[brushfoot]

Even if you agree that Google should have to prop up Canadian news organizations—which I don't, but for the sake of argument—predicating that support specifically on hyperlinks seems like a terrible precedent to set. Any site of any size should be free to link wherever it wants, freely.

[tannhaeuser]

"Link tax" is clearly chosen in bad faith, like in other cases where that wording had been used. What's really happening is that Google Search shows previews of large parts of articles such that visitors don't go to origin sites to generate page impressions/ad playouts. Which is especially problematic since Google/Alphabet is either an ad provider on the skipped news site or its competitor.

[gspencley]

This is a popular myth about the way the law works in Canada. The myth may be based on how the Australian version works, I wouldn't know. But here in Canada, this is the relevant text of the bill:

> digital news intermediary means an online communications platform, including a search engine or social media service, that is subject to the legislative authority of Parliament and that makes news content produced by news outlets available to persons in Canada

"Making available" includes just links. In other words, there is no qualifier in the bill what-so-ever that requires that previews and snippets be extracted and provided. A single link with no text or any other content whatever qualifies as "making available."

Here is the entire text of the bill: https://www.parl.ca/DocumentViewer/en/44-1/bill/C-18/royal-a...

[Nextgrid]

> makes news content [...] available

That sounds like a simple link wouldn't count?

[gspencley]

I suppose I could have included other "relevant" sections of the bill. But there's a reason I linked to the full text.

> "Making available of news content (2) For the purposes of this Act, news content is made available if

(a) the news content, or any portion of it, is reproduced; or

(b) access to the news content, or any portion of it, is facilitated by any means, including an index, aggregation or ranking of news content."

So not only is there no qualifier saying that content needs to be reproduced, but (b) specifically says that "facilitating" access BY ANY MEANS counts as "making available."

[kevingadd]

Indexes don't work without consuming the news content first in order to build the index, right? So it's reasonable to argue that a news search index is a derived work that draws some amount of value from the source news content. Without the news sites offering their news for free to google's spider, google's index would be useless. You could make a similar argument for aggregators since they usually display excerpts and thumbnails.

Ranking seems harder to justify.

[ethbr1]

It does beg the cascading question too.

What if I create a search product that directs you to a site, based on information gleaned from another site?

Trivial example: a "Top 10 Topics in the News Today" list of Wikipedia links, based on scraping a news site's front page daily

On the one hand, "It's free info and scraping should be allowed." On the other hand, if everyone did that, it'd highjack all of the news site's visitors, depriving it of revenue and destroying the very resource it's built on.

[Spivak]

This line of reasoning is dangerously close to saying that mapmakers need to pay license fees for your land to be shown on a map because "without the land your map wouldn't be very valuable."

[SiempreViernes]

Sure, but the act only applies if you are as powerful as Google (Section 6: Application), and since they are the primary target the scope needs to be broad because Google can be counted on producing the most bad faith reading of any law to try to escape paying for the practice of scraping content from journalists in an effort to keep the traffic on their own pages.

[kevingadd]

The text of the bill says that links are "news content"? A pure hyperlink does not make content available, since if the target server goes down the link is useless. It just directs the user to the content.

Unless somebody loses a court case over hyperlinks, it feels extremely disingenuous to claim this law is going after hyperlinks when Google's content was blatantly more than hyperlinks (they provide excerpts).

[SiempreViernes]

The definition of "making available" is actually very broad, presumably to head off at the gate whatever technical work-around Google was planning to use to bypass a the more narrow "the news content, or any portion of it, is reproduced".

Like, I can see Google pivoting to providing custom links and a script that runs in the users browser that will dynamically produce the snippets on the pages without the content technically going trough their servers...

[gspencley]

A couple things of note:

1. Meta was the other "initial target." And so if we're in a mindset where we just dislike Google or "big powerful corporations that will stop at no end to ... whatever" then the application of the bill ALREADY hit a point where we can start to see how loosely the regulators will interpret Section 6. And according to the wording of that section, they just need to consider the "size" of the entity,"strategic advantage", and "prominent market position" in order to determine if they decide that the digital news intermediary has "a significant bargaining power imbalance."

2. With respects to what you think Google might do in order to try and "work around the law" ... I'm not a lawyer, but from what I've heard from lawyers, courts tend to be very intolerant towards people who try and apply a strict interpretation of the wording of a law in order to try and skirt around what it makes illegal. Common law, precedence and judicial interpretation really exists in large part to try and avoid that type of thing. Judges will look at things like the intent of the legislators, past court decisions and the intent of the accused in order to determine whether the accused is in violation.

