5 ms·

HackerNews is starting to become less and less useful for reading tech news as of late. I come here to read stuff that's interesting in tech/science/etc, becaus

HackerNews is starting to become less and less useful for reading tech news as of late. I come here to read stuff that's interesting in tech/science/etc, because the signal-to-noise was great. The vast majority of comments on this article are complaining about a single thing: a bluetooth API behind permissions check, and then spilling over into political arguments about what the Web should be (documents only!) and hyperbole (e.g. "i'll be bombarded by permission requests" -- lots of Web APIs can ask for permissions already like the Mic and Webcam, why isn't there an epidemic of permission spam from those, but you expect BT?) For example, if I hadn't read the link in detail, I wouldn't have even known about the Web Authentication API extension (https://developer.mozilla.org/en-US/docs/Web/API/Web_Authentication_API https://developer.mozilla.org/en-US/docs/Web/API/Web_Authent...) to the Credential Management API. Improving authentication on the Web should be an interesting topic for HN, but no one even mentioned it. The ship has sailed, we're 20+ years into this open ecosystem of ephemeral, mobile code execution, with Web Assembly poised to take it even further. We're not going back to documents-only, so either we improve the Web, it's security, performance, and capability, or we freeze in place, and hand our lives over to the App Stores of the world (the security world of native-platforms with side-load app capability isn't very impressive at this point)

Grazester 8y ago
I absolutely feel the same way and was thinking about this when I read the comments. Its like we have become old and disenchanted. We no longer see the the positive potential of an underlying piece of technology(or even the technology itself) but see ways to further exploit us.
- dylan604 8y ago
  Somebody has to though. If we just blindly accept things, then we end up in a nightmare we have with unsecure IoT devices. It should be accepted that if something can be exploited, it will be exploited. As developers, we shouldn't be blind of that.
  - krapp 8y ago
    But almost no one goes any further than complaining about the modern web and javascript and the existence of "web apps." If these tangents came with constructive ideas about how to improve the situation, that would be fine. But the only suggestions people seem to offer are to abandon the web as a lost cause or break the web somehow. And heaven forbid anyone mention Webassembly without someone predicting that the entire web will inevitably be converted into closed proprietary WASM blobs and HTML will be deprecated... for reasons. It's getting unreasonable and counterproductive. I feel like if some people had their way, computing would have never advanced beyond the terminal, or at best, at the state of the art in 1990.
    - eitland 8y ago
      > If these tangents came with constructive ideas about how to improve the situation, that would be fine. But the only suggestions people seem to offer are to abandon the web as a lost cause or break the web somehow. Some of us are suggesting constructive alternatives: - test your site in all major browsers - make sure it works without JavaScript and is readble without styling. Thst is unless you are creating a frontend application. We've been complaining about this since IE6 and we had it more or less nailed for a while around web 2.0 until enough devs started thinking - it was enough to target Chrome and - JavaScript is available everywhere. There's no reason why most blogs and news sites should need JavaScript. And there is no reason why Chrome monoculture should be more acceptable than IE6 monoculture.
    - dylan604 8y ago
      Regardless of the industry, bad actors always ruin it for the rest of us. I am familiar with the saying of "don't bring me problems, bring me solutions". However, not every problem has an immediate solution, or just because I found the problem doesn't mean I understand it or even in the correct place to fix it. I would think that applies to most people finding issues and trying to report it as well. What I find counterproductive is that we know there are bad actors out there just waiting for the next thing to come out that they can exploit, yet very little is done to counter act it during development. Yes, develop the thing, but before releasing it into the wild, maybe do a little pen-testing on it before releasing the next delivery device for the whatever ill will someone might use it for. Plus, this is Hacker News where part of the group are the ones cracking things open to see what's under the hood while another part of the group are the ones building the next "thing". Keeping these kinds of articles front of mind might actually make one of these devs actually think about adding some extra security.
xg15 8y ago
> ...so either we improve the Web, it's security, performance, and capability, or we freeze in place, and hand our lives over to the App Stores of the world Considering the most-important browser and the second-most important app store are made by the same company, with the company having about the same capabilities of control in both, I wonder if there is really that much of a difference anymore.
- bun_at_work 8y ago
  Wow. You completely missed the point OP was making and instead did what he was complaining about. Incredible.
sanderjd 8y ago
I think my problem is that the comments are so often predictable. I wouldn't mind if it were annoying niche complaints from a unique perspective, but this static web discussion comes up all. the. time.
- cromwellian 8y ago
  Exactly. You can read the repetitive arguments over a document-web vs an application-web everywhere, just like I can read Android vs OSX, or Emacs vs VI everywhere. This has been going on for decades with no resolution. Often what I love about HN is someone in comments raises an interesting or unique point, and posts a link to some paper, article, or video that goes in-depth into the issue in a thoughtful way. More of that, and less of "FooFramework was announced? Sorry but do we need another FooFramework!?" style stuff.
sdegutis 8y ago
HN was really interesting because the entire purpose of the site was to engineer and moderate it in such a way that signal is inherently promoted and noise is downplayed. But the fascinating thing to me is that its users have evolved to post content that matches what HN's algorithms have considered "signal", despite it not actually being signal, and HN's algorithms and moderation haven't evolved with this statistically significant influx of faux-signal! The very goal of the site is put in danger by its own users learning how to use it, almost like natural-selection. Ironic!
- selestify 8y ago
  Is it even possible for any open community to be immune to natural selection effects?
tialaramex 8y ago
> The vast majority of comments on this article are complaining about a single thing: a bluetooth API behind permissions check, A HackerNews staffer wrote, some time ago: > We've reverted the title from “Chrome 70 supports Web Bluetooth on Windows 10”, which breaks the guidelines by editorializing. So that's your proximate cause. Somebody editorialized the article and the initial responses are to that editorialized title.
- cromwellian 8y ago
  Good point, and an excellent justification for the no-editorializing-policy! They should link that policy to this thread to show what happens to S/N when people editorialize for clickbait.
blub 8y ago
Perhaps fawning over the latest piece of technology and APIs has become unpalatable as more people have woken up to the idea that technology is neither politically nor morally neutral. That's a sign of the industry growing up: those working in it are starting to think about the consequences of what they are doing and are speaking up against misbehaving companies like Google. Who cares about some silly web authentication API when Google will engineer the thing to its maximal data collection potential anyway?
- cromwellian 8y ago
  Did any of the critics making unsubstantiated claims like this even bother to read the Working Group mailing lists where people are discussing the security and privacy requirements of these APIs? These specs are being developed in the open, with input from organizations like Mozilla and security engineers from all over. If you have a specific claim, you can reference the spec and make it. Also, the whole point of the "silly web authentication API" is to do away with passwords and usernames, and let users supply authenticated credentials, which would enable -- if desired -- creation of as many user credentials as you like, without having to tie everything to your email + password, which could make it a lot easier for people to have many different identities that can't be linked between sites, as well as make it easy to create 'throwaway' identities. This 'silly' API doesn't increase data collection, it simply fixes some of the biggest problems with web security: passwords and 2-factor authentication. This is the tech equivalent of the following fallacious kneejerk reasoning you see in anti-vaxxers: "big pharmaceutics/Monsanto is evil, therefore since big pharma also makes vaccines, any new vaccines that comes out is bad because they just want your money." This kind of knee-jerk conspiracy theory stuff doesn't help public health and isn't linked to any specific claim. If you've got a specific technical critique of a new spec, you should make it, but to dismiss a new security API as "silly" is a non-sequitur.

