I must admit, I do have a great deal of experience with many programming languages, many frameworks and libraries, many operating systems, and many platforms. Some of this experience predates the web.
I have indeed worked on multiple large JavaScript code bases. Some of this work was in the late 1990s (using Netscape Enterprise Server's support for server-side JavaScript, unfortunately), and some of it within the past few years.
Having worked on large JavaScript-based systems, in addition to ones written in languages like C, C++, Java, C#, Fortran, COBOL, PHP, Erlang and several proprietary languages/platforms, the problems with JavaScript become excruciatingly clear. The wider your experience, the more obvious it becomes how terrible JavaScript is.
The problems with JavaScript, especially for any sizable development projects, run far deeper than merely needing to stick to some "good practices". It just doesn't provide the infrastructure needed for anything but small-scale development. Its many flaws render it further inadequate for any serious use.
Now, that won't stop some people from trying to push it far beyond what it's capable of. The results are never good, however, and generally inferior to what could be obtained when using other technologies.
I love JavaScript, but there's no denying that large aspects of the language are somewhat ill-conceived. Crockford himself says this.
Once you know which pitfalls to avoid it's possible to write some really elegant code, but the mere presence of those pitfalls causes a ton of problems when you bring in engineers from other programming languages.
I find it kind of concerning that when someone criticizes JavaScript, everyone's first response is, "he must not have any experience with the language." That's an unfounded personal attack and ignores the crux of the argument.
Look, we love JavaScript, but it doesn't do us any good to pretend it's perfect. It'd be much easier to ramp up new engineers on the language if we admitted its flaws and made efforts to educate "the right way."
"crockford himself" has done great things fo Javascript but he's also not the last word on it. He himself admits that some of his ideas about JavaScript were wrong and resulted from his pre Javacript baggage. E.g. The model he suggests for conventional inheritance in "The Good Parts" does not, iirc, actually work as advertised and in any event isn't needed.
The interesting thing to me about JavaScript is that insofar as it lacks "crucial" language features it turns out to provide the necessary tools to build them yourself. Exactly gow do you want inheritance or modules or whatever to work? E.g. While JavaScript doesn't impose a module structure it's very easy to come up with one, and because you build it yourself you can tweak it to deal with your exact needs rather than adhering mindlessly to "patterns" developed for working around the shortcomings of more tightly defined languages (Java, cough)
Java programmers moving to JavaScript seem to write horrible code because the freedom to actually just do stuff rather than build boxes with tiny weird-shaped holes cut in them to package pieces of functionality is so alien.
I wish Python, with its (fewer) warts were as ubiquitous and useful as JavaScript, but it's not, and Javascript is pretty darn good.
I'm not disagreeing with you in any way, shape or form. JavaScript is an AWESOME language for both productivity and results. But I think we're doing ourselves a disservice if we assume that anyone jumping right into the language is going to immediately understand it.
Like you said, bad code is somewhat inevitable when you drop a classically-trained OOP programming into JavaScript. And that specifically is not a fault of the language. But I think we can all do a better job of pointing newbies in the right direction.
I have indeed worked on multiple large JavaScript code bases. Some of this work was in the late 1990s (using Netscape Enterprise Server's support for server-side JavaScript, unfortunately), and some of it within the past few years.
Having worked on large JavaScript-based systems, in addition to ones written in languages like C, C++, Java, C#, Fortran, COBOL, PHP, Erlang and several proprietary languages/platforms, the problems with JavaScript become excruciatingly clear. The wider your experience, the more obvious it becomes how terrible JavaScript is.
The problems with JavaScript, especially for any sizable development projects, run far deeper than merely needing to stick to some "good practices". It just doesn't provide the infrastructure needed for anything but small-scale development. Its many flaws render it further inadequate for any serious use.
Now, that won't stop some people from trying to push it far beyond what it's capable of. The results are never good, however, and generally inferior to what could be obtained when using other technologies.