In other words, Section 4 "Purpose", would be considered:

> Purpose 4 The purpose of this Act is to regulate digital news intermediaries with a view to enhancing fairness in the Canadian digital news marketplace and contributing to its sustainability, including the sustainability of news businesses in Canada, in both the non-profit and for-profits sectors, including independent local ones.

As well as Google's intent. If the court were to gather that Google's intent was to try and work around the law, they go back to the Purpose, look at Google's actions and the EFFECT of their actions and will say "Sorry, you don't get to do shit like that to try and weasel your way out of the law. It still applies." And those types of actions can often be used as evidence of an intent to break the law, so any lawyer would likely advise their client to not even consider doing slimy shit like that.

[SiempreViernes]

Ah, its a good point that the main target is the even more bad faith actor in the form of Facebook! Even more reason the technical criteria has to be broad.

I am however confused why you think the law being applied to the company having 91% of search in Canada is mission creep, you never actually say why claiming that Google has a "a significant bargaining power imbalance" is "loosely" interpreting the criteria....

And sure, the state might have a good chance of winning the court case about Google and Meta trying to avoid the law by not technically delivering the snippets themselves, but it'll take years for sure to go trough the courts. And why would you accept those several years where the damage the law is trying to prevent continues because you wanted the technical definitions to be narrow as when you already had criteria on company size.

[supriyo-biswas]

Now that you say that, I wonder if this is what SXGs[1] were about.

[1] https://web.dev/articles/signed-exchanges

[jsnell]

Except that the actual text of the law does not match your claim. It's about "facilitating access to news content by any means" including an "index" or "aggregation". Not about reproducing large chunks of the original content.

And it should be pretty obvious that this is the case. If your interpretation was correct, Meta would not have removed Canadian news sources entirely. They'd just have removed the previews (hell, the news companies could have removed the previews themselves; they already have the controls for that).

[llm_nerd]

>If your interpretation was correct, Meta would not have removed Canadian news sources entirely.

Meta's actions prove absolutely nothing. I could similarly say that Google didn't remove any news, therefore the law has zero effect?

Meta has, for months in advance of anything coming into force, embargoed Canadian news and replaced it with an appeal to the reader. That is political action, and it seems to have failed given that the government didn't blink. Now that Google has an agreement, it looks especially silly.

If Meta continues their embargo of Canadian news, personally I consider that a good thing. The Canadian public accepting that Meta is not a good source or "homepage" for news is a wonderful outcome of all of this.

[ipaddr]

I welcome it for the opposite reason. The few powerful families that control the news cartel in Canada need less reach.

[SiempreViernes]

You "welcome" media control going from a "few families" to a single person?

[ipaddr]

When you mute corporate media all of these little voices get louder.

[wesapien]

If I had to choose a lesser evil it would be Facebook

[brushfoot]

The act targets more than content previews. Notice (b):

> (2) For the purposes of this Act, news content is made available if

> (a) the news content, or any portion of it, is reproduced; or

> (b) access to the news content, or any portion of it, is facilitated by any means, including an index, aggregation or ranking of news content.

[Dylan16807]

I'm pretty sure I've seen multiple critiques that this specific law is in fact a link tax, and doesn't care if there are previews or not.

Is that wrong?

Sure, google currently has previews, and a law could be written to target previews, but I don't think this law targets previews.

[SiempreViernes]

It does target previews very explicitly, but it also has a much broader condition that I imagine is there to prevent Google and others from making previews something technically not provided by them but instead created by some payload they give the users of their website.

[CivBase]

Are news articles considered intellectual property, like books? If so, wouldn't it already be considered piracy for Google to copy and redistribute significant portions of the articles? If so - and if this law is not actually a "link tax" like you claim - then what is this law really preventing?

[nradov]

How many characters of preview would be reasonable? 50? 100?

[soganess]

Why would you even use number of characters?

A better criteria is: what fraction of people does it stop from visiting the source (because they got the information they were looking for from the summary). Yes, yes that system has it's own problems, but it is more holistic that string.length().

[nradov]

Determining that fraction would require the search engine to run a series of A/B tests with varying preview lengths, and keep detailed records of test results. That seems like an unreasonable burden.

[kridsdale1]

That’s exactly the kind of activity the google is doing every single day, thousands of them in parallel.

[CivBase]

How much am I allowed to quote from a book before I'm considered a pirate?

[nradov]

In the US we have fairly clear guidelines for fair use of copyrighted materials such as books. Rules in Canada may be different.

https://www.copyright.gov/help/faq/faq-fairuse.html

[cmrdporcupine]

It is amazing the web-illiterate/ignorant way this is being reported in Canada, too. Lots of mention of Google "posting" or "publishing" content, when all it's doing is linking to the publisher's own site.

I'm no fan of Google these days, but I wish they had played hard ball on this. This is a total joke.

