[Posted on january 31, 2016]
Nearly nine years ago (time flies!) on this blog I wondered aloud about errors in mathematical papers (errors that are made even by the best authors, in the best journals with very competent referees), and the problem of the subsequent errata that are possibly being missed by readers (at least for a good few months maybe). So I proposed at the time as a solution a centralised system listing all errata in a single place.
In the meantime, MO was born, and I asked there a question about when to take published results for granted at all, the consensus seemingly being: never.
So in the world (indeed, the era) of paper journals, it was hoped that the readers would find the erratum early enough that it doesn’t affect too nastily any work building on the problematic proofs. But in the age of Open Access online journals, one might wish things to change a bit.
To wit, the Open Access Electronic-only journal Algebraic Geometry has issued its first erratum at the end of last year. But, as of today, if one downloads the original paper, nothing warns the reader about the erratum.
Wouldn’t it be wise, and very feasible technically, to add above the title in the pdf of the original article a note like “an erratum for this paper has been issued” with a link to it ?