An underlying question is why become embroiled with bibliographic management software, and part of the reason it's now worth revisiting is the broader problem of managing information about a broad range of information resources (which the RE Lee project is addressing). The fact that EndNote is capable of connecting to and interacting with important databases (Web of Science, SciFinderScholar, FirstSearch, others) is probably reason enough. The problem of teaching/supporting is, however, non-trivial.
The question of how to move EndNote off the individual desktop --how to distribute bibliographies-- led me to thinking about the practicality of moving EndNote material into Microsoft Access, and a weblet at University of Missouri (OutcomeLit) offers a multi-step solution to the problem... and then goes on to note that Thompson ISI ResearchSoft has a product called Reference Web Poster which is intended to fill just this niche (list $149.95). A review from UC Berkeley is quite favorable.
A key bit of software for installation on local machines: ISI ResearchSoft Export Plug-In for Windows ("the free plug-in that installs the utility and import filter required to export references from various Web-based products and services into your EndNote 5/4, Reference Manager 9.x/8.x and ProCite 5/4."). I tried using Web of Science with this and had no difficulty adding references to the bibliography. I found that I was also able to add the necessary filter to import records from CSA's Biological Sciences database, WITH the keywords, abstract, etc. Similar success with GeoRefS, and with HOLLIS (there are many filters available, to add to those already in place in the installed version). Many of the databases Greg uses do have filters available.
I also used the 'connect' feature to retrieve records directly from the Amherst (Five Colleges) Online Catalog --but the INNOPAC filter doesn't work, because Annie doesn't give us the option to SAVE in EndNote format, and probably because we don't have Z39.50. Other pages with filters
The question of the moment: should we be supporting users, and encouraging the use by others? The possibility of output to the Web seems to me to tip the balance, but what we really need is a worked example to show how it's an actual improvement.
I explored SciFinderScholar, and found that I needed another filter (for SFS2001) to make it work. Google led me to the SFS page with the requisite filter, but this exemplifies the care and feeding necessary to support EndNote: one has to know that there might be a filter and hunt it down and install it... and then make it available to other users who might want to use it too.
The next question: are there practical ways for multiple users to enter things into an EndNote library? Could members of a working group use one successfully? Or would it be better to merge what members had done separately? And how does all this fit with Reference Web Poster?
I did some experimenting with exporting tab-delimited records and was able to get them into Excel, but the various formats make for a messy transfer (/digilib/collab.xls is a substantially cleaned up version).
I've been exploring Reference Web Poster (see current state) and find its primary appeal to be as a negative example: it does make content of various EndNote libraries available via the Web, but is clumsy and counterintuitive to use. We can do better...