Chromecast only works from iOS for apps that have integrated with the Chromecast SDK. Like Netflix and Youtube.
Maybe the EU can ask Google to open Chromecast up so Apple can integrate it into iOS. Then you can use it from any app without needing special integrations.
Apple has all the information they need, available publicly, to implement an OS-wide Chromecast client. They haven't done so because they don't want to.
> I believe there are people who want results sooner rather than later
In Denmark all ballots are hand-counted. It takes about 6 hours from polls close to every precinct reporting a preliminary result. Wanting it faster isn't really necessary, other than to feed the 24/7 news machine.
Italy has 60M people and paper ballots, results are out the next day. Population does not matter since polling stations can be scaled up proportionally.
I've been in voting where we had a dozen ballots per person (referendums) so this would be more than the total paper ballots in the US, it works fine.
Minor miscounts happen but nobody has ever seriously questioned the overall vote results.
Uh no. This election I had a president, two senators, a state senator, a state assembly, a county executive, city council, school board, prosecutor recall, and half a dozen ballot measures. It took four legal size paper surfaces.
Some of the regional elections in Germany have comically large ballots with dozens of options and a very complicated counting system (16 votes that can be split between individuals or party lists). The hand-counted results are generally available by the next morning. There is really no excuse for using electronic voting. In Germany it has been ruled unconstitutional since it cannot be checked by the voters.
Those could easily be prioritized. First count the federal elections, then the state elections, then the county elections. You get the presidential and congressional results within a day, the state election results within two-three days, and the more local ones within the week. Is anyone going to seriously complain that it took a week to find out who is on the school board?
So does Denmark, why do you assume it's any different? Counting doesn't take that long when you split it across 10K voting districts, which is what the U.S. did for most of its existence. We want ACCURATE elections free from corruption and hacks. You can't undo a fucked up election.
In Denmark, I don't even lock up my helmet. My bike, sure. But the helmet just casually hangs on the handlebars. I've never experienced, nor heard of, anyone losing their helmet when doing this.
> My belief is that retribution is the immune system of civilization.
Have you engaged with any justice system that isn't based on retribution? Have you seen the results they achieve? Norway is the classic example. They give more lenient punishment for crimes, their prisons are vastly more comfortable, and actively try to educate their prison population. All things Americans (as a general point) would scoff at.
The result? Norway has a vastly lower recidivism rate. 20% vs 75+% in the US. Norway is a uniquely good example, but similar results play out across Northern Europe.
> Have you engaged with any justice system that isn't based on retribution?
Most people here who live in SF have done just that, not by their own choice - and the results are not pretty. "Restorative" justice might or might not work in places like Norway, but let's just say other implementations have been quite bad.
> On June 12, 2014, Tesla announced that it will not initiate patent lawsuits against anyone who, in good faith, wants to use its technology.
with the following definition:
> A party is "acting in good faith" for so long as such party and its related or affiliated companies have not:
> * asserted, helped others assert or had a financial stake in any assertion of (i) any patent or other intellectual property right against Tesla or (ii) any patent right against a third party for its use of technologies relating to electric vehicles or related equipment;
> * challenged, helped others challenge, or had a financial stake in any challenge to any Tesla patent; or
> * marketed or sold any knock-off product (e.g., a product created by imitating or copying the design or appearance of a Tesla product or which suggests an association with or endorsement by Tesla) or provided any material assistance to another party doing so.
In other words, Tesla is giving away all of their patents only to entities that grant Tesla access to all of _their_ patents. That is hardly open-source in the software development sense of the word.
Interesting, thanks. But I’d disagree that this makes it “hardly open source”. It makes it more ‘GPL’ than ‘MIT’, but both of those qualify as open source. It seems a pretty reasonable condition in this case, imo. In any case it would be a stretch to call it a “lie” that they open sourced their patents, which is the claim I challenged here.
> The article mentions bounds checks, but gives no measurements. Have they ever measured?
Related, the prominent Rust community member "Shnatsel" wrote an article [1] about exactly that in January. It's a great look at the problem, and gives some actual data to evaluate with.
Thanks for the reference. Several of the examples given could benefit from loop versioning and/or iteration space partitioning, i.e. the compiler splitting a loop that it cannot prove is always in-bounds to an in-bounds part and a possibly-out-of-bounds part. The HotSpot C2 compiler does this under some circumstances.
E.g. the simplest case:
for (i = 0; i < some_expr; i++) {
a[i] = other_expr; // cannot prove is in bounds
}
Becomes:
var t = some_expr;
if (t <= a.length) {
for (i = 0; i < t; i++) {
a[i] = other_expr; // no bounds check necessary
}
} else {
for (i = 0; i < a.length; i++) {
a[i] = other_expr; // no bounds check necessary
}
throw OutOfBoundsError; // throw exception at first OOB iteration
}
This approach will have exactly the same behavior as the original program and run without bounds checks in either case. It will even throw at the right time. The cost is having two copies of the loop and a check up-front. This specific case could be done with one version, and other cases can be done with a different partitioning of the loops.
I feel people are lacking imagination on what an awesome buffet of compiler transforms are available when it is both absolutely required to do bounds checks and absolutely necessary to try to eliminate them everywhere. In practice, many many loops are trivially in bounds, many other loops are cold enough or big enough that the cost of a bounds check isn't much. You don't need to unroll and version every loop in the program. And like I said, for the loops outside of this set, where iterations are not affine or the indexes are irregular, you either live with bounds checks or you go deeper on allowing static assertions to prove accesses are in bounds. That's exactly the cases where you probably want checks anyway.
> As such most ebikes give the same fitness benefit but let you go faster
This is simply not true. A pedal-assist bike will go faster with the same amount of W put into the pedals, yes. But will people put in the same amount of W if 60 % will get you to your "target speed"? I doubt it. And then you get less health benefits for the same distance traveled.
On my 8 kilometer commute I average 150 W. Not because I use it as exercise. That's just where I find my comfortable level of output. Every time I've ridden on ebikes I've put in much, much less effort. I'd be surprised if I put in even a third of the energy. That's great if you just need a mode of transport. Bikes are practical, efficient, and planning for them improves cities. Even ignoring the potential health benefits. But claiming that a pedalassist bike gives the same fitness benefits just doesn't pass the smell test.
I can only state for myself that I'm putting more effort in (since I can feel the bike slow down more when I don't). Plus the ebike allows trips that because of distance I wouldn't use the regular bike for.
> At the end of the day the counter to this is starting and building companies with remote work forces.
You'd think so. CrowdStrike, a self-proclaimed remote-first company, instituted a return-to-office policy earlier this year. Despite never requiring office presence before the pandemic. They still have blog posts like [1] on their website.
reply