I have a hunch that the downfall of the "Plato" real-time downloader wasn't the Indian court case but rather the fact that it helped publishers trivially identi

I have a hunch that the downfall of the "Plato" real-time downloader wasn't the Indian court case but rather the fact that it helped publishers trivially identify the university accounts through which the downloads were happening. Even if the appearance of papers were delayed by a random number of days, there are other pitfalls now, and most importantly, publishers started caring. In particular, Elsevier now slaps UUIDs onto all PDFs you download from them, and no, I'm not just talking of visible watermarks. Other publishers seem to be doing similar things (there was a recent twitter thread on this, retweeted by @textfiles, which I can't find). The rational solution for Sci-Hub seems to be to buffer their uploads and release them in yearly batches, maybe programmatically removing various kinds of watermarks and diffing against the same paper downloaded from a second IP. If this is what they are doing, I'm not surprised. Not sure how much of a winning strategy they have in the long run, though. Guys: post your papers on the arXiv.