Update, January 18: EU ministers have failed to approve the compromise text—with Germany, Belgium, Poland, Sweden, Luxembourg, the Netherlands, Finland and Slovenia, Italy, Croatia,  and Portugal all voting against the current Article 13/11 proposal.

Keep up the pressure! If you’re in the Czech Republic, Luxembourg, Germany, Poland, Sweden, or Belgium —tell your government to oppose Article 13 and 11.

Politicians are meant to broker compromises in the pursuit of the public good – though in a year that is already overloaded with government shutdowns and Brexit logjams, that skill seems in short supply.

But sometimes there are no compromises to be found. Sometimes, even the most talented diplomats are handed an impossible task. The Romanian Presidency is struggling to finish negotiations the Copyright in the Digital Single Market Directive together. But two parts of that law —Article 13, intended to introduce compulsory copyright filters, and Article 11, a new licensing requirement on reproducing snippets of news articles—are so controversial that they risk sinking the entire process.

Just hours before a key vote on this Friday, the Presidency has presented their proposed compromise to the negotiators. The text, leaked to Politico Europe, shows just how far they will have to go to bring all the parties together.

On Article 13, the Council and the Parliament are struggling over whether small and medium-sized businesses should be excluded from the crushing demands and liability Article 13 would impose on Internet sites. This was one of the concessions that MEP Axel Voss offered in a last-minute attempt to get the Article’s provisions past Parliament.

But that’s not good enough for the article’s lobbyists, who believe that any site that allows users to put their content online should be treated as a pirate’s den—even if it’s a small European Internet site hoping to compete with deep-pocketed, US-based Big Tech companies.

The trouble is that the whole pitch to Europeans for accepting Article 13’s excesses was that it was aimed at clawing back money from YouTube and other big, foreign hosting sites. MEPs and other elected officials aren’t likely to be so keen on a provision that will also bleed money from the fledgling EU digital sector into the coffers of the established rightsholders.

Or, indeed, bleed money from individual European Internet users. Since the draft Directive passed Parliament, another huge battle has emerged over whether Internet users should be covered by the licenses that the Big Tech companies will need to negotiate with Big Content. The rightsholders want the ability to be able to be able to “double dip”—suing to extract cash from the YouTubes and Twitch TV’s of the world, and then again, suing individual Net users to get some more money from prominent YouTubers and Twitchers.

(Well, at least, that’s what the recording industry wants. At this point, it’s only the music industry that has any hunger left for Article 13 – all the other major European rightsholders have backed away from its dangerously vague language, and have now turned against it. Given the choice between the current status quo, and a Directive oscillating so wildly between wild demands and unclear carve-outs, they’d rather stick with the current, functioning, Internet. Even the recording industry has denounced the latest proposal, wanting to rewind to earlier versions that lack even the veneer of compromise.)

A similar pattern has emerged in Article 11. Lobbyists for Article 11 said it was to intended to stop Big Tech wholesale stealing news articles and money from journalists, and had nothing to do with mere linking to news stories. The negotiators have taken them at their word, and now have suggested that quoting “insubstantial parts” of news articles should be acceptable, and that also some of the money made from the new licenses should go direct to the authors of the articles. Not so fast, say the lobbyists: we still want to limit linking to only allow “individual words”, plus who said anything about giving money directly to journalists?

The tragedy in all of this horse-trading is that nobody in the room (or shouting into the keyhole) is actually fighting for Internet users, or the public good. They’re just trying to act as referees between various industrial sectors.

Meanwhile digital rights groups, and everyone from actual journalists to actual Internet experts to actual copyright experts have said that there’s no compromise to be made here, because the whole system of copyright filters and special news licenses simply won’t work to remunerate creators, and will instead just break everything online.

None of the proposed compromises fix the underlying problem. Even if you exclude small companies, you’ll just be creating a two-tier Internet with the tech giants controlling the licensing, and European Internet startups struggling to stay small enough that they won’t be sued into oblivion. Even if you grant individual Twitchers and YouTubers some protection from being sued, they will still suffer, because of the unaccountable black-box algorithmic filters will constantly block them from broadcasting or recording their legitimate works. And no matter how you quibble about who gets the right to link or quote news stories, the absolute lifeblood of gathering an audience for an independent news site – through people linking to it, and quoting it – will be chilled by companies refusing to take the risk of a lawsuit, and individuals and non-profits being unsure as to how they can keep safe from the 28 national versions of an ambiguously-written law.

Meanwhile, in the fantasy world where all of these consequences might be avoided with a few carefully-worded recitals, members state negotiators now have less than 24 hours to work out that perfect wording.

The Romanian Presidency is working under a major, self-imposed deadline—if agreement cannot be struck before mid-February, it will be too late to present the text for a final European Parliamentary vote. But a directive that risks sabotaging the entire Internet should not be rushed like this.

Compare this pell-mell hurtling into chaos to the leisurely pace of another EU Internet law: the E-Privacy Regulation. That proposal, which takes its aim at the privacy issues surrounding Big Tech and Big Media, has been stuck in pre-Trilogue negotiations for two years. It’s a sad indication of the priorities of the European establishment that improvising new, untested, copyright law is more urgent than tackling the clear challenges of digital privacy.

But if the clock really is ticking for the Copyright Directive, the correct compromise on Article 13 and Article 11 is to remove them entirely from the Copyright Directive, and send them back to the drawing board. The European Union’s negotiators are struggling to strike a deal with internally incoherent and unenforceable language. This trilogue has deleted previous articles in its pursuit of a reasonable deal; it can delete these two as well. Anything else is sacrificing the stability of the Internet for laws that nearly everybody opposes, and no-one needs.

Countdown to Catastrophe: Article 13 and 11’s Key Dates