To some academia is about publishing… lots of journal articles that is. Which may have been nice when the only way to gain information was through the journal subscription you received in the mail once a month. Now there are more academics and more journals, and more pressure to publish lots. Looking back on three decades of computer science articles, there are amazing articles to be found, often in smaller journals, or older issues no longer viewed because they are deemed “old”. But such journals often contain excellent articles which have barely seen the light of day. Techniques discussed, but never tried out. Other articles have faired better of course. Some have made it into textbooks consistently for 20 years, are well cited and often used… but *may* not be the most appropriate. We use them because others do. Sad really, but it goes to prove a point. No matter *how* good academics think journals are, most articles will barely be looked at. The articles most referenced may be those whose implementation is available to all, or are easy to reproduce. Code gets pushed about becuase it exists, and other algorithms are left to linger in the vortex of time. Publishing in a journal guarantees nothing. Writing a book is better, but with the high cost some publishers attach to books, I do wonder who buys them.