Grazester 8y ago

I absolutely feel the same way and was thinking about this when I read the comments. Its like we have become old and disenchanted. We no longer see the the positive potential of an underlying piece of technology(or even the technology itself) but see ways to further exploit us.

dylan604 8y ago
Somebody has to though. If we just blindly accept things, then we end up in a nightmare we have with unsecure IoT devices. It should be accepted that if something can be exploited, it will be exploited. As developers, we shouldn't be blind of that.
- krapp 8y ago
  But almost no one goes any further than complaining about the modern web and javascript and the existence of "web apps." If these tangents came with constructive ideas about how to improve the situation, that would be fine. But the only suggestions people seem to offer are to abandon the web as a lost cause or break the web somehow. And heaven forbid anyone mention Webassembly without someone predicting that the entire web will inevitably be converted into closed proprietary WASM blobs and HTML will be deprecated... for reasons. It's getting unreasonable and counterproductive. I feel like if some people had their way, computing would have never advanced beyond the terminal, or at best, at the state of the art in 1990.
  - eitland 8y ago
    > If these tangents came with constructive ideas about how to improve the situation, that would be fine. But the only suggestions people seem to offer are to abandon the web as a lost cause or break the web somehow. Some of us are suggesting constructive alternatives: - test your site in all major browsers - make sure it works without JavaScript and is readble without styling. Thst is unless you are creating a frontend application. We've been complaining about this since IE6 and we had it more or less nailed for a while around web 2.0 until enough devs started thinking - it was enough to target Chrome and - JavaScript is available everywhere. There's no reason why most blogs and news sites should need JavaScript. And there is no reason why Chrome monoculture should be more acceptable than IE6 monoculture.
  - dylan604 8y ago
    Regardless of the industry, bad actors always ruin it for the rest of us. I am familiar with the saying of "don't bring me problems, bring me solutions". However, not every problem has an immediate solution, or just because I found the problem doesn't mean I understand it or even in the correct place to fix it. I would think that applies to most people finding issues and trying to report it as well. What I find counterproductive is that we know there are bad actors out there just waiting for the next thing to come out that they can exploit, yet very little is done to counter act it during development. Yes, develop the thing, but before releasing it into the wild, maybe do a little pen-testing on it before releasing the next delivery device for the whatever ill will someone might use it for. Plus, this is Hacker News where part of the group are the ones cracking things open to see what's under the hood while another part of the group are the ones building the next "thing". Keeping these kinds of articles front of mind might actually make one of these devs actually think about adding some extra security.

xg15 8y ago

> ...so either we improve the Web, it's security, performance, and capability, or we freeze in place, and hand our lives over to the App Stores of the world Considering the most-important browser and the second-most important app store are made by the same company, with the company having about the same capabilities of control in both, I wonder if there is really that much of a difference anymore.

bun_at_work 8y ago
Wow. You completely missed the point OP was making and instead did what he was complaining about. Incredible.

sanderjd 8y ago

I think my problem is that the comments are so often predictable. I wouldn't mind if it were annoying niche complaints from a unique perspective, but this static web discussion comes up all. the. time.

cromwellian 8y ago
Exactly. You can read the repetitive arguments over a document-web vs an application-web everywhere, just like I can read Android vs OSX, or Emacs vs VI everywhere. This has been going on for decades with no resolution. Often what I love about HN is someone in comments raises an interesting or unique point, and posts a link to some paper, article, or video that goes in-depth into the issue in a thoughtful way. More of that, and less of "FooFramework was announced? Sorry but do we need another FooFramework!?" style stuff.

sdegutis 8y ago

HN was really interesting because the entire purpose of the site was to engineer and moderate it in such a way that signal is inherently promoted and noise is downplayed. But the fascinating thing to me is that its users have evolved to post content that matches what HN's algorithms have considered "signal", despite it not actually being signal, and HN's algorithms and moderation haven't evolved with this statistically significant influx of faux-signal! The very goal of the site is put in danger by its own users learning how to use it, almost like natural-selection. Ironic!

selestify 8y ago
Is it even possible for any open community to be immune to natural selection effects?

tialaramex 8y ago

> The vast majority of comments on this article are complaining about a single thing: a bluetooth API behind permissions check, A HackerNews staffer wrote, some time ago: > We've reverted the title from “Chrome 70 supports Web Bluetooth on Windows 10”, which breaks the guidelines by editorializing. So that's your proximate cause. Somebody editorialized the article and the initial responses are to that editorialized title.

cromwellian 8y ago
Good point, and an excellent justification for the no-editorializing-policy! They should link that policy to this thread to show what happens to S/N when people editorialize for clickbait.

blub 8y ago

Perhaps fawning over the latest piece of technology and APIs has become unpalatable as more people have woken up to the idea that technology is neither politically nor morally neutral. That's a sign of the industry growing up: those working in it are starting to think about the consequences of what they are doing and are speaking up against misbehaving companies like Google. Who cares about some silly web authentication API when Google will engineer the thing to its maximal data collection potential anyway?

cromwellian 8y ago
Did any of the critics making unsubstantiated claims like this even bother to read the Working Group mailing lists where people are discussing the security and privacy requirements of these APIs? These specs are being developed in the open, with input from organizations like Mozilla and security engineers from all over. If you have a specific claim, you can reference the spec and make it. Also, the whole point of the "silly web authentication API" is to do away with passwords and usernames, and let users supply authenticated credentials, which would enable -- if desired -- creation of as many user credentials as you like, without having to tie everything to your email + password, which could make it a lot easier for people to have many different identities that can't be linked between sites, as well as make it easy to create 'throwaway' identities. This 'silly' API doesn't increase data collection, it simply fixes some of the biggest problems with web security: passwords and 2-factor authentication. This is the tech equivalent of the following fallacious kneejerk reasoning you see in anti-vaxxers: "big pharmaceutics/Monsanto is evil, therefore since big pharma also makes vaccines, any new vaccines that comes out is bad because they just want your money." This kind of knee-jerk conspiracy theory stuff doesn't help public health and isn't linked to any specific claim. If you've got a specific technical critique of a new spec, you should make it, but to dismiss a new security API as "silly" is a non-sequitur.