I mean, this is good news, but why was the efficiency of the system so misunderstood at the design phase? I hope someone's interested enough to find out!

Ahoi! Participated to the development of a simulation tool for that purpose. Many factors need to be taken into account:

- The gradients of the track

- How agressively you want to drive at different time of the day

- How many trains

- The timing of the trains. When one train brakes, it can power another one that accelerates at the same moment

- The efficiency of the power chain

- The resistance model of the train

With so many parameters the results can be quite volatile. However, rule of the thumb in the business is that 30% saving can be achieved with good energy management. Hope this clarifies a little.

Update: bullet points

Interesting! Thanks for the great info. Do you know if anyone will have the opportunity to integrate observations into the model so it can continue to improve?

We did use actual data from existing lines to measure the gap between our models and reality. We were +/- 5% accurate.

Maybe they didn't factor in ridership on certain part of the track gradient, which would lead to more regen braking than expected?

...but why wouldn't they use AI as an oracle? From an outsider's perspective, it seems that there's already plenty of incentive to test the margins of acceptable academic practice in order to produce more papers or publish more quickly. Sadly I feel like it'll become the norm to have a chatbot interpret your results and write your paper rather than using those expensive grad students.

I don't have answers; just the lingering question "why are we building this?"

We're building this because the ability to make narrow, specific predictions can be narrowly and specifically useful. This works if you have a good understanding of both the tools and the domain you're looking to make predictions in.

Unfortunately, from an outsider perspective, this looks like being widely and generically useful. If you don't understand your tools, you're going to misuse them, and this hype cycle is the result.

The first recorded use of the word "busybody" was in 1526, per Merriam Webster. Some people just feel like it's their duty to tell the world how to be. I think it's just human nature.

As a former iOS engineer at Apple, and a dvorak user, I can verify that this is exactly the thinking (during my time) of why dvorak support wasn't a development priority. The properties which make it good for typing make it bad for a phone keyboard.

This seems like sound logic as long as iOS is only used for phones and an on-screen keyboard; but isn't it used for the iPad as well? And wouldn't a lack of Dvorak support mean you can't use it with your standard external keyboard?

I don't know, I'm not an iOS user but it would hugely inconvenience me if the only way I could use Dvorak on Android is to use a keyboard with a hardware Dvorak layout.

Edit: Indeed, it appears that iOS supports Dvorak layout for external keyboards already; this is supporting the native on-screen keyboard

Yes, your assumption here is off - iOS supports remapping standard external keyborads to Dvorak. I’m typing this on Dvorak on my iPad with a standard Logitech keyboard connected via the Smart Connector.

That's good to know, thanks!

It made sense to me. Luckily, Dvorak support for hardware keyboards has been always available on iOS.

I'm surprised to hear this. It seems like the point is clearly to allow existing Dvorak (desktop) users to avoid switching formats when they switch devices. It shouldn't matter whether Dvorak is optimal for typing on a phone.

This really doesn't matter because the "muscle memory" doesn't carry between the two formats.

Disagree. I've been almost an exclusive Dvorak user for 15+ years, and the muscle memory absolutely carries over.

The muscle memory of touch typing, by definition of muscle memory, does not carry over to thumb typing.

https://en.wikipedia.org/wiki/Muscle_memory

Good point. Though a significant part of typing speed is your _spacial_ memory (where is the J key on the keyboard?).

Your _muscle_ memory makes you good at moving your fingers (or thumbs) to an absolute position on your keyboard (or screen) - but to do that you must first decide what absolute position your finger (or thumb) should move to (where is the J key?).

And yet, somehow, I feel like it was a lot faster for me to learn how to type on an iPhone--nearly instant, really--because it was using QWERTY instead of having the letters in a random order, so it isn't clear to me that what you are saying matters even if it were true.

Yes, and as others point out iOS has long supported Dvorak (and Colemak and a few others) where that desktop muscle memory matters: when using a hardware keyboard via Bluetooth.

Qwerty is useful for "swipe typing" on a touch screen and Dvorak/Colemak is great for touch typing on hardware and the way "muscle memory" works those are such different media/muscle movements that they have separate "muscle memory".

True. I was happy when I switched to Android because Dvorak was available as an on screen keyboard. I didn't last for more than a day with it. Dvorak is terrible for on screen typing.

Was there any research into what keyboard would be the ideal phone layout (presumably something entirely new)?

Back in the Days of Yore we had something called "Fitaly", which was supposed to be optimized for pen typing: https://textware.com/fitaly/fitaly.htm

I loved Fitaly and the associated ecosystem. Palm, Handspring, the usable stylus on a 160x160 screen...

Few years ago someone posted a link on HN of a keyboard that was designed for optimum swiping, fewest overlaps and what not.

It faced the same problem as all new keyboard designs, getting people to use it.

If you are on Android, you can find plenty of research project keyboards that are indeed more efficient to use, after you get over the learning curve.

I've been working in climate tech for about 7 years now, at https://sense.com. Very fulfilling. We're hiring in lots of roles. Feel free to get in touch. (Email in profile.)

Should be marked [2016], it seems.

I was wondering if you could defeat the beans/lentils by injecting some water vapor in there (while maintaining the vacuum & placement of the items), and freezing the whole thing. Then cut it open under hard freeze and maybe everything sticks together.

Might even be doable without water vapor at cold enough temperatures. Vapor could damage or cause the rice/beans to start to rot later on, dead giveaway of tampering. Maybe the plastic beads are better in general because of their lack of moisture and general lack of organic weirdness.

plastic beads are also infinitely reusable, although probably have a much higher environmental impact than grains. Gravel could work.

I made one of these halos 2 years ago, and my anecdotal experience is that the older (fully grown) sparrows can't figure it out, but the smaller, younger ones get through. Then, weeks later, I see fully grown sparrows at the feeder, ignoring the halo. Did the little ones "fit" and then grew too used to it to be spooked? Did the big ones learn from the little ones? I don't know.

Another observation: Because of the number of sparrows, I did pull the whole feeder down and kept it down for months. Then, when winter approached, I put the feeder and the halo back up, and no house sparrows made it through again. Did the adults who were defeating it before forget? Move on? Die? Who knows!

Check out fwarnock's reply. They explain exactly this behavior.

No. It only works if you're so familiar with the previous UI than your pattern recognition kicks in. Someone who sees this for the first time will have no idea what it means. It's inaccessible to the novice.

Also, tabs without a favicon don't start with an icon at all.

> It's inaccessible to the novice.

It doesn't seem to be the novice people speaking up against it? Or are they all acting in defence of novices they don't know?

If you haven't yet, check out Scott Mabin's blog. He's working on ESP32 & Rust: https://mabez.dev/blog/posts/esp-rust-18-10-2021/