I feel like this is something that needs to happen. The observation about $80 GSM shields is spot on. Amazing things would happen if we could have a $30 IoT board with built-in LTE and a country-by-country easy-to-follow guide to how to buy a SIMM and get online.
My one critique of the project is that I think the most important part (from a marketing perspective) is the "make your own connected IoT board for cheap!". The OS part is a huge enabling factor but not what gets me pumped about it.
I almost skipped clicking on this link because I was imagining strapping a full heavy smartphone (plastic case and all) onto a drone with zip ties.
RadioShack is closing a lot of stores, many of which sell Arduino things. I got two ~$60 MSRP GSM shields for $24 a piece. Pair that with the minimal T-Mobile $3/month sim ... that's a pretty good deal.
For anyone interested, the $30 smartphone seems to be from a special Christmas time deal from CherryMobile [1] (a Filipino phone retailer) and may only have been valid at brick and mortar stores. It'd be much more reasonable to compare something available online with worldwide shipping even if it is out of active production. A Geeksphone Keon is comparable in price to an Arduino with equivalent shields and a heck of a lot easier to work with if you're not a hardware person.
The 30$ phones are based on the Spreadtrum SC6821 chipset and are sold around Asia. There are four different ones in India alone (Intex Cloud Fx f.e.) that go for this price.
Probably can get some SC6820 phones as well and make a port. Don't know if those are readily available in western countries.
Its though to get your hands on the super cheap handsets in US / Europe. But if there's sufficient interest we can go to an odm and just order 20,000 for <25$.
Edit: the cherry mobile phone went for 23$(!!) During their intro offering
I went to the trouble of getting an Intex phone. Even for its price it's not worth it - even if I can build the JanOS/FFOS sources I don't think I have the driver to flash the phone.
If anyone is interested in US cell plans for this type of application, I've used US Mobile [1] in the past. For a single day or two of heavy use, a T-Mobile daily plan is probably best. But for a long period of light activity, these guys have worked very well for me, especially for SMS.
I've been casually looking around for data-only plans with no (or really low, at least) monthly fees and pay-as-you-go per MB data... I'm not coming up with much, but roamline [ https://www.roamline.com/roamline-sim ] might do the trick. No monthly feed, twenty eurocents per MB.
Does anybody know of anything else? This seems like a market that typically buys access for thousands of devices at a time, pushing just a few MB per month. There's not much out there for the hobbyist looking for just a SIM or two.
Interestingly, it looks like Truphone is free for incoming voice and SMS. So if you wanted to control something remotely... that could be extremely inexpensive!
A quick scan of their website suggests there is no monthly fee to hold the number either. Am I missing something?
Yeah, true, but the building blocks are available to port it to any cheap Android 4.x phone. I just don't have them available. I also chose this particular list because all of them ship rooted and have downloadable images available in case someone f*ks up, so I don't brick anyone's device :-)
An open source kernel also helps when you do GPIO support etc.
I have been able to use the modified System application on my own firefoxos-based boards. However, for GPIO I have used a background process based on NodeJS (which exposes the ports using a REST-api). Getting GPIO working on some phoneboards would still prove a challenge as mostly they do not have these easily exposed.
... of the GPIO on the SoC, you only have some of the exposed buttons a device might have. Far from ideal... also, at the moment FirefoxOS itself does not have a notion of GPIO as a WebAPI. But great to see this, nice article.
This is what I've been looking for. I didn't want to have to create something like this myself, but wanted to be able to reuse older smartphones lying around. I don't know if I want to use JS though.
Calling it a fork would be too much; it is actually just a modified main System application for FirefoxOS to allow it to run 'headless'. The rest of the infrastructure is nearly unmodified B2G.
My one critique of the project is that I think the most important part (from a marketing perspective) is the "make your own connected IoT board for cheap!". The OS part is a huge enabling factor but not what gets me pumped about it.
I almost skipped clicking on this link because I was imagining strapping a full heavy smartphone (plastic case and all) onto a drone with zip ties.