[noelwelsh]

Most of the freely available technical books I'm familiar with (programming, machine learning, and maths) don't appear to have had an editor and don't seem to suffer from it. Perhaps it's because the people who write them (mostly academics) typically write a heck of a lot. I agree editors are useful but I don't think they're essential. If an author felt they needed one, they could be hired without going through a publishing house as an intermediary. Either way I don't think this point significantly alters my argument.

[tel]

Have you ever seen the process transforming a set of lecture notes to a textbook? That process is almost inconceivable for the professor (who has teaching and research engagements) to complete alone. Often they have a lot of help from graduate students, but not always.

Worse, I've read a great deal of math books that were unedited. Arguments end up in the wrong orders, sections are missing, notation changes without notification or regard for readability, numbering fails to match up (not even LaTeX can help if you change your labels midway), figues go missing, indices are wrong, exercises are mislabeled (or accidentally impossible!).

Even with an editorial team there's often a hefty and vital errata for most major textbooks. It's just a lot of organization.

I don't think editors are going anywhere for a long time. The role they serve is crucial. I do totally agree that they don't need to be tied to publishing-as-a-process though.

[jamesgeck0]

The most popular online Python book of years past, Dive Into Python, needed an editor badly.

Other online books have had technical editing done by the readers.

[unreal37]

I've been an editor on programming books. Believe me, most programming books have two editors - one for the spelling and grammar side, and one for the technical side.