4 ms·
> The problem comes not from the language JavaScript, but from the fact that browsers, by default, will pull in JavaScript code that gets sent to them and run i
by aidenn0 4y ago
> The problem comes not from the language JavaScript, but from the fact that browsers, by default, will pull in JavaScript code that gets sent to them and run it to do anything at all. Emacs is supposed to defend your freedom. It's supposed to help you to defend your freedom, and lead you to defend your freedom, which means it shouldn't lead you to throw your freedom away as soon as you visit a site that tries to send you a non-free program to run straight off of that other machine. So it's important not to lead users to do computing this way.
I don't know how this is related to extending Emacs with JavaScript, but it's more or less correct about how JavaScript works in the browser (modulo "anything at all" being "anything at all in the browser sandbox"). If you only want to run free software, you necessarily must disable JS in your browser.
- BeetleB 4y agoTechnically, elisp can do the same.
- dubya 4y agoThe "by default" is an important distinction.
- BeetleB 4y agoThat's not a language issue, but a platform (i.e. browser) issue. It's entirely possible to enable JS and not have this as the default. In Emacs, you can install any package that will then do most/all the evil that JS can. In principle, you should be very careful about the packages you install.
- dubya 4y agoThat's what he said, more or less: "The problem comes not from the language JavaScript, but from the fact that browsers, by default, ..." I think it's the cultural issues he's getting at. Minified or otherwise obfuscated elisp packages would rightly be viewed with suspicion. Unauditable JS is the norm, right?
- natrys 4y agoHe had devoted a part of the talk about browsers too, starting with: > But one thing I think we really shouldn't have is the equivalent of a modern web browser. I think people are focusing on the Javascript part alone too much, when he made it clear numerous times that he is talking about the browser model, not javascript language itself.
- masukomi 4y agobut... if he doesn't want the equivalent of a modern browser, then downloading and executing JS isn't really an issue, and thus the whole JS=BAD because you download it and execute it goes out the window. I don't get it. Like, i don't care if we can write emacs stuff in JS or not, but i don't see how it's any more dangerous than Lisp.
- cnity 4y agoI agree, his reasoning is strange: JS is used in other contexts to compromise user control. We want user control. Thus we will avoid Javascript in this context.
- natrys 4y agoI mean he is only saying JS=BAD in the context of downloading and executing obfuscated non-free javascript. The scenario where JS isn't downloaded and executed (e.g. outside browser) isn't even being what's talked about here much less being also lumped into the BAD group, so it's weird that people here are making this about JS at all, when it's clearly not.
- krapp 4y agoI don't understand the argument that Javascript isn't free. Every library I'm aware of is MIT or otherwise FOSS licensed, and (at least before Node.js and NPM came along) it isn't owned or licensed by a specific corporation. It's probably the most free programming language ever created. It's RMS, so I know he's going to throw the baby out with the bathwater on this, then throw out the bathtub, then disassemble the plumbing, brick up the bathroom and never bathe again on principle, but Javascript has probably served as a greater benefit for FOSS culture than any other language. Certainly more so than lisp. Making it out to be an implicitly evil language seems... out of touch? quick edit: I know the argument about privacy invading javascript software, but software can be free and still invade your privacy. That's got nothing to do with the language.
- teddyh 4y agoJust to explain, the non-free code JavaScript code he is referring to is the inline JavaScript code on individual web pages (which in turn often uses the large MIT-licensed libraries).
- zelphirkalt 4y agoThe JS mess has brought us megabyte huge blobs of minified JS that does who knows what, slowing down getting information to a crawl compared to what could be. Not all JS is like that, but a lot of people are actively participating in creating that mess.
- Unrectified 4y agoI do think reading https://www.gnu.org/philosophy/javascript-trap.html https://www.gnu.org/philosophy/javascript-trap.html is helpful to understand his position. Specially if you're familiar with his 'four freedoms' concept
- brenns10 4y agoI'm not a fan of RMS, but as you noted, his blurb in the talk wasn't "Javascript the language is not free", or even "Javascript the ecosystem is not free", it's about the privacy invading factor of JS in a modern web browser. Since, as you said, software can be free and still invade your privacy. He just doesn't want this GNU project to include support for a language whose primary purpose seems to be invading privacy -- even if it has a lot of great secondary purposes and a good open source ecosystem. Given that the GNU/FSF projects are all about advocacy in the area of software freedom and privacy, it seems like a reasonable perspective to have. Yeah, RMS definitely takes things to an extreme, but I don't see the "JS is implicitly evil" thing you're calling out, just refusing to appear to endorse or support something that goes against his principles. And of course ironically enough those very principles mean he (should be...) perfectly happy with the emacs-ng fork, since that's part of the freedom he's advocating for.