There have been many books about the compilation of the original _Oxford English Dictionary_, and this is as it should be. The monumental work started publishing its A volume in 1888 and finished the Z in 1928, with 15,000 pages defining over 400,000 words. At a banquet to celebrate its conclusion, it was hailed as "unrivalled in completeness and unapproachable in authority; as near infallibility, indeed, as we can hope to get this side of Rome." Millions of people trust that near-infallibility; to say "The _OED_ says..." about a word is to give the strongest of evidence. Yet although the dictionary was finished in 1928, it was not really finished (and never could be), and it was far from error-free. In 1951, a co-editor of the original _OED_, C. T. Onions, wrote that the great work had "hosts of wrong definitions, wrong datings, and wrong crossreferences. The problem is gigantic." How the lexicographers and the Oxford University Press handled the problem of updating and correcting the dictionary by supplements, abridgements, electronic versions, and the current online version is the subject of _Treasure-House of the Language: The Living OED_ (Yale University Press) by Charlotte Brewer. Anyone who uses the _OED_, and anyone with any interest in words and the roles dictionaries play in the use of language, will find stimulating this scholarly history of the _OED_ after its first edition was completed.
Since the dictionary took forty years to print from A to Z, and since English had plenty of new words and new emphases by the time the great task was completed, everyone knew that there would have to be a supplement volume to cover the years when the dictionary was in production, a supplement that would have plenty of entries for the early part of the alphabet and fewer for the later part. This first supplement came out in 1933, and some thought that this would be the end of the _OED_ effort, since such a gargantuan project could not be completely revised, and indeed work on the dictionary did go into hibernation in some ways. However, the readers and contributors who had from the inception of the dictionary sent in slips with examples of words and usage simply continued; they were in the habit of excerpting quotations from everything they read and did not stop. There were, indeed, four more supplements issued, the last one in 1986. This was unwieldy, because if you wanted to be sure about a word, you had to look it up in several different volumes. The need for a second edition was eventually clear, and in 1989 it was published with the help of the electronic storage and retrieval available at the time. These twenty volumes were heralded as a masterpiece, but Brewer explains that it was more than anything a marketing triumph. The second edition had essentially the same content that had gone before, but was a financial step in bringing forth a complete revision.
That revision is in progress. There may be a third edition printed, maybe twenty or thirty years from now, but maybe a print edition will never emerge. Brewer's book ends with a review of the latest version, the _OED Online_, launched in 2000. I can confirm that this really is the best _OED_ ever from a user's view; the subscription fee is daunting, but plenty of libraries can get you on for free, and the website is much more fun to use and browse around than was my old microprint _OED_, the one with the inescapable magnifying glass. Brewer, who does research on the _OED Online_ to investigate the _OED_ itself, explains that in the web format, "_OED_ was poised to escape the tyranny of alphabetization." Searching does not have to go by alphabet, but can be done electronically by dates, suffixes, etymologies, cited authors, and more. Also, there is plenty of space and no worry about how many column inches would be used up by cramming in a new word or new quotation; this is a fundamental difference in how the dictionary is put out. Revised chunks are issued every quarter, the full revision-in-progress including even changing the wording of definitions, replacing "the late nineteenth- or early twentieth-century locutions that now look quaint..." Even in the online age, however, the dictionary-maker's "dull plodding... is still an inextricable part of the lexicographical process." Brewer gives fascinating details about how that process has changed, and stayed the same, over the past century.