Ciro Santilli @cirosantilli 37

The idea of writing the history pleased me very much: I felt that historical justice had never been done to the Carthaginians. I spent my leisure time in making a study of the ruins of the Old City, with the help of a contemporary survey, and familiarizing myself with the geography of the country in general. I also learned the rudiments of the language well enough to be able to read simple inscriptions and understand the few Phoenician words used by authors who have written about the Punic Wars from the Roman side. When I returned to Italy I began to write the book concurrently with my Etruscan history. I like having two tasks going at the same time: when I tire of one I turn to the other. But I am perhaps too careful a writer. I am not satisfied merely with copying from ancient authorities while there is any possible means of checking their statements by consulting other sources of information on the same subject, particularly accounts by writers of rival political parties. So these two histories, each of which I could have written in a year or two if I had been less conscientious, kept me busy between them for some twenty-five years. For every word I wrote I must have read many hundreds; and in the end I became a very good scholar both of Etruscan and Phoenician, and had a working knowledge of several other languages and dialects too, such as Numidian, Egyptian, Oscan, and Faliscan. I finished the History of Carthage first.

It is conjectured that the algorithm runs in time $O(n^{\alpha })$ and space $O(n^{\alpha })$, where $\alpha =log(2)/log(3)\approx 0.631$