An idea occurred to me recently and I want to see whether anyone thinks there might be mileage in it. It is to create a wiki that tells you for each mathematics journal which of its articles can be found in preprint form online. 

What would be the point of such a wiki? After all, if you want to find an article online, you'll surely just put the title of the article into Google rather than first find out what journal it's in and then go and look in some wiki under the heading of that journal. 

I can see two possible motivations. One is that occasionally articles are available but hard to find. This is not the usual situation, but sometimes -- for example, if the article is on someone's home page and not on arXiv -- it can happen. So occasionally it might be convenient to check in the wiki. Also, if the wiki says that an article is not available, then the knowledge that someone has looked for it and failed to find it could save someone else time that they might have futilely spent looking for it.

But I think its main use would be more as a barometer of how mathematicians are doing in making their articles available, and therefore how much we are relying on the official published versions. This could help to convey two messages: first, that libraries might well be able to do without subscriptions to many journals (were it not for bundling, but demonstrating our lack of need for journal subscriptions is the first step towards persuading librarians not to accept horrendously expensive bundling deals); and second, that those mathematicians who have not yet made preprints available are being antisocial and should do something about it.

To give an idea of what such a wiki might be like, I've indicated which papers are available online from the latest (not yet physically published) issue of Discrete Mathematics, published by Elsevier. I managed to find four out of nine. All the four I found were on the arXiv and showed up first in a Google search of the title. It was a slow process, so doing this for several volumes of several journals is something that would have to be a collaborative effort of many people. One of the purposes of this post is to find out whether anyone likes the idea enough to think that they might be ready to put in some effort of that kind. If I get an encouraging response, then I'll consider organizing something. 

Update. In the light of the first few comments below, I think I should perhaps have asked a different but related question. I'm not interested in trying to create a search tool in an unnecessarily labour-intensive way. What really interests me is data about what we would lose if we lost access to the official versions of journal articles. Is there a quick mechanical (or partly mechanical) way of determining how many articles in a given year of a given journal are freely available in preprint form online? Given what we have already, it sounds as though it wouldn't be too hard to write a program to gather such data. The reason I'd be interested in this is that it's information that I think should be widely publicized, particularly amongst people who are negotiating with the big publishers.
Shared publiclyView activity