Thoughts on wireless hardware

[bzar]

Well, as another usage scenario: I only listen to internet radio, since all the FM stations (save for one that's on the air one week per year, and they have internet streaming too) only feature generic list music which I don't prefer, news I get faster and better from the internet, weather I get faster and better from the internet and commercials I'd rather live without. An FM receiver adds no use cases for me, but a cellular modem adds plenty. My cell phone battery can't take working as a modem as long as I expect Pyra to so that's no solution either. Unlimited cellular internet plan is 6-13€ per month (alternatively 1€ per day the plan is used) here depending on speed, so that's not too bad either. For some of us internet streaming is just a better option.

 

[Granitehead]

I don't doubt that, but that might not be true for everyone. I think we should get on with the poll instead of discussing.

So we have:

GPS - ?

Cellular radio/3G/etc - ? (probably included as an option anyway)

magnetometer/compass (especially useful in combination with GPS) - $5

FM/FMHD/DAB/DAB+ - $11 and easy to integrate (or with just FM/AM for similar prices. Do we need AM?)

DVB - $5-$15 but hard to integrate

IR - <$1

We are still missing prices and also:

I would be happy paying more than <estimated additional cost of feature> for <feature>

I would be happy paying less than <estimated additional cost of feature> for <feature>

I would not want to pay any extra for <feature>

I would be put off buying the Pyra if it had <feature>

Maybe your more money-oriented approach is better. I would like to suggest some minor changes. For clarification, mainly.


I would be happy even paying more than <estimated additional cost of feature> for <feature>

I would be happy paying <estimated additional cost of feature> or less for <feature>

I would not want to pay any extra for <feature>

I would be put off buying the Pyra even if it had <feature> for free

Changes in 1+2 made to make it clear which one people should choose who would be okay with about <estimated additional cost of feature> as a limit.

Changes in 4 made to further distinguish from 3; 4 isn't supposed to be about money.

It came to my mind that if we are making a nice poll about additional wireless hardware, why not include other types of hardware too, to have just one poll for all additional hardware? Say stuff like accelerometers, gyroscopes, cameras, etc.

There is lots of discussing going on about stuff about which it doesn't make sense to discuss because there is no base to the discussion. To really be able to argue against/in favour of the inclusion of <feature> we really need to know how many people would like it.

 

[WizardStan]

Thoughts on cellular:

First off, GSM is out. While it is possible to get quad band GSM modules for as little as $25 which would cover almost everyone on the planet, 2G speeds are simply not sufficient. People are already complaining when their wifi can barely reach 30KB/s, EDGE is designed for 18KB/s; you cannot watch netflix at that speed, I wonder if you could even youtube reasonably. 2G is great for emails and GPS, just download the map once and only periodic updates as you move so it's ok to be slow, but you start browsing your facebook at 18KB/s and you will want to tear out your hair. A lot of carriers are dropping GSM support now, as well, so longevity of putting it into the Pyra is questionable.

LTE is still being rolled out, and conditionally replaced already, getting something working here is probably not a good idea.

That leaves us with UMTS/HSPA: 3G speeds are more than enough for current uses and it's basically deployed globally.

Even so, to be global, we need a multiband module. Quectel sells such a module, the U10: a tri-band UMTS board that covers almost everywhere for 95USD. It would be better if it were quad-band (or even 6 band) but the ones I found that could cover that much were $150+. As a bonus, the U10 also includes a quad-band GSM module (presumably because once the transceiver is in, GSM is a virtually free feature) so you can use that if you really want/must. Alternatively, they also sell another module, the UC20 for 60USD which comes in two flavours: one for UMTS850/1900 (ie, Americas) and UMTS900/2100 (ie, everywhere else), plus quad band GSM of course; this would allow ED to produce to versions without needing to change the board at all, just alternate the modules. If they also had UMTS1700 I would be SO happy, but AWS is a fairly underutilized frequency band and I don't expect to ever find a module that supports it, but I digress. There are a number of other companies that produce region specific UMTS modules (some supporting up to 8 different bands, which makes me very very happy) that could be used instead but they're mostly automotive related and I can't find any details on prices: even resellers say you need to contact them, but let's go with what I found and assume it applies across the board: $60 and you need to pick your frequency band when you order, or take a $60 saving and don't include the module at all.

Issues with this: CC is now doing different population. This probably wouldn't cause a major problem, they'd just produce one type one day and another type the next. What would be problematic is figuring out how many of each to order. Suppose the minimum quantity is 1000, so he orders 1000 of each: because of the "no radio" option, does he then get left with 1600 unused modules even after selling 10'000 Pyras? There's also the certification required, several thousands of dollars is required to do the tests and you want to try and spread that across as many users as possible. Finally there's the size: these modules seem to measure around 30mmx50mm: basically 1/3 the of the current board would be consumed by this module. If it can't be placed directly on the board (and I'm almost certain there won't be the room for it) it'll need to be stacked, which means increasing the thickness by 3 or 4mm instead.

As a bonus, in my searches I noticed that Quectel also produces a very good GPS module for 20USD that measures 10x10mm: about the size of one of the key contacts, relatively cheap, wouldn't require certification, and is really simple to interface with. I'll look into these modules later, I may be able to find something even better/cheaper.

 

[WizardStan]

There's a company called "Cinterion" that also produces a line of M2M modules, the PLS8 (US and E flavours), which could be swapped in and out and around as desired, but the only source to buy them I can actually find is alibaba and you don't trust that for prices and certainly not for future existence. What I can find though puts them in around $20-$35. It's a 30mmx30mm piece, and as an added bonus actually includes GPS built into it.

If someone can find a legit source to buy this PLS8 (or something similar) in quantity that isn't a risk of vanishing overnight, it might be viable.

edit: ahah, Cinterion is the product line, the company is called Gemalto. They are conveniently German. ED may want to talk to them.

 

[Natsu]

Why is certification required? Shouldn't that be the job of the supplier?

 

[Binky]

Because its a radio transmitter of sorts, and governments like to regulate who can transmit what?

 

[WizardStan]

Every electronic device produces electromagnetic interference. Everything needs basic certification to demonstrate that it won't interfere with other devices even if they don't have a radio. When they're intentionally broadcasting a signal (like wifi) there's even more certification. When the broadcast range exceeds a certain limit (like 30 meters or something, I'm not sure exactly how far, basically cellular signal) there is even more certification. Basically they want to make sure your radio transmitter isn't going to knock out a plane flying over head or something.

 

[Natsu]

From my understanding, Quectel M2M modules are already built to certification.

 

[slaeshjag]

That doesn't mean anything.

If you integrate it into a device, it's a new device. A new device that needs a complete re-test.

The only thing the certification on that module tells you is that the module in question passes the requirements on its own.

 

[Natsu]

That doesn't mean anything.

If you integrate it into a device, it's a new device. A new device that needs a complete re-test.

The only thing the certification on that module tells you is that the module in question passes the requirements on its own.

Yeah, I figured after that post. Such a pain for the seller.

 

[thatgui]

I can't argue much against it

You argued against an FM receiver because it adds cost, adds size, adds complexity of integration, requires official support, and will be underutilized.The xG module adds even more cost, requires more space, is signifcantly more complex, requires a lot more in support, and regardless of the number of use cases you can think of it will still be extremely underutilized for a number of reasons.

Every argument you gave against radio applies even more to the xG module. The only argument in favour of the module vs the radio is that ED mentioned an interest in it. If you're going to argue for one thing over another it is imperative that you use arguments that can only be applied to the thing you don't want.

To answer your rhetorical question "Why not put in an RC Tranceiver to control RC - Cars and such ?", because no one asked. If no one asks then there's no point in investigating it. Do you seriously want such a feature? Do you want me to look up some stuff on it and demonstrate why it specifically wouldn't be a good idea?

Sigh, as it seems (as often) i'm not able to "deliver" my thoughts correctly:I reread my posts int this thread and I simply can't find anything about me arguing pro including a xG module - all I did was offering the functionality of a xG module to compensate for the lack of a radio module - which isn't a certainty (but I can't see where I hinted that either) as Ed only stated that he is thinking about it.

So while it is true what you are saying, I don't see any connection to the argument at hand, as it never was a "why not a radio module if xG is included"

or "why include a radio module if we already have a xG module" discussion at all.

I'm sceptical about including a xG module myself, and whenever it came up in the past I always argued against it with the arguments mentioned (needs certification, etc.) following Eds leed. After Ed changed is mind about it(or at least is less disinclined), I changed my "direction" in that particular case to "whatever".

As to the "RC Transceiver", that was just a "drop in" example, you can put anything you can think of in it instead, that meets the

- cheap hardware costs

- could be practical

- appeals to a minority

requirements. In my opinion including a radio module is clearly in the "featuritis" category: including stuff that could be usefull, but generally isn't very usefull, but adds to the complexity of the device (in several aspects), like built in ir module (and no "I" want it included, not as an addon), built in home automation control (like x10), citizen band radio, etc *.

*Disclaimer: just to be sure, I have no interest in any of those beeing included

Suppose 10% of users save  $1000 per year by having a radio receiver instead of streaming.

Wouldn't that require that there are 10% of Pyra users that will listen to radio with their pyra regardless of the costs (or the availability of a cheap radio receiver, they probably already have) ?

Does it cost $100 per device to add a chip, circuit board tracks and pads, and deal with any additional production issues (including returns due to failure)?

Don't know, but where is the number 100 coming from? If your hinting at the "but its so cheap" argument - please see above ("featuritis").

Big companies have been phasing out buttons and clamshells too. They make production much harder. Maybe the Pyra shouldn't have either of those?

I doubt you can draw such clear relations here, buttons aren't really needed as there is already another method of input. Why drive up costs and complexity of the product if it is not necessarey at all? And please don't make a comparison to the Pandora, the Pandora was specifically designed with buttons in mind as there was a niche that big manufacturer did not want / were not able to cater. I don't see a niche for a gadget with a readio module, nor are there posts in the past where people wished the Pandora had a radio module included, but I can see a lot of posts praising the buttons. That does not mean there are not more people than the few that argue pro including it, all I'm saying is it is a niche in a niche that is not worth catering. It nearly the same for clamshell, I doubt the clamshell design was introduced just for the sake of the Pandora beeing a clamshell device. I guess the question at hand was at that time "where to put the screen" and it was either another requirement or a nice side effect that utilizing a clamshell design would also protect the controls. Again for big manufactueres this wasn't a big problem, as buttons where designed flat or weren't even there.

There is lots of discussing going on about stuff about which it doesn't make sense to discuss because there is no base to the discussion. To really be able to argue against/in favour of the inclusion of <feature> we really need to know how many people would like it.

Looking at official polls (= made by Ed) beeing the only ones that exceed 100 votes, I guess we won't get any reliable numbers (pro or contra including something).

 

[Natsu]

It would kind of be a waste tho, since I just got a phone that can use 4g.

 

[WizardStan]

I reread my posts int this thread and I simply can't find anything about me arguing pro including a xG module - all I did was offering the functionality of a xG module to compensate for the lack of a radio module

You provided many arguments why an FM radio was a bad idea. You then suggested that an xG module would be a better replacement: I did not mean to imply that you said it WOULD be included, simply that your arguments were saying that an xG module would be a better replacement to a radio receiver.I used your same arguments to demonstrate that if the FM radio is a bad idea then an xG module is even worse (for the reasons you gave). That is the extent of my point: you provided a poor argument against FM radio which I was then able to use against the xG module. If you want to propose the module vs an FM radio then you need to use arguments that can only be applied to one and not the other. Either that or simply agree that if the FM radio is a bad idea for the reasons you gave then the xG module would be even worse and that neither should be used.

As to the "RC Transceiver", that was just a "drop in" example, you can put anything you can think of in it instead, that meets the

- cheap hardware costs

- could be practical

- appeals to a minority

No. There is one more important thing required: someone has to ask for it. Yeah, we could drop any number of examples in but unless someone actually asks for it there's no point in discussing it. As soon as someone asks for it and gives a coherent example of how it would be used it becomes an interesting thing to discuss.No one asked for an RC transceiver, no one asked for 7.1 channel audio, no one asked for an underside touchpad: the list of things that are cheap, could be practical, and appeal to a minority is endless, but no one proposed them. Once you add that fourth factor, someone asking for and explaining what it is for, it becomes a point of discussion.

I understand where you're coming from, you're afraid of feature creep, but rejecting features for the sole purpose of avoiding feature creep is incredibly arbitrary. Why consider a cell transceiver but reject immediately an FM/DAB receiver? We need debate! We need discussion! We need to actually consider why one is better than the other. Neither is strictly better than the other: a cellphone is more feature rich than a radio receiver and can be used to stream radio (thus seemingly being a superfeature of the radio), but the radio receiver is smaller, cheaper, and significantly easier to use.

In my opinion including a radio module is clearly in the "featuritis" category

That's a fine opinion to have. Opinions are great, they help us quickly evaluate the world around us and make decisions when there isn't time to think about the choices. We have lots of time here, weeks if not months: explain why other people should share this opinion, carefully, without using an argument that could apply to something else. Do it wrong and I am not above using your own arguments to prove that a radio receiver would be better than speakers.

 

[Binky]

where is the number 100 coming from?

I was thinking in terms of overall gain.

If 10% of users get $1000 benefit from the feature (that is, the savings in mobile download costs for audio streaming 30min per day for a year), then the average benefit per user is $100.

If the cost of the parts, design effort, repairs, etc is less than $100, the feature is, overall, of benefit.

This is only a rough example though, so don't take it too seriously

 

[bzar]

Potential savings is not savings. Would they stream that much if they had to pay $1000?

 

[xiongxioi]

I usually agree with you, Brinky. But not on this topic, because you have not consider the effect of raising the pyra's price tag by $100 on most people. 

Sure, one can keep the Pyra for years and years unlike most smartphones, but it is still a negative because most people would rather have a quick discount than a prolonged benefit which is only a promise. It's also true that most of us would buy the pyra regardless if it can emulate the gamecube, or what not. Having all those things is not a total deal maker for most people, but making it $100 more expensive is likely to be a deal breaker. (have a look at the xbone)

Of cause, in reality, it would not be costing that much. I would be totally fine paying 10 or 20 bucks extra for FM radio. Even though, if I have to choose, I would say spend that extra money on prettier case and better buttons. 

 

[thatgui]

You provided many arguments why an FM radio was a bad idea. You then suggested that an xG module would be a better replacement: I did not mean to imply that you said it WOULD be included, simply that your arguments were saying that an xG module would be a better replacement to a radio receiver.

I used your same arguments to demonstrate that if the FM radio is a bad idea then an xG module is even worse (for the reasons you gave). That is the extent of my point: you provided a poor argument against FM radio which I was then able to use against the xG module. If you want to propose the module vs an FM radio then you need to use arguments that can only be applied to one and not the other. Either that or simply agree that if the FM radio is a bad idea for the reasons you gave then the xG module would be even worse and that neither should be used.

I understand what you did (or not) but I see now where my error lies. While I never suggested that an xG module would be a better replacement - I only stated that it can be used to compensate. I guess the not clevery placed "can" is the problem here. I should have use "could" instead to make it more clear, that I see an xG module mereley as another option not as a feature to "repell" other request. I did mention another alternative though: USB Dongles. If I was so inclined about presenting an xG module as the "better solution" would it make sense to mention alternatives ?

I used your same arguments to demonstrate that if the FM radio is a bad idea then an xG module is even worse (for the reasons you gave). That is the extent of my point: you provided a poor argument against FM radio which I was then able to use against the xG module. If you want to propose the module vs an FM radio then you need to use arguments that can only be applied to one and not the other. Either that or simply agree that if the FM radio is a bad idea for the reasons you gave then the xG module would be even worse and that neither should be used.

I'm sorry, but I can't make much of this. If my argument against FM radio is that poor, why is it working against an xG module well ? Its either poor or not, if you are able to turn it against me. Which in itself is something I don't understand, as I never stated that the xG module could be a substitute for a radio module, so why writing about the same theme of argumentation ("I turned your own argumentation against you")over several posts while there is no real reason.The same with your last sentence. This implies that both options would offer the same amount of advantages that would need to be weighed against the same amount of disadvantages. Which simply isn't the case => but I'm still not arguing for including a xG module.

No. There is one more important thing required: someone has to ask for it. Yeah, we could drop any number of examples in but unless someone actually asks for it there's no point in discussing it. As soon as someone asks for it and gives a coherent example of how it would be used it becomes an interesting thing to discuss.

No one asked for an RC transceiver, no one asked for 7.1 channel audio, no one asked for an underside touchpad: the list of things that are cheap, could be practical, and appeal to a minority is endless, but no one proposed them. Once you add that fourth factor, someone asking for and explaining what it is for, it becomes a point of discussion.

I understand where you're coming from, you're afraid of feature creep, but rejecting features for the sole purpose of avoiding feature creep is incredibly arbitrary. Why consider a cell transceiver but reject immediately an FM/DAB receiver? We need debate! We need discussion! We need to actually consider why one is better than the other. Neither is strictly better than the other: a cellphone is more feature rich than a radio receiver and can be used to stream radio (thus seemingly being a superfeature of the radio), but the radio receiver is smaller, cheaper, and significantly easier to use.

I beg to differ - unless this is some last chance to include some features phase we are currently in. Who says that such a feature will not be proposed in the near future ? Putting the radio module in, puts you in an argumentational disadvantage. Or how do you want to repell a request for a feature that only a minority of the userbase will really make use of, if you can only hope that its a totally obscure request where you can find enough reasons to argue against ?

That's a fine opinion to have. Opinions are great, they help us quickly evaluate the world around us and make decisions when there isn't time to think about the choices. We have lots of time here, weeks if not months: explain why other people should share this opinion, carefully, without using an argument that could apply to something else. Do it wrong and I am not above using your own arguments to prove that a radio receiver would be better than speakers.

But you did nothing. You turned arguments against me were there wasn't even a base to do so, leaving other factors completely aside. In the end you had no real argument pro including a radio module other than "its cheap" and "I see use cases for me and a small portion of the user base" - thats it.

I was thinking in terms of overall gain.

If 10% of users get $1000 benefit from the feature (that is, the savings in mobile download costs for audio streaming 30min per day for a year), then the average benefit per user is $100.

If the cost of the parts, design effort, repairs, etc is less than $100, the feature is, overall, of benefit.

This is only a rough example though, so don't take it too seriously

But (as said alread in the last post) this would require that 10% of the future Pandora users will definately use an xG connection to stream radio. What about the users that have a different dataplan, or that even do not have to pay for their dataplan (like me), or simply continue to use the device they already use to listen to readio on the go?But how much of your rough example would really be added to the unit price of the Pyra ?

 

[WizardStan]

If my argument against FM radio is that poor, why is it working against an xG module well ?

It's poor argument BECAUSE it can be used against the cell module. They are excellent arguments against an FM receiver, but what makes them excellent arguments against an FM receiver makes them even better against the 3G (I'm just going to forgo the xG from here on, 2G is not enough for streaming, got to be at least 3G) module, and a lot of other features that are already being included. Whether you used the word "can" or "could" or whatever is irrelevant, the arguments you made to demonstrate that an FM receiver is a bad idea work even better against a modem.

You turned arguments against me were there wasn't even a base to do so, leaving other factors completely aside.

There were no other factors. You said FM was bad because it added cost, size, complexity, and would be underutilized. That's it. You didn't mention anything about any other factors, that was the extent of why you said the radio receiver was bad. Those exact same arguments apply stronger to 3G: it adds more cost, adds more size, much more complexity, and can potentially be even more underutilized.

The same with your last sentence. This implies that both options would offer the same amount of advantages that would need to be weighed against the same amount of disadvantages.

I'm sorry, I didn't mean for it to imply that. All I'm saying, all I have been saying (all I will say again and again) is that you gave reasons that a radio module should not be included, and that those exact reasons apply to a modem: if you are against an FM receiver because it adds cost, space, complexity, and will only be used by a minority then you should be against a 3G modem for those exact same reasons. If you are genuinely anti-radio and pro-modem and you want to see it included then you must have arguments why it is better than a radio receiver, yes? Arguments that don't involve size, cost, complexity, and user adoption. Give those arguments!

Who says that such a feature will not be proposed in the near future ?

That's fine, we can discuss it when the feature is proposed. Trying to figure out all the things someone might propose in the future is a exercise in futility and rejecting features for the sole reason of preventing feature creep is still arbitrary. How can you decide what features are genuinely useful and which aren't if you flat out reject a feature without discussion?

Or how do you want to repell a request for a feature that only a minority of the userbase will really make use of, if you can only hope that its a totally obscure request where you can find enough reasons to argue against ?

I don't want to repel requests for features at all, I want to discuss them. If someone proposes a feature and neither I nor anyone else can find a reason to argue against it then it should absolutely be included. Literally, what you are saying is that some future feature has been suggested and no one can find even a single reason not to add it. If there are absolutely no reasons not to add it then where is the problem with adding it? If you do have a problem with adding it, even one tiny little doubt, then that's an argument against it so we have a discussion and grounds to reject the feature. Ultimately it doesn't matter, we're just discussing features for the fun of it and ED is the one making the final decisions, regardless of whether we as a forum decide a feature is useful or not. Eventually there'll come a point where ED says "the prototype is done, no more features can be added" (in truth, ED will probably stop considering features long before then but there's absolutely no harm in people discussing them here) then you can go ahead and reject all the features you want without discussion because by then it is too late. Until then, where's the harm in discussing features that people would like to see?

In the end you had no real argument pro including a radio module other than "its cheap"

That's because I'm not pro-radio. I'm not anti-radio either, all I've done is laid out the facts on the various radio modules. I'm not even pro or anti-3G, I've also just laid out the facts on the various modules.I'm anti-bad-argument. Everything I've said to you has been for the purpose of demonstrating that if you are anti-radio for the reasons you gave then you should also be anti-3G module for the exact same reasons, and that if you want to prevent feature creep then the best way is through discussion and not arbitrarily rejecting ideas that you think are "features for features sake" without evidence.

I never stated that the xG module could be a substitute for a radio module,

Yes you did. From page 4

both functionalities can be "compensated" with a xG modem

You made arguments against a radio receiver and then said that if a 3G module were included it could be used for those same features. I have been saying this entire time that if you are against a radio receiver for the reasons you gave then you should be even more against the 3G module because the reasons are even stronger for it. You shouldn't have been discussing about how a modem could compensate unless you have other arguments that make a modem strictly better than a radio receiver, and if you do I want to hear them because from my standpoint, everything I know about both, depending on what your concerns are, neither is necessarily better than the other.

 

[Binky]

To everyone who has been poking holes in my argument about the average costs of radio listening:

I  agree that its a weak argument. I came up with it when someone was suggesting that streaming over 3G was a viable alternative, to illustrate the huge savings to be made for radio listeners by having a real radio module.

Its also rather convoluted, with hypothetical costs and values everywhere.

I never suggested that anyone would have to pay an extra $100. The 10% of users would listen to radio if they could figure is about as about as meaningful  as the statistics in a politician's speech.

Please stop disagreeing with something that hardly made sense in context, let alone out of it!

 

[levi]

Internet or DAB radio would be a nice to have for me, but with two large SDs on board, I can include a large number of CDs I've probably only ever listened to once or twice so far, so I can limit the times I feel I need to listen to completely new music to times when I have wifi access.


I wonder whether having on-board GPS would enable me to stop relying on my aging Nokia phone for mapping? There seem an incredible number of graphical GPS apps for GTK Linux, none of which I've tried out yet. With a suitable map viewer and selected OpenStreetMap data pre-loaded I feel it could give my phone a run for its money at providing me with maps while I'm out and about, and if it included an electronic compass it would probably even surpass my phone. Has anyone tried out these apps on the move to see how realistic it is to have this feature using pre-loaded map data should we have access to a GPS module? Or should I expect to have to look for another device to replace my phone for mapping purposes?