David Gerard wrote: By an interesting coincidence, just this morning I've been working on my own little database of CAS numbers. You won't find, anywhere on the net, a nice little tab-delimited file of chemical information containing columns for name, formula, CAS number, etc., precisely because of this CAS claim on the copyright of their numbers. I was wondering how long it'd be until CAS complained about Wikipedia. Whether they're non-profit or not, CAS acts precisely as jealous of its set of numbers as any other commercial database company. The impression I get is that this *is* a significant nuisance for chemists and other scientists. Other entities (I can probably find the details) have attempted to establish their own, freer sets of unique identifiers for chemical compounds, precisely in hopes of avoiding the cumbersome restrictions placed on the use of CAS numbers. But CAS has sued -- and I think successfully -- to discourage this, claiming either that they own the idea of a single master database of unique identifiers for chemical compounds, or that having a competing set of identifiers would sow confusion. Whatever the legality of the situation, I'm betting Wikipedia isn't the first information service to have incurred CAS's wrath. We can certainly learn from the experiences of others. For example, here's a set of URLs I collected a little while ago which permit free web-based lookup, by name or CAS number, of information about chemicals, including their CAS numbers: http://www.emolecules.com/ http://www.cdc.gov/niosh/npg/ http://www.inchem.org/documents/icsc http://pubchem.ncbi.nlm.nih.gov/ http://chemfinder.cambridgesoft.com/ I get the impression that CAS is grudgingly tolerant of services which allow one-at-a-time, interactive lookup of chemicals, but what they're adamant about is that no one create a simple database of chemicals using the CAS number as a primary or secondary key. Wikipedia isn't quite one of those, of course, but you could do a decent job of creating one by writing a script to go through every article in [[:Category:Chemical compounds]] and extract the relevant information from the {{Chembox}} at the top -- precisely as I've currently got a script in the background doing. I work with a guy who used to work at CambridgeSoft; on Monday I'll ask him what he knows about the CAS situation.

On Sat, Mar 8, 2008 at 8:13 AM, David Gerard <dgerard(a)gmail.com> wrote: I don't have informed legal commentary, but my guess is -- and their actual comment bears this out -- that ACS's objections are due to the extremely stringent license requirements that one agrees to for when one subscribes to SciFinder. Basically, redistributing information from SciFinder/CAS Abstracts to other people who didn't subscribe (i.e. via Wikipedia, or otherwise) is almost certainly not kosher. I make this guess based on the rules that libraries agree to when they subscribe to SciFinder. There are nearly always specific provisions in the contract (not just for SciFinder, but for most databases and journals) that say who can get access to the data -- for instance, only the faculty, staff and students of a university. Obviously enough, the many glorious readers of Wikipedia are unlikely to fall into this faculty/staff/student classification. Assuming this is true for most contracts they sign, then probably anyone systematically posting data from one of these systems is violating some provision of their contract. This doesn't have anything to do with copyright law per se, but it is a terms of use question. The separate issue of CAS numbers being copyrighted is very unfortunate, but everyone is right that the CAS numbering and indexing and data organization scheme that is the core of SciFinder and related tools is very much the lifeblood of ACS, and they are afraid of losing it. Anecdotally, though the ACS is a nonprofit, they have historically not been inclined at all to go to open access/open licensing/open content schemes, unlike some other major publishers who are starting to see the wisdom of making information freely available. So I'm not sure that an appeal to their better nature would be terribly likely to work. But, it's always worth trying. -- phoebe