[drawkward]

>when all it's doing is linking to the publisher's own site

That's not true, however. It is also often including a summary and an image, which is certainly republishing or reposting the content. In no way is it "web-illiterate/ignorant" to refer to it as such, because it is more than just a link.

[cmrdporcupine]

Google News is ridiculously spartan. It's very old school Google. A headline, a classification, and sometimes a thumbnail. No description or summary or text of any kind

Facebook, yeah, that's something else. One thing that is legitimately bad about that is the way the content can be posted on whatever wall and then a pile of unmoderated public comments put on it, that the publisher has no control over.

[kridsdale1]

Heaven forbid.

[rossjudson]

I highly encourage you to read Google's statement on how it supports and handles news. The word "often" is load-bearing in your comment, and this provides the details.

https://blog.google/supportingnews/#overview

[bryanlarsen]

Google/Meta do more than link, though. They also grab a hero image and an excerpt. Likely fair use, but definitely getting into the grey area. I don't know how the Canadian law is worded, but it would be much more acceptable to me if Google/Meta could get out of paying if they only linked and didn't excerpt images or text.

[randomdata]

Open Graph, which all the major news organization in Canada use, was created by Meta/Facebook to ensure that the news organizations have editorial control over what is "grabbed". If the news organizations want them to have less information, all they have to do is give less information.

[bryanlarsen]

Canadian journalists could have done this without the government by adapting tactics similar to unions.

A union strikes not because they want to withhold work from their company, but because even though both the company and the union members want to work, the strike shifts the power balance.

Similarly, Canadian journalists could have gotten together to withhold links from Google & Meta until they got a better deal. They'd both be worse off during the duration of the strike, but negotiations could get a better long term deal. And in this metaphor, Meta is doing a pre-emptive lockout.

The biggest difference is that the government is helping journalists combat the defector problem.

[AlecSchueler]

It comes down to culture and what kind of onus society expects to put onto worker to create and police de facto legislation, which is prefered in some places to allowing the state to enact the same de jure.

They could have done this but now they don't have to.

[babypuncher]

I feel like that could backfire since "scabbing" would be really easy, especially for less scrupulous news organizations more strongly tied to corporate or political interests than actual news reporting.

Basically while Reuters or AP are "on strike", Drudge Report, Jacobin, or Fox News would step in and become the de facto news sources across all of social media.

[bryanlarsen]

This is what I was getting at in the last sentence of my comment. The reason it's a government action rather than a collective action is that way it prevents defections, aka scabbing.

[ipaddr]

But smaller news outlets are not part of this cartel.

[wharvle]

Devil’s advocate: “You can choose what to give us and we’ll display it… but if what you give us isn’t good enough, who knows what might happen to your links’ prominence on our site?” isn’t quite some fully-voluntary thing when Google’s the one saying that.

[Dylan16807]

"You can choose between prominence and opaqueness" seems like an okay deal in a vacuum.

I suppose the issue is more that there's too much competition for news, rather than google making the competition unfair.

[s1artibartfast]

It is about the link itself, not the image. Meta refused to negotiate, because it doesnt control what links people post.

[KennyBlanken]

Bullshit. Facebook as a matter of policy manipulates people's feeds - the chances of something appearing, in what order, etc.

People have found that certain sites, topics, and words will trigger an outright shadowban on that post - your friends never see it.

[s1artibartfast]

not sure which part you are calling bullshit. Facebook certainly controls the how much a post is circulated, but it doesnt control which links people put in their posts, or how often they post.

The idea that facebook should have to pay a fee each time a user decides to share a link to news is stupid.

[echelon]

None of this is going to matter. In ~5 years everyone will be using AI summarizes and sharing this stuff with no chain of origination.

With the rise of rage and clickbait, that might be a really good thing to disincentivize many of the bad ways that media is produced and presented.

[AlecSchueler]

The acceptable way you're hoping for is exactly how it's worded. The headline here is fearmongering.

[whythre]

Yes. Adding more walls to the increasingly walled gardens of the internet is terrible.

[jsnell]

It's not at all clear what these payments are predicated on, the deal structure is obviously very different from what the bill originally required. All that's being annouced is that Google is paying $70M/year for something, but it might not be for links.

That's what happened in Australia's law from a couple of years ago. The law was ratified, but doesn't actually apply to any company, so what the law said was totally irrelevant.

[randomdata]

Especially when the news organization is free to deny the request referred by a linking party who has not offered the expected compensation. It is not clear why a new law is needed here when existing contract law already sufficiently covers the issue.

[wubrr]

How would news organizations deny the request referred by a linking party? Are you talking about a technical denial - as in reject the HTTP request? I guess technically they can base something like this on HTTP referer header, but you can have links without referer info as well.

[cmrdporcupine]

Referrer is exactly how you would do it.

The idiocy here is that most of these publishers are likely customers of Google's ad networks anyways, and the clicks through to the articles are yielding ad revenue to them, and they likely are getting analytics and tracking that identifies exactly where the inbound traffic is coming from.

It feels like a shakedown by people who are in other parts of the business totally disconnected from the web content / publishing arms, who likely know better?

[wubrr]

Referrer is basically optional though, you can specify a link to have no referrer : `<a href="example.com" rel="noreferrer">link</a>` , among other ways.

It's true that the links only increase the revenue/traffic to their website though, so they should really be supporting the referrers rather than blocking them.

[randomdata]

> you can specify a link to have no referrer : `<a href="example.com" rel="noreferrer">link</a>` , among other ways.

But you wouldn't do that if you want the link to resolve. If you don't want it to work, why would you go to all of the trouble of creating the link in the first place?

[Dylan16807]

And block people that use a bookmark or share the link via chat programs? I think almost no sites would do that.

[wubrr]

Huh? The link would still resolve and work, just the referrer header wouldn't be set.

[acdha]

What happens when they set a referrer policy no-referrer? I’m sure Google and Meta would say that’s to protect user privacy.

https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Re...

[randomdata]

The simplest option is to treat it as a rejection. Users aren't going to be dreaming up (choosing a recent article at random) URIs like /politics/federal/chiefs-of-ontario-says-trudeau-s-carbon-price-is-discriminatory-and-demand-a-review/article_9c995f63-1e26-56c6-9048-79781a9b649c.html in order to get there without some referring party.

Negotiations between Google/Meta and the news organization when defining the service contract can explore alternative options (access token, for example) should it be an imperative, for some reason, that the referrer still be inaccessible. People are allowed to talk to each other.

[cesarb]

> Users aren't going to be dreaming up (choosing a recent article at random) URIs like <long url> in order to get there without some referring party.

What about bookmarks? I don't think clicking on a bookmark on your browser is going to send any Referer header.

[randomdata]

If that case is beyond "who cares?", which I suspect is not, once the user is granted access to the news the first time then they can be given subsequent access regardless of referrer.

[acdha]

Users are going to be opening links from emails, SMS/MMS/RCS, other apps, and sites which do care about their users' privacy so, yes, this would be a problem for many people.

[randomdata]

It can only be a problem for the news agencies – and only if those users are a significant number to impact profitability. If those users don't matter, restricting their access doesn't matter. Who cares if someone who isn't helping your bottom line can't access the content you expect payment for? That’s the whole point of this – to keep out those who are not helping the news business.

Now, each successful entrance into an article via allowable referrer would come with an access token to allow future access absent of referral. When sharing in the small scale, the news agency can simply allow these links to be shared. But if the reach grows wide, suggesting that a major tech source has picked up the link and should be using their contractually settled upon authentication mechanism instead, then the token can be invalidated.

[Dylan16807]

These are the people trying to visit your website because of very specific interest. You should want their views a lot, not try to keep them out!

The access token plan is workable, though still causes annoying linking problems.

[Y_Y]

And if somebody sends you the link in an IM? Or if the browser fakes a referral from the news sites landing page?

[randomdata]

Any IM service not operated by a major tech company that the news organization wouldn't also want to collect compensation from is going see such a small number of referrals, who cares? Same goes for the number of people who are going to take the time to hack around it. Who cares?

[drawkward]

>Any site of any size should be free to link wherever it wants, freely.

Maybe that was true in the days before big tech, venture capital and advertising monetized the internet, but the internet has been vastly changed by those three groups, and not always for the better. They certainly do make a bunch of money though!

[ouEight12]

So.. we should break a fundamental aspect of an 'open and free' internet by introducing a link tax, just because the horse and buggy people haven't taken the last 25 years to figure out how to sell automobiles by now?

[drawkward]

The free and open internet died long before this link tax.

[acdha]

The law explicitly requires a power differential and $1B in revenue, so almost any other site won’t be affected.

I do wish there was more distinction between a link and content display. I think there’s a very real concern that summaries lower click-through rates, which Google has been pushing for years and will be turbo-charged with LLMs summarizing content in the future. It would be interesting to see if there might be some future nuance around that.

[oh_sigh]

Is it based on hyperlinks? I ctrl-f'd on the c-18 bill and there was no reference to "link". I would imagine the law is more about using the blurbs from the news outlet, not the link itself.

[jsnell]

The relevant part of the text is:

> (b) access to the news content, or any portion of it, is facilitated by any means, including an index, aggregation or ranking of news content.

[s1artibartfast]

Nope, it is the link itself.