LTS -

For a fully professional and accurate rendering, an infinite number of monkeys might be needed, although one can get into serious debates about how many monkeys actually would be needed to replicate the 'cat "within a reasonable probability" of success and with an acceptable level of error. Since an exact replication of the previously existing 'cat is not really mandatory, a reasonably large number of monkeys would give a finite likelihood of replication of the prior 'cat database with good expectation that the number of errors of reproduction would be limited to an easily definable percentage of data elements.

Unfortunately not all participants at the 'cat are "fully randomized" (although the presence of "bias" producing repetitions of the same stuff over and over will likely be a surprise to some). Including defective monkeys might seem to lead us into the requirement for higher order infinities. I attempted to avoid the arguments about "my infinity is bigger than your infinity" by using the less accurate "ten thousand monkeys." In fact, since the inclusion of "defective monkeys" in any infinite set doesn't affect the "size" of the included set of functional ones (assuming some limit on percent defective) that would be another endless and repetitive argument.

The less precise "colloquial" term is perhaps more descriptive of our behaviour here, anyhow.

For us other monkeys:

A quick check of the first 20 threads in the top section as of now shows posts out of order in 5 of the 20. Short threads started after the "recovery" seem pretty good. Threads that were "dormant" for a while before the data corruption seem mostly to be okay, although there were not enough in the sample to be sure.

Long threads (many posts) are more seriously garbled, and surprisingly out-of-order posts pop up among the newer ones posted "post-recovery." I had sort of expected that posts since the recovery would be in pretty good order. While this indicates some "error propogation," I don't think the database grows rapidly enough for this to be cause for serious concern.

In most cases, individual readers should be able to make sense of the threads, if they are aware that some posts may be out of order. The more serious concern would be with posts that are simply "missing." It's reasonable to assume that some of these may still be in the database, although there may be some actual loss. Finding unlinked records, if any exist, may be a bit of a trick; and may not be worth the effort.

The recovery thus far has been remarkable, and the 'cat lives.

John