Live data from Hacker News

Introducing the Photon

spark.io

51–60 of 100 posts

Re: Introducing the Photon

#52

Spark seems very cool, but I think what worries me as a product builder is any gaurentees that the cloud service will be around in a couple years. Certainly not suggesting that it won't be, just that if I build a product with this, I will want my own version of the cloud as an open source project that I can license or something.

The cloud back-end is open source so you can do that:

http://www.github.com/spark/spark-server

Re: Introducing the Photon

#53
Correct me if I'm wrong, but this is similar to the Electric Imp in that you have to connect via their cloud service (you can not just make requests directly to any web endpoint). While this might be cheap, and great for prototyping, I dont want to build on top of something that locks me into your cloud service. They get you to buy a cheap piece (pricepoint) of hardware, but then you have to pay them to connect.

I'd rather use an RaspberryPi or Arduino Yun. While more expensive up front, they allow you to connect to any service you would like.

Re: Introducing the Photon

#54

Earlier quoted context omitted.

Interesting, I generally have a fairly positive opinion of the Imp. In particular the integration with Squirrel means that much more complex things can be done, more easily, and without touching the running hardware. The Spark (Core) on the other hand is much like an Arduino in that you have it in your presence to code its behavior, and then release it to the wild. Slightly different target apps, hugely different use…

Electric Imp is trying to patent (something around) using a light and a photo diode for communication. That's not a good sign, but enough about that, the Photon (assuming they actually ship) and its price point are much more interesting to me as a hobbyist. They also have schematics and gerber files for their hardware which is more than I can say for Electric Imp. > In particular the integration with Squirrel means t…

   > How much of the Imp's advantage here is due to features 
   > of Squirrel the language itself, and how much is due to 
   > their 'cloudification' of the ecosystem?
I believe it all relates to the cloud aspect of it. The systems model is that the Imp is simply an intelligent channel controller at the end of the connection. This gives two advantages, one is that "infinite" state can be kept off chip, and the other is that updates can occur without the modules participation.

I don't doubt that a similar setup could be recreated for the Spark with some standardization on what the Photon expects from the network.

As for the patent, at least what they are trying to patent is actual hardware not some 18th century practice but done with a computer :-)

Re: Introducing the Photon

#55

This looks really, really interesting. It also seems like they're thinking a lot about tooling, which I appreciate - there's a lot of stuff out there right now, I assume the true differentiator will be in the tooling. On that note, there's a weird "don't go back" loop at https://www.spark.io/signup .

Tooling? You mean software toolset or mechanical tooling?

Re: Introducing the Photon

#56
post #24
post #3

This seems like a version of the Tessel[0] that isn't as overpriced[1]. [0] https://tessel.io/ [1] https://shop.trycelery.com/page/tec

You can program Spark using JavaScript too: https://github.com/spark/sparkjs

Those two don't seem to be the same thing. Tessel runs javascript on the micro so you can write your programs in js instead of C and compile it.

What you've linked to is a js library for connecting to Spark cloud.

Overall I think the Spark is much better bang for your buck.

Re: Introducing the Photon

#57

Correct me if I'm wrong, but this is similar to the Electric Imp in that you have to connect via their cloud service (you can not just make requests directly to any web endpoint). While this might be cheap, and great for prototyping, I dont want to build on top of something that locks me into your cloud service. They get you to buy a cheap piece (pricepoint) of hardware, but then you have to pay them to connect. I'd…

Nearly everything is open source; firmware, hardware design files, and an api compatible server: https://github.com/spark

Re: Introducing the Photon

#59

Spark seems very cool, but I think what worries me as a product builder is any gaurentees that the cloud service will be around in a couple years. Certainly not suggesting that it won't be, just that if I build a product with this, I will want my own version of the cloud as an open source project that I can license or something.

The cloud back-end is open source so you can do that: http://www.github.com/spark/spark-server

How hard is it to reprogram the device to use your own server using that back end? How about if you wanted to do your own back end?

Re: Introducing the Photon

#60
post #43

I know nothing about IOT. Is this revolutionary? It does look awesome.

The IoT revolution is inherently price conscious. It's not that any of this stuff is new, it's that I can buy a device cheap enough to consider adding wifi to everything, including a temperature sensor on my toaster, to show me temperature graphs for the perfect toast. It could be done 10 years ago, but spending $200 to internet-enable my toaster is ridiculous, but for $25 ($20 + a temperature sensor), it's a fun and…

A temperature sensor in the fridge would be much more useful than one on the toaster. I may have to get me one of these.
Post reply on HN