These questions seem pretty good!

In the max heap question, I'd argue that add should take a parameter by value, rather than a const reference, since this prevents having to needlessly copy temporaries (and is less verbose than overloading for T&&).

List of iterator categories is missing input iterators.

To that end, I highly recommend the command line utility 'sl', which runs an unskippable ASCII animation of a steam locomotive. Makes for great fun explaining it to the flummoxed coworker borrowing your computer!

Notably, the "al" in algebra and algorithm really is the Arabic 'al-'— algebra and algorithm are latinized forms of al-jabr and al-Khwarizmi respectively.

That seems pretty harsh. From context, 'been there, done that' sounds to me like sympathizing with the article's author, not dismissing them.

Calling Scheme a "new language with a strong Lisp influence" is a bit rich. It's certainly not new, and in many respects it's more representative of Lisps-of-Old than Common Lisp.

Lisp started in 1958. Popular Lisps in the 60s/70s were Lisp 1.5, Lisp 1.6, Interlisp, Maclisp and Standard Lisp.

Scheme grew out of research from 75-80 (the Lambda Papers), investigating Actors and bringing ALGOL and Lisp together.

Thus Scheme is the newer language.

>it's more representative of Lisps-of-Old than Common Lisp.

Common Lisp contains a direct core of the original McCarthy Lisp. One can develop in Common Lisp in the typical styles of the 60s and 70s. With PROGs and GOTOs, symbols and property lists, dynamic binding, procedural macros, ...

Scheme was different: it introduced a more functional style, lexical scope, closures built-in, away from symbols and dynamic binding, ..., continuations, hygienic macros, TCO (recursion instead of direct iteration constructs), ... for some years people tried to bridge the gap (for example by providing a Scheme in Common Lisp, by reusing the concepts or by writing to a compatibility layer), but nowadays there is very little sharing.

"Historically, languages designed for other people to use have been bad: Cobol, PL/I, Pascal, Ada, C++."

In what sense was C++ designed for other people? IIRC Stroustrup came up with it to use on his own research projects.

C++ is the classic design-by-committee language. However it started it, it's far from one person's design now and hasn't been for decades.

Mehmed II? In any case a lot of the territory Suleiman conquered did remain part of the Ottoman Empire into the 20th century.

I think the intention is less "the tipping model makes more sense", and more "given that the tipping model is in effect, you shouldn't unfairly punish the housekeepers by not tipping them".

I don't think the tipping model makes sense at all, but as long as we're talking about people we pay to clean up after us, it's worth pointing out that hotel housekeepers have a particularly raw deal, and despite the fact that people frequently seem surprised to hear this, you are expected to tip them.

But how does that (semi-)relate to tptacek's other points?

Uncharitably, I'd say the old logo looked much more professional, and so didn't jive with Google's corporate ethos of "seem as much like a preschool as humanly possible". The new one definitely fits better in that respect.

When you're collecting the world's data, it's in your best interest to come off as cuddly as possible. It works.