ZeroMQ and its descendants will outlive us all. Pieter, you should be proud to have achieved so much.

Actually, this is a great opportunity for YC alums to deliver some of PG's knowledge to non-alums.

I look forward to discussing my ideas with PG-bot.

I consider this to be a failure of a GPL-licensed open source software project. Samba is not being used by Solaris, nor by Mac OS X.

Developers who use the GPL don't want as many users as possible; they want freedom and users can take it or leave it.

...and they have mostly left it in this case. Fine; but lets not pretend open-source GPL projects are significant to the world at large; they are backroom projects with no future.

are you kidding me? gcc?

I use gcc. Its a constant struggle compared to commercial, supported compilers.

Upvoted, as I agree with you philosophically, that the GPL is not exactly a corporate developer's best friend.

However, neither Solaris nor Apple are typical corporations in terms of using licensed 3rd party code; given the terms the GPL poses, wouldn't you expect to see more smaller companies, or companies that operate primarily on the web to adopt GPL'd projects?

prewett, which are the tools that you think already exist, exactly?

Many Mozilla developers make good use of valgrind. Leaks that can be found that way often are, and rarely cause problems for long.

No, UncleOxidant, it is definitely not "especially true for C++". There are plenty of powerful languages out there, many with more subtleties than C++, and longer roads to mastery.

No one disputes that C++ has a long road to mastery, but there are plenty of other languages for which that is true. So, C++ is not special in that way.

What languages do you have in mind?

Haskell probably has a long road to mastery. That said I think it might still be shorter than that for C++. it is hard to think of a language that takes longer than c++ to master. Interesting!

I disagree. Haskell is more difficult to master. There's far more to it than C++. I am highly proficient in both languages, it so happens.

"I disagree. Haskell is more difficult to master"

Fair enough. I found haskell easier than c++ to grok. It may have helped that I had a good grasp on type theory before I learned Haskell (worked through TAPL a few years ago). Subjectively, I found the various components of Haskell almost always fit together in very logical fashion with a very clean syntax, while C++ felt arbitrary, (with exceptions to every rule and exceptions to those exceptions and exceptions to those). I share your intent of not starting a language war. Just expressing my experience.

Of course, there can be no objective measure easily made, and I do not want to engage in language wars.

That said, I am a C++ expert, but contend that each of Lisp, Haskell, and Scala are far deeper languages, with more involved in their mastery than C++.

Learning to program well in language X can actually be done using language Y, for some X's and Y's, especially when Y has the features of X.

So, I continue to argue that asking precisely for language X may well exclude better applicants - ones smart enough to program in Y instead of X!

Obviously, I disagree. Such a filter will exclude precisely the wrong candidates. That is my argument in the article.

kailoa makes a good point: removing keywords from the job posting risks not appearing in a potential employee's search.

I think that the trick is to have "just the right" keywords, and no more. The particular example that I give in the article has too many. So, fewer is better, in that case.

I liked The Haskell School of Expression. It was a good first Haskell book. Read it, but also learn by doing, of course.

The walk score for my home was only 30, and I walk everywhere. I don't even own a car. So, the score is not meaningful.

Cool idea, but I agree that the implementation needs work. The score for my office is 68, but it is really not walkable at all. Whereas the score for my home is only a 62, but I never need to use a car at home.