I imagine the quip ID is a unique identifier for every answer. As a database developer, I would likely put all answers into one table, thus there would be a sequential primary key which is likely the number in the permalink (or would be if I had designed this site).
Your test, combined with looking at my comment directly above the two of yours, would support this theory. Your two quips had two sequential numbers, presumably because you very quickly posted your second test post. However this sequence is not carried backwards through my answer above yours, so I feel it’s safe to conclude what I theorized above is pretty close to what they have set up.
That would mean there could be close to 425,000 answers posted through the history of the site (assuming they started with quip1).
Hey, this reminds me of the time we found out the first ever Tweet. Let’s see if I can pull a similar trickery… we’ll load up the first ever question asked here. It has a discussion number of 1, which lays the theory that the questions are another database table from the answers (I’ve based this on the URL structure pretty much, and intuition). And sure enough, quip number one is Ben’s first reply to his question. That question is actually very interesting metadata-wise, you can skim down the list of quip URLs and see big gaps where presumably other comments in other questions were added – sure enough, looking at question number two shows that the quip numbering remains continuous across the both of those two questions. Thus, the evidence for there being more than 425,000 answers posted between the first ever question and us writing these words right now is very strong. However, without manually verifying each of them exist, I can’t say definitively that there are this many comments, as simply changing the primary key starting value to 399,999 would produce comments in the 400 thousands.
That was a big derail! But yes, if there was somehow no timestamp in the DB, the quip ID could be used to infer the same data. As a developer I find the lack of such timestamps to be a bit surprising, as I would have included such a field on such a constantly-accessed table. Then again with more than 400,000 records, not including the timestamp for each record saves a lot of space, and since questions have timestamps included, the need for comment stamps may have been reduced in priority. Here’s one older discussion on timestamps in Fluther which has some interesting discussions about the issue. Andrew’s comment at the end is, due to his nature with the site, particularly insightful about the public-timestamp issue. Based on his answer, I think it’s clear that all quips do have the server’s timestamp associated with them. That helps with our pseudo-algorithm building here.