Hackaday Links: September 6, 2020

Hackaday Links Column Banner

That was a close shave! On Tuesday, asteroid 2011 ES4 passed really close to the earth. JPL’s close approach data pegs its nominal distance from earth at about 0.00081083276352288 au! Yeah, we had to look it up too: that’s around 75,000 miles (120,000 kilometers), just ten times the diameter of the earth and only about one-third the distance from the earth the moon. It got within about 52,000 miles of the moon itself. Bookworms who made it all the way through Seveneves are surely sweating right now.

There’s a low current arms race when it comes to lighting up LEDs. The latest salvo in the field comes from [Christoph Tack] who boasts a current of 1.36 µA at 3 V for a green LED that is roughly 10x brighter than a phosphorescent watch dial. Of course, the TritiLED is the design being chased, which claims to run 17.6-20.2 years on a single CR2032 coin cell.

Proving once again that Hanna and Barbera were indeed future-tech prophets, flying cars are now a thing. Sky Drive Inc. made a four-minute test flight of a single passenger octo-rotor aircraft. Like a motorcycle of the sky (and those are a thing too) this thing is single-passenger and the cockpit is open air. The CNN article mentions that “The company hopes to make the flying car a part of normal life and not just a commodity”. Yeah, we’re sure they do, but in an age when electric cars are demonized for ranges in the low hundreds of miles, this is about as practical for widespread use as self-balancing electric unicycles.

Just when you thought the Marble Machine X project couldn’t get any bigger, we find out they have a few hundred volunteers working to update and track CAD models for all parts on the machine. Want a quick-start on project management and BOM control? These are never seen as the sexy parts of hardware efforts, but for big projects, you ignore them at your own peril.

Google and Apple built a COVID-19 contact tracing framework into their mobile platforms but stopped short of building the apps to actually do the work, anticipating that governments would want to control how the apps worked. So was the case with the European tracing app as Elliot Williams recently covered in this excellent overview. However, the United States has been slower to the game. Looks like the tech giants have become tired of waiting and have now made it possible for the framework itself to work as a contact tracing mechanism. To enable it, local governments need to upload a configuration file that sets parameters and URLs that redirect to informational pages from local health departments, and users must opt-in on their phone. All other tracing apps will continue to function, this is meant to add an option for places that have not yet adopted/developed their own app.

And finally, it’s time to take back responsibility for your poor spelling. Auto-correct has been giving us sardines instead of teaching how to fish for them ourselves. That ends now. The Autocorrect Remover is an extension for Google Docs that still tells you the word is wrong, but hides the correct spelling, gamifying it by having you guess the right spelling and rewarding you with points when you get it right.

15 thoughts on “Hackaday Links: September 6, 2020

    1. We need AI that can detect the difference between poor virtual keyboard use, and genuinely not knowing the correct spelling. I think it would be nice to have help realizing you can never remember if it’s gauge or guage and giving you the business about it.

      1. Good point. I would not be surprised if a lot of younger people who grew up with virtual keyboards and auto correct have been crippled in their typing skills as well as their spelling.

        Personally I disable a lot of auto correct on my work laptop, and prefer to correct my own mistakes. My theory is that if I let auto correct fix it for me, I’ll probably make the same mistake again. But if I force myself to retype words correctly, then maybe muscle memory will help me out next time.

        A more useful way I use auto correct is like a macro for inserting symbols. Like *C to [deg]C (sorry, apparently can’t even type the degree symbol on my virtual keyboard), ohm to capital omega, uF to [mu]F, things like that.

        1. once i’ve got a phone bill that i know nothing about, so i talked with that phone company, they told me to go to the police and state that i’m not using any such a provider, so i went, and to my biggest surprise one police lady brought me a piece of paper and a cheap pen and asked me to write down my case :) now that was scary, not the spelling but the cursive writing what i’ve done only in school and it was and still is ugly as hell :) so i was there for a long time writing with all capital letters :)

  1. If course, the first thing when you hear a about a near asteroid was how fast was it moving and how big was it? ie what would we do if had hit the earth?

    The one above was doing 8km/s ie 29,000kmh. It was 12.7km by 5.8km. Assuming it was made out of something lightish (say 2g/cm3) then it if it hit the earth head on, it would have wiped out a fair bit of the life on earth I suspect (think the energy of about 80,000 big atomic bombs all at once).

  2. Just imagine the noise made by hundreds of those flying cars flying in a city. Just unthinkable, so not gonna happen. Anything flying with blades or turbines is too noisy for a city environment. May be a few ones will exists, like helicopters today.

  3. In the linked NASA spreadsheet, they now say 0.00988652085573321 au for Sept. 02 07:49. It seems this has been corrected since the previous calculations made in January, since 0.00081083276352288 au value can be found in the Google cache. It seems there was quite an uncertainty in the previous calculations.
    But that makes a huge difference, since it is rather 1.479.000 km or 919.000 miles. It was close, compared to Neowise’s 103.000.000 km, but not so close.
    If the uncertainty made the previous calculcations wrong in the other direction, it could have been 0.00008 au, and that would have been a very close shave!

  4. One simple question. If user data isnt personally identifying, and modern devices use a random Mac address when scanning, how is contact tracing supposed to function?

    Probably because our data is personally identifying no matter how much they say otherwise and a random Macs won’t confuse a framework that also fingerprints the probe requests you send out.

    Remember, every Google device beams back to the overlords every single device it hears along with the devices last known at GPS location.

    All hail the Google Overlords.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.