest days of the site, and indeed (after LOTS of binary searching) one finds that https://stackoverflow.com/questions/2529632 redirects to https://math.stackexchange.com/questions/106814. So, I believe, the origin of this question is the following:

• On March 27, 2010, Jason posted a question about relational calculus on StackOverflow and later self-answered it.

• On July 20, 2010, MSE was started in earnest, and the first "real" questions were posted to the site; at this point, and for the next 18 months, looking at MSE would have shown the oldest question as being from July 20.

• On February 7, 2012, at sometime late in the UTC evening, a particularly diligent StackOverflow moderator scouring the archives saw this question, thought it was better suited for MSE, and moved it, whereupon it (and its associated answer) were assigned the next two ID numbers in the queue, 106814 and 106815. Thus, I think by most reasonable definitions it is not MSE's oldest question, despite appearances.

I think this mostly wraps up the title question, but I'm still curious what's up with that 2009 Internet Archive snapshot. Any MSE veterans who might know?

Answer 2

I don’t know the history, but a close look at the 2009·10·15 Internet Archive (“Wayback Machine”) backup of http://math.stackexchange.com reveals indications that it is a prototype or stub:

• The author of the question is shown as “Brian R. Bondy♦♦”.  As we know, only Stack Exchange moderators get to display a diamond (“♦”) at the end of their name; double-diamond is never used.

• The page says

  Welcome to StackExchange

  A place to ask questions about _______.

  which looks like it was copied from a template.

• It also has headings “Sidebar” and “Sidebar Low” (placeholders?), each of which is followed by a (non-archived) link to “Customize this”.

• The page ends with “Some example footer content.” — literally (obviously a placeholder).

   

• Also, Brian R. Bondy is a Stack Exchange user of long standing, who is interested in mathematics in general and verifying primes in particular.  He followed the Area 51 proposal for Mathematics Stack Exchange and posted answer #24 to Math SE question #8 a mere 17 minutes after question #1 was posted (on July 20, 2010), but has not been active since then.  (His Stack Overflow activity pretty much ended in summer 2011.)

So, we can guess that the MSE private beta (or the Area 51 site) briefly (accidentally?) went publicly visible in October 2009, and that Brian Bondy, as user #2, asked question #1.  And the Internet Archive managed to grab a shallow copy of the site home page (without digging down to any of the pages that the home page linked to).  And then the page got restricted, the way private betas are supposed to be.  (Or maybe the site was just idle for the next several months; I can’t explain the fact that the January 10, 2010 snapshot is identical to the one from 2009.)  And then, on July 20, 2010, all of the contents of the beta site got purged, and the current site was opened to the public.