jayjg wrote:
No, the accurate statement is "This specific online search by Daniel P.B. Smith on this specific date did not return any results".
I highly doubt the identity of the person who looked in the catalog is relevant. I suppose it would be technically possible for an online catalog to deliberately return different results to different users, but I would not really call any site that did that a "catalog", much less a reliable source.
So we're down to "as of this date, no books by this author were listed in the official online catalog of this library." Doesn't sound all that ridiculous to me.
(Elsewhere you've complained about the supposed difficulty of repeating the search. You do realize that most, if not all, library catalogues allow you to link directly to search results?)
And, of course, there's no guarantee that 1 minute after Daniel P.B. Smith does his search, the book will not be acquired, or entered into the catalog, or re-indexed properly because it had been improperly entered before, or...
If it indeed happens one minute later, what will happen is that Daniel P.B. Smith will look foolish -- or would, if the catalog did not show when the record was added.
If it happens later, there are two possibilities. One is that no-one will notice, in which case we will continue to cite an old version of the source until someone does. Assuming the original citation included the date of access, as our guidelines recommend for all ephemeral sources, this is not a problem: we do not pretend out information is always up to date.
The other possibility, of course, is that someone (perhaps even Daniel P.B. Smith himself) does check the source and finds that it no longer matches the claim in the article. The reasonable thing to do in that case is to change the article. The only problem occurs if the person who notices the mismatch fails to assume good faith and, mistakenly, accuses Daniel of deliberately lying -- but hopefully, even if no objective record of the change remains, other neutral editors will have previously verified the source and can confirm that Daniel's claim in fact used to match the source.
If not, the outcome is the same as in the one-minute scenario: Daniel may suffer a blow to his credibility. Note that, in any case, the risk is _only_ to Daniel's personal reputation, provided that the article is corrected as soon as the nmismatch is noticed; if Wikipedia's reputation were in any significant way harmed by mistakes that were corrected when found, we wouldn't have any reputation left to harm anyway. (Some may, of course, argue that this is in fact the case.)