Can't We Have Anti JavaScript-Only-Sites Search Engine? | Hacker News Reader

Can't We Have Anti JavaScript-Only-Sites Search Engine?

8 pointsvezycash14 comments

Am I the only one who wants a search engine that will prioritize sites that with with Javascript Off?

And also can be filtered for pages size say under 1mb

Won't this be a better solution than AMP?

I use ublock origin with scripts and large media blocked by default (which makes sites load faster, like AMP is supposed to as well).

For some sites, depending on their purpose, it would be unreasonable to expect them to not use JavaScript. For example https://www.openstreetmap.org (where the scripts are even free/libre).

In such a case, it is easy to (permanently) unblock scripts and large media in ublock origin, but it would not be fair for a search engine to deprioritise a map site for that reason, would it?

Also, most news sites pull in massive pile of JS, but they still work when it's blocked. Yet, even some blogs can fail to show any text when JS is blocked. So I feel the issue is not so much about JS alone, more about graceful degradation.

Because surely it's fine to have add a bunch of non-existential JS? Salon has a Monero miner on their site[0], but works fine without it being loaded.

[0]: https://arstechnica.com/information-technology/2018/02/salon...

s/non-existential/non-essential/

I partially agree. What I want is for stuff that is only available with JavaScript to not be indexed; if it contains some text that is always available and some that is added due to JavaScripts, only the text that is part of the downloaded HTML file should be indexed. (Pages that are empty except for the JavaScript code would not be indexed at all, although if there is a page that describes and then links to it, and the description does not require JavaScripts, that page would still be indexed, so you could still find it.)

For some webpages it is useful to have some JavaScripts, but most ones with just text or whatever, you should not need it. Even ones that use it just to access its internal data, should still have descriptions that do not need JavaScripts (whether you want to access the data in a different way or just read the description only, or whatever other purpose you might want to do).

As hjek has mentioned below, sites that clearly need JavaScript (e.g. interactive street map) shouldn't be de-prioritised.

However I definitely agree that it would be nice to be able to filter out sites that you would expect to just contain plain text and images (e.g. blogs, news, product pages), yet show a white screen or infinite loading animation when viewed with JS off.

Yes, Javascript can be annoying. But what matters is the quality of the content that is being searched, not if a website has annoying Javascript or not.

Would you rather read a really bad article that has no JS or a very well written one with JS enabled?

> Would you rather read a really bad article that has no JS or a very well written one with JS enabled?

That's not always the choice though. There are lots of good articles, so you could ask: Would you read one good article that requires you run a (non-free) program in your web browser, or would you rather read another good article that doesn't require that? (or simply read the article elsewhere, if possible)

Presumably OP doesn't have an issue with sites that have "JS enabled", but sites that don't work at all without JS. For example Wiki Wiki Web[0] fails to display anything but a spinning loader animation when JS is disabled.

[0]: http://wiki.c2.com/

In the case of the c2 wiki, you can still download the raw text and write your own parser, at least.

Although I agree with the issue being sites that don't work at all without JS enabled. (That is why I suggested ignoring JavaScript when indexing.) (The c2 wiki should be made to work without, or at least, to add a <noscript> which links to the raw text, perhaps.)

> at least, to add a <noscript> which links to the raw text, perhaps.

Yes, even just that would make all the difference.

Apart from this obvious problem, there is also a problem of freedom: https://www.gnu.org/philosophy/javascript-trap.html

I am not an expert, but maybe free search engine YaCy[0] could be modified for that purpose.

[0] https://yacy.net

Oh man my buttons use JS just to log in. But what is a way around this? Making logins with <form></form>?

Exactly. Just like the site you're logged into now.

Code your <form> login as part of the page dom, visible on load, then add your javascript to hoist it out, hide it, and stick it into a modal dialog or whatever you'd do currently with javascript.

The term you're looking for is 'progressive enhancement'

Reply on news.ycombinator.com