No prototypes for GamesCom, but a Wifi Chip and a Pyra in the dark!


Is that all that jumped out at you in the comparison post? Anything you want to say to ED about the looping brightness, the layout, something less trivial?
Trivial is in the eye of the beholder, I saw most of that spewing as trivial...
Although I could mention(again) that putting modifiers on the game controls(like the D-PAD) are a horrendous idea.
 
Last edited by a moderator:
something like 3 levels (low - medium - high) with wraparound
You forgot "Screen Off"... :)
I never use that. Do you?

If I want the screen to be off, I close the lid (or wait until the automatic screen saver turns the screen off).

I don't see the point of having the screen off with the lid open. It's confusing: you can think the screen is off because of the screensaver, so you press a button or touch a nub to bring it back, but nothing happens. After a while you remember that you turned off the screen with brightness down, so you press brