tonfa

I assume taste was meant in term of coding. "taste" is still often the lacking trait that LLMs have when it comes to code design.

If the SVP says it, it must be true :) What's their sources? I'm sure they could find a couple of anecdotes, doesn't make it significant.

More surprising it didn't pass the "majority of cantons" either (both are required for initiatives like this), I would have expected it to pass (there are a lot of smaller/rural/alpine cantons which tends to vote more conservative).

> The current system permitting freedom of movement across the continent while devolving immigration policy entirely to members creates a fundamental tension the EU needs to resolve. Because otherwise, Berlin can basically dictate EU immigration single handedly, which is bound to generate backlash even if they run a perfect programme.

You do realize German nationals (followed by French) are the top contingent in term of immigration to Swizerland.

(Only EU citizens benefit from freedom of movement to settle in Switzerland)

What's the issue with WSJ? "people familiar with the matter" is standard lingo, means the journalist and editors have vetted the sources (multiple).

Seems like it's 4 per hour on Rotterdam/Utrecht, seems similar to Geneva/Lausanne with 6 per hour.

In any case, I think commuters are fine with every 15 min, as long as there's enough seats. (for long distance like trains, my feeling is that frequency below 15min doesn't have a lot of impact, unlike shorter distance public transport like tram/bus/subway)

Yeah there's tons of work ongoing. Lots of line close to the big hubs have ongoing construction to eventually switch to 15min takt.

Improvements on various train station (new underground stations in Geneva and Luzern, extra platforms, etc.).

https://company.sbb.ch/en/railway-development/future-rail/na...

(for example, there's also lots of tram, etc. projects)

Well there's anyway going to be a referendum about the bilateral. (which is why I find the initiative somewhat stupid, you can vote on the real deal in a few years, about whether people want or do not want to have agreements with the EU, instead of hiding it behind a fake/emotional reason)

> The public version of Gemini is ridiculous. At least half their search "answers" are just wrong.

That's not Gemini, that's AI Mode (in Search), they're different products built by fairly different part of Google (actually one is built by Deepmind).

(I don't think it's much comparable to https://gemini.google.com/app at least in the past you'd get very different results)

Especially after we saw how happy the EU was to negotiate (they didn't budge) when https://en.wikipedia.org/wiki/2014_Swiss_immigration_initiat... passed.

The new initiative is basically the same, but with no leeway to ignore it.

(that said I suspect if it passes, there will be something tied to the bilateral referendum in 2027/28 to try to supersede it)

> This referendum is an attempt by the members of SVP/UDC, the right-most party, to show that on immigration topics they have more popular support than what their relative power

Not really about immigration but EU relationship. Almost every SVP initiative tries to create a contradiction in the constitution with foreign agreements to force an "exit".

> The strong point of the Swiss political system is that the government is, by law, made up by all significant parties.

It's a tradition, not a rule (the composition of the council is simply the result of an election by the parliament).

https://en.wikipedia.org/wiki/Magic_formula_(Swiss_politics)

Unlike UK, the impact to the EU is minimum and Switzerland doesn't have leverage (if the EU still stands).

Of course if you have EU dismantlers in power anyway in FR/DE, they'll just be happy to sabotage.

Germany (16% of recent immigration), followed by France and Italy (12% and 11%).

https://cms.news.admin.ch/fileservice/sdweb-docs-prod-nsbcch...

(page 5)

"the swiss equivalent"

As OP explains, freedom of movement can't be stopped in isolation from the rest of the bilaterals.

(btw funnily Schengen is just about the border control, we're talking about freedom of movement which is a different thing, e.g. UK wasn't in Schengen but the freedom of movement applied to UK as well before brexit, tho I guess people use Schengen interchangeably)

> This is only about the Schengen, Switzerland is not a part of the EU

Not really, the bilateral are a package and the EU doesn't want CH to pick and chose.

If freedom of movement stops, a whole lot of thing also stop. It happened the last time SVP got something similar voted on (introduction of quota for foreign immigration), on a smaller scale (erasmus and horizon which are the higher ed and academic research collaboration, CH was a heavy recipient of the latter).

If you're truly convinced there's nefarious reasons for including megacap IPOs in passive index, you can always short the stock (or use derivatives) by the same amount.

I'm not sure you'll come out ahead. (Personally I don't get the outcry, except for nasdaq which has fairly stupid rules, delaying the inclusion of megacaps won't make the problem go away, but probably increase since the float will be massively larger). It's inherent to being a passive index.

The float will get bigger as you wait tho, since it's common for early investors to be locked for e.g. 6 months. You can argue it's better to smooth the entry as float gets unlocked rather than being front run by all the hedge funds in a single day on a massive capitalization.

Also they needed to wait for the person to travel to the US.

> Also, they are only renting out colosus 1 ($10B), not colosus II ($18B).

The news from the S1 is that they're renting both (see OP).

Afaik this is because there is a community and database around them.

+1 I learned so much about software engineering (Software design, network protocol incl how to handle backward compat in sane ways, ...) thanks to the mercurial community.

> but at a slower rate that can be sustained by inference revenue. reply

also it's possible that the scale of inference needed (e.g. Jevons paradox) keeps growing to the point that training costs can fully be absorbed (since training cost is one off vs. inference that can scale).

(I suspect that might be the thinking, I don't know if it will be true, it's also possible that no model will create a moat big enough to attract enough of the inference traffic to make it true).

Depending on the chips/architecture used, the off-peak traffic from inference can also subsidize the training costs.

> Did anyone try to map this code back to the original codebase to see if this memory problem exists in the original codebase?

FWIW what is being discussed is not memory problems, it's breaking rust invariants (the unsafe code has to follow specific rules, e.g. annotate lifetimes properly).

> Windows is practically banned.

FWIW I don't think this is accurate (was kinda true in the 2010s?). I wouldn't be surprised if it's almost easier to get windows laptop than linux one now.

Since it's mostly browser tabs, as long as you have ample memory (eg 16gb) it's good enough.

And the counter intuitive thing is that people in countries with lots of renewables and not so many external links (e.g. Scandinavia with hydro) might be against adding more links since it will increase electricity prices.

Bloomberg has a long read about this: https://www.bloomberg.com/news/articles/2026-05-01/why-uk-gu...

(pottty training strategy changes were apparently somewhat pushed by researchers close to disposable diaper industry)

> Why did google code end? Was it too hard for them to monetize?

My guess is that abuse (people hosting files/data that google didn't/wasn't allowed to host) made it untenable for a service that wasn't generating revenue and had limited headcount.

Something like Google drive or yt could spend a lot more energy stomping it rather than the handful of folks from the open source programs team.

For me, can do anything complex in terms of history rewriting without ever shooting myself in the foot (worse case just `jj undo`).

And the UI is pretty intuitive so don't really have to search for a solution, usually the command to use is obvious (tho I guess now you could delegate the UI to an LLM to ask it to give you the right incantations).

Do you use a mega-merge + absorb workflow on top of the faned-out changes?