Re: Parochial control of historical railroad data
While bureaucratic paralysis is undoubtedly true at some, perhaps most institutions, there are some very notable counterexamples, and (to the extent permitted by copyright), millions of books will soon be available online:
Google announced in December, 2004, agreements with major research libraries to publish the full text of their book collections online over six years, including all eight million books at Stanford University and all seven million at the University of Michigan. Additional material will come from the Harvard and Oxford University Libraries and the New York Public Library. Search results for copyrighted books will be limited to short excerpts.
The Online Archive of California has available "over 120,000 images; 50,000 pages of documents, letters, and oral histories; and 8,000 guides to collections ... "
The American Memory Project of the Library of Congress has also been very busy. "The National Digital Library exceeded its goal of making 5 million items available online by 2000." The Library of Congress railroad map collection, available online, is spectacular:
The University of Michigan and Cornell University's "Making of America ... collection currently contains approximately 9,500 books and 50,000 journal articles with 19th century imprints."
Progress always depends on what a very few are doing, so to get an accurate overview, it is essential to appreciate that the accomplishments where things are actually happening dominate, not the pervasive inaction elsewhere. As the bureaucratic institutions eventually come to realize that their competition is making them irrelevant, there will be a scramble to catch up, but all that will be left for them to scan are their unique holdings, as Stanford, Michigan, Harvard, etc. with Google's help will have already provided the rest.