Vmware launches Liota (Little IoT Agent), a vender-neutral open-source SDK | Hacker News Reader

Vmware launches Liota (Little IoT Agent), a vender-neutral open-source SDK

github.com

Does anyone actually believe that IoT is anything more than marketing fluff? What are the benefits of my LED lightbulbs having an IPv6 address?

Yes there are definite benefits. The problem at the moment is that nothing is interoperable. Everything is proprietary.

I'm hopeful that eventually we'll get some semi-open standard that most things support. Probably Homekit or Weave. They have the huge advantage of native support on a mobile OS. Weave looks better, but Homekit has the advantage of actually existing.

I would say, wait 3 years then ask that question again.

And some possible benefits of connected lights are:

* Automatically turn off when you leave the house (or room if you have good enough sensors).

* Automatically adjust colour temperature over time (e.g. using the weather).

* Gently fade on when your alarm goes off in the morning.

I lived in a flat with motion-sensing lights in the bathroom and I wish they had been connected so I could have made them dimmer at night.

Wouldn't it be better if they were just automatically dimmer at night? Wouldn't you still prefer a real dimmer switch, even if it moves automatically on a schedule? I can't see wanting to pull out a phone (which I may not even have with me at night) to dim the lights while my hands are wet. I don't even have a smartphone, and my flatmate broke hers irreparably the other day and has been without while waiting for a replacement.

I really don't see any reason why people should be expected to control devices which are right in front of them with something in their pocket.

There's a reason we still carry cash: when the lights go out, and the ink on the credit card recorder runs dry, you can still pay for things. People who do not carry cash deserve a nomination for the Darwin Award.

I agree that only being able to control stuff using your smartphone is dumb, but that's a separate question from whether IoT is useful or not, as you could have both.

As far as I know, those things run on a schedule most of the time but maybe you want to leave early and turn off all the lights at once. Or if you want to pull an all nighter and want normal light so you don't get sleepy quicker.

There are a lot of cases when you want to change your schedule and your phone helps you do that.

Furthermore, you will be controlling the lights less with an app and more with your voice. For instance, if you tell Siri good night (and you have home kit enabled), Siri should dim or turn of the lights. Combined with Amazon's Echo, this can really be handy (although I have a lot of privacy concerns regarding Echo).

Of course it's better if things just do what you want them to automatically, and that's entirely possible with current connected bulbs. I've got some with physical buttons to control the lights, and some smarts (so that I can hit the same button for full brightness at some times, a nice evening dimmed at others, and really low level at night).

The issue at the moment is that it's not really accessible to everyday users, I had to interact directly with the system's slightly obtuse API to do that. Your everyday user just gets switches to activate a particular scene.

The tech and especially interoperability isn't there now. It has promise though. I'd like the coffee maker to heat up when I wake up. I don't wake up the same time every day, so my fitness tracker could tell it when I wake up. I'd like to turn off all the lights, TVs etc that the family leaves on at night without walking downstairs. The Nest could actually work better than a programmable thermostat if it didn't rely on something wandering past its motion sensor which my family does not do on some days we're home and certainly don't do when we take a day trip.

IoT will be huge in industry. Manufacturing, supply chain, logistics, etc. will be transformed dramatically.

I would love to have my AC unit controllable by my computer. Then I could automatically scale the temperature settings between comfortable and good enough for the cat. The real payoff would be the bedroom since our sleeping hours aren't consistent.

I think it's a form of forced obolescence. Pre-IoT, pre-"smart" devices could last for decades with only a little maintenance, but these new devices with software that is inevitably buggy, needs updates, may require continuous connections to a third-party service, and may become completely unsupported within a short time basically force you to buy a new one every so often. LEDs theoretically could outlast incandescent lamps, but the ones in these new lamps are driven at such high power that they thermally degrade and fail quickly.

The security/privacy issues are also very concerning, because these systems are so complex that they become incredibly hard to understand, troubleshoot, and maintain. I can understand and repair essentially all the appliances I have because they are simple, non-IoT, minimal-software systems with no interaction between them. The tradeoff is between arguably less convenience when they work but low MTTR when things go wrong, and convenience when they work but very high MTTR or replacement cost when they don't.

Related announcement post: https://news.ycombinator.com/item?id=11833620

That example code isn't going to run. Hyphens are illegal in python identifiers.

Would be better if they called it Lolita :)

Or just "Iota", which already carries the connotation of "little" without the preceding "L".

No. Because it is intended for Goodfellas.

Say all you want, but it's an example of how not to name things, especially an open source project meant for wide adoption. In written form, it's basically a variation of "Lolita", which leaves a corresponding impression even after it's re-read properly.

That ... didn't even enter my mind, FWIW.

Whereas my first thought was "Ray Liota", which apparently has two T's. Also with the fact that I would pronounce the two words quite differently, I think you're wrong.

Lolita isn't a very common term outside of *chan.

It's one of Nabokov's more notable novels, but I guess to each his own, including *chan for some.

That's because you spend too much time combining literature discussion and back water IRC channels! More seriously I agree, terrible naming, especially given another poster pointing out Iota would have been fine.

Reply on news.ycombinator.com