Hacker News new | ask | show | jobs
by bastawhiz 1049 days ago
Let's say I pay for Kagi. Kagi is a tool that I'm using to avoid doing hard work manually. With relatively few exceptions, I can probably accomplish what I use a search engine for manually, but with much more time and effort. So I'm paying for a tool to assist me with my use of the web. A "user agent", you might even say.

It simply doesn't sound right to say which tool a user can use. It's literally the same as arguing that you should be able to block Firefox from accessing your website and it's Mozilla's fault that they don't respect your wishes as a webmaster to block Firefox exclusively. Or that a VPN doesn't publish its IP addresses so that you can block it. Or a screen reader that processes the text to speech in a way that you disagree with.

Philosophically it seems intuitive to say "I should be able to block a third party that is abusing my site" but it's ignoring the broader context of what "open web" and "net neutrality" actually mean.

I run a service for podcasters. There are podcast apps and directories that either ignorantly make unnecessary requests for content or have software bugs that cause redownloads. I could trivially block them, but I don't because doing so penalizes the end user who is ultimately innocent, rather than the badly behaved service operator. The better solution is primitives like rate limiting, which I use liberally. Plus, blocking anyone literally has a direct effect of incentivizing centralization on Apple, Spotify, etc. and making the state of open tech in podcasting even worse.

> the Brave search API allows you (for an extra fee) to get the content with a "license" to use the content for AI training? Who allowed them the right to distribute the content that way?

I don't think there's any court at this point that would back you up that freely published content annotated with full provenance cannot be scraped and published for a fee. Services like this have existed for decades. If you don't want your content scraped, put it behind a login. Especially considering this only applies when you allow other search engines and if you think Google and Bing aren't using your content to train AI, you're off your rocker.

1 comments

> With relatively few exceptions, I can probably accomplish what I use a search engine for manually, but with much more time and effort. So I'm paying for a tool to assist me with my use of the web. A "user agent", you might even say

1. User agents should identify themselves

2. A crawler is not a User agent - it's an agent for Brave

>I don't think there's any court at this point that would back you up that freely published content annotated with full provenance cannot be scraped and published for a fee.

You can't end-run copyright like this: just because something is publicly available doesn't mean anyone can redistribute it. Look at the legal issues & cases relating to Library Genesis.

> User agents should identify themselves

There is no rule that this is true, and many user agents exist _specifically to not be identified_. See Tor and other privacy-centric user agents.

> A crawler is not a User agent - it's an agent for Brave

You know, I thought "what does Wikipedia have to say on this matter?" and sure enough:

> Examples include all common web browsers, such as Google Chrome, Mozilla Firefox, and Safari, as well as some email readers, command-line utilities like cURL, and arguably headless services that power part of a larger application, such as a web crawler.

I can't even make that up.

> just because something is publicly available doesn't mean anyone can redistribute it

You're mistaking reselling content with providing access to it. By your logic, caching proxy servers would be illegal on the grounds of copyright. The physical act of downloading files necessarily creates copies of the data every step of the journey from the source server to you. There's a material difference between paying someone for a copy of some content and paying someone to fetch content for you on your behalf. Nothing about copyright law specifically requires the person physically acquiring the content is the one who ends up consuming it.

Downloading something isnt redistributing it. It is your website. You provide what is on it to me. I send you an HTTP request. You dont have to respond. You do. I am not copying anything. Copyright simply isnt engaged at any point in this process.