[black_knight]

Surfing on a smart phone is a real pain. Pages take longer to load than in the 90s and contrary to the 90s you can't start reading before it all has loaded.

Java tried to be C++, but run on every machine. That turned out to be difficult. But I don't think the author is thinking of Java. I guess he more has in mind domain specific languages, which are abstract enough in nature to be executed faithfully on any system with given capabilities.

Security also goes hand-in-hand with this form of abstraction. If the language can only express safe actions, the program will not be malicious. In pure languages, such as Haskell, one can use type-guarantees to enforce these restraints. One could imagine a virtual machine with this kind of typing.

[JadeNB]

> If the language can only express safe actions, the program will not be malicious.

Because 'safe' is not well defined, I can't argue rigorously against this, but it seems like the sort of thing that falls afoul of Rice's theorem (https://en.wikipedia.org/wiki/Rice%27s_theorem): for most reasonable definitions of 'safe', you can have a proveably safe language or you can have a Turing-complete language, but not both.

[black_knight]

Rice theorem is not really applicative here. It is a beautiful theorem, but it has to do with what we can compute, while "safe" has to do with what one can access.

A function taking Integers to Integers in a pure language cannot do I/O, and thus is "safe" to run, in the sense that I can be guaranteed it does not contain a trojan making my computer into a peer on a botnet. This is true, even if I allow it to compute any computable function.

Computational expressiveness and "safeness" are in a sense orthogonal. And just as I don't think it is always appropriate for any function to do I/O, I am not convinced all functions should be able to perform any computation. But that's a different discussion.

Regarding definedness of the term "safe", I would say it is defined by your threat model. It not an absolute term, but dependent on context.

[Boldewyn]

Of course, if the language of a program is stripped down, the interpreter can be made much safer. But then, we gain nothing from trying to shove down the users’ throats other platforms compared to, say, WebAssembly, which _will_ be shipped in almost all browsers in a year or so.

Just look at the long history of people trying to bring Python in the browser, or the fight to get Java applets _out of the browser_ again.

Basically I read the article as “scrap that web thing, and just begin from scratch”. And this is not a worthwhile path to go for many reasons.