4 ms·

I don't think things have really changed. The games market is saturated, but the IoT is only just beginning, and the next "minecraft" (sorry I don't know a lot

I don't think things have really changed. The games market is saturated, but the IoT is only just beginning, and the next "minecraft" (sorry I don't know a lot of indie games) will probably be something AR or VR. AR is posing to really take off. I mean, I work for the government and I'm not very creative, so these upcoming things might seem lame to a lot of people, but we can basically cover a 15 km radius with slow "internet" (LoraWAN) for around $1000. In this you can build all sorts of small devices. You can monitor street temperatures and giving the data to the public via open data. You can monitor if elderly people are alive by tracking whether or not an IoT lightbuilb turns on. You can build tourism apps that emit small stories via Bluetooth. I mean, the options are endless, and it's basically shaping up to be the Wild West that indie games were 10 years ago.

sillysaurus3 9y ago
AR isn't really going to become a thing. Not with current technology, anyway. The problem is twofold: There's no way to get enough power to the device to do anything interesting, and there's no way to make things darker. You can only add light, not subtract it. That means your AR interface will be basically unreadable in sunlight. The power problem is the real killer, though. Oculus is only successful because it's tethered to a high-end rig. When you don't have that, you're basically tethering a mobile phone to your eyes. Not only do mobile phones die within a few hours if you play 3D games on them continuously, but their graphics aren't very impressive. There's a third problem worth mentioning: When you're using AR, you're making an overlay on the real world. The real world is instant. Your overlay has to be at least 90fps to look seamless. That means you have to recognize a target + render the overlay within 11 milliseconds. That recognition step is a long way from being that performant. I predict Magic Leap will be shown to be an elaborate ruse, and they were able to raise money mostly by tricking investors with tech demos that can't possibly advance any further. It's easy to make this claim, but the fundamental problems with the technology seem insurmountable on our current tech tree. It's not quite the same thing as being negative toward a new startup -- most startups have a chance of succeeding if they get users. But you can't get users with AR without wishing for new advances in technology that seem beyond startups' capabilities for the moment.
- swolchok 9y ago
  > there's no way to make things darker https://en.wikipedia.org/wiki/Liquid-crystal_display#Twisted_nematic_.28TN.29 https://en.wikipedia.org/wiki/Liquid-crystal_display#Twisted...
  - sillysaurus3 9y ago
    I thought so too, but unfortunately this doesn't work either. The reason it won't work is the same reason scratches on your eyeglasses don't appear sharp. Your focus is in the distance, so you won't see whatever dimming has been applied Not unless you use it to darken e.g. an entire quadrant of your lenses. Interestingly, this problem applies to text rendering too. It seems unlikely that you'll be able to project clearly readable text over anything when it's an inch away from your eye.
    - jon_richards 9y ago
      Can you elaborate? Isn't the basis behind VR that you can show different images to each eye to trick people into thinking they are looking at something far away? Is there something about the cornea/lens shape that prevents this from working with AR?
      - sillysaurus3 9y ago
        Yeah, so the reason VR works is that it controls the entire scene. The light being projected towards your eyes is from a consistent focal plane -- the light comes from a flat surface located exactly N inches away from your eyes. That means you can focus sharply on it. AR is a very different beast. In the real world, your focus is usually several feet away, or sometimes across the room. Your focus also changes almost instantly. Yet AR still needs to project light toward your eye from exactly N inches away. You can simulate this: Find some glasses and put some black tape over part of the lens. You'll notice the tape doesn't appear sharp and crisp. It's a blur. (Or just close one eye and put a finger close to your other, and look out in the distance.) I'm not sure AR has any hope of rendering text in a readable way. It'd be like writing some letters on your finger, then putting it up to your eye while looking out in the distance. Even if the letters glow, so that they're a light source, you still can't read it at all.
gdulli 9y ago
> AR is posing to really take off. Even if that turns out to be true, the pattern of 1-2 Minecraft-like winners and 99.9% losers is likely to hold. Existing platforms are already accessible to indies. Though if your goal is to make it a hobby instead of a living, "losing" on income may not bother you.
WalterSear 9y ago
IoT, AI, biotech and possibly VR/AR (it's too soon to tell) are so development intensive, they are the purview of giant corporate teams, not indie developers. The web spoiled us. It's hard problems from hereon out.
- eksemplar 9y ago
  I think there are a lot of opportunities for indie development. We build a lot of more-story elderly housing and in those you have a doorbell and gate opening system. This used to be pretty costly because you needed a professional system with wiring and what not. These days we do it with IoT. Basically we run our own private and encrypted internet, LoraWAN, and we setup a small $10 door control in each apartment which connects to the main door via our slow private net. The components and software was developed by a single student at a hackaton. Similarly we have a lot of town history on record. We've given people access to this, and a group of student went around town setting up 957 little Bluetooth transmitters which emit these stories to their app all over town. Another example is a medicine cabin that monitors it patients/elders take their meds and automatically alarms staff if they don't which is something we offer today. The plastic bag and electronics that does it were developed by a small local bakery with two employees. I do agree with you in that some areas are going to require google sized companies. But things like the raspberry pi and all these small/cheap "standard" components which anyone can build themselves are really shaping up to change a lot of things and it's not big companies who are driving it forward.

sillysaurus3 9y ago

AR isn't really going to become a thing. Not with current technology, anyway. The problem is twofold: There's no way to get enough power to the device to do anything interesting, and there's no way to make things darker. You can only add light, not subtract it. That means your AR interface will be basically unreadable in sunlight. The power problem is the real killer, though. Oculus is only successful because it's tethered to a high-end rig. When you don't have that, you're basically tethering a mobile phone to your eyes. Not only do mobile phones die within a few hours if you play 3D games on them continuously, but their graphics aren't very impressive. There's a third problem worth mentioning: When you're using AR, you're making an overlay on the real world. The real world is instant. Your overlay has to be at least 90fps to look seamless. That means you have to recognize a target + render the overlay within 11 milliseconds. That recognition step is a long way from being that performant. I predict Magic Leap will be shown to be an elaborate ruse, and they were able to raise money mostly by tricking investors with tech demos that can't possibly advance any further. It's easy to make this claim, but the fundamental problems with the technology seem insurmountable on our current tech tree. It's not quite the same thing as being negative toward a new startup -- most startups have a chance of succeeding if they get users. But you can't get users with AR without wishing for new advances in technology that seem beyond startups' capabilities for the moment.

swolchok 9y ago
> there's no way to make things darker https://en.wikipedia.org/wiki/Liquid-crystal_display#Twisted_nematic_.28TN.29 https://en.wikipedia.org/wiki/Liquid-crystal_display#Twisted...
- sillysaurus3 9y ago
  I thought so too, but unfortunately this doesn't work either. The reason it won't work is the same reason scratches on your eyeglasses don't appear sharp. Your focus is in the distance, so you won't see whatever dimming has been applied Not unless you use it to darken e.g. an entire quadrant of your lenses. Interestingly, this problem applies to text rendering too. It seems unlikely that you'll be able to project clearly readable text over anything when it's an inch away from your eye.
  - jon_richards 9y ago
    Can you elaborate? Isn't the basis behind VR that you can show different images to each eye to trick people into thinking they are looking at something far away? Is there something about the cornea/lens shape that prevents this from working with AR?
    - sillysaurus3 9y ago
      Yeah, so the reason VR works is that it controls the entire scene. The light being projected towards your eyes is from a consistent focal plane -- the light comes from a flat surface located exactly N inches away from your eyes. That means you can focus sharply on it. AR is a very different beast. In the real world, your focus is usually several feet away, or sometimes across the room. Your focus also changes almost instantly. Yet AR still needs to project light toward your eye from exactly N inches away. You can simulate this: Find some glasses and put some black tape over part of the lens. You'll notice the tape doesn't appear sharp and crisp. It's a blur. (Or just close one eye and put a finger close to your other, and look out in the distance.) I'm not sure AR has any hope of rendering text in a readable way. It'd be like writing some letters on your finger, then putting it up to your eye while looking out in the distance. Even if the letters glow, so that they're a light source, you still can't read it at all.

gdulli 9y ago

> AR is posing to really take off. Even if that turns out to be true, the pattern of 1-2 Minecraft-like winners and 99.9% losers is likely to hold. Existing platforms are already accessible to indies. Though if your goal is to make it a hobby instead of a living, "losing" on income may not bother you.

WalterSear 9y ago

IoT, AI, biotech and possibly VR/AR (it's too soon to tell) are so development intensive, they are the purview of giant corporate teams, not indie developers. The web spoiled us. It's hard problems from hereon out.

eksemplar 9y ago
I think there are a lot of opportunities for indie development. We build a lot of more-story elderly housing and in those you have a doorbell and gate opening system. This used to be pretty costly because you needed a professional system with wiring and what not. These days we do it with IoT. Basically we run our own private and encrypted internet, LoraWAN, and we setup a small $10 door control in each apartment which connects to the main door via our slow private net. The components and software was developed by a single student at a hackaton. Similarly we have a lot of town history on record. We've given people access to this, and a group of student went around town setting up 957 little Bluetooth transmitters which emit these stories to their app all over town. Another example is a medicine cabin that monitors it patients/elders take their meds and automatically alarms staff if they don't which is something we offer today. The plastic bag and electronics that does it were developed by a small local bakery with two employees. I do agree with you in that some areas are going to require google sized companies. But things like the raspberry pi and all these small/cheap "standard" components which anyone can build themselves are really shaping up to change a lot of things and it's not big companies who are driving it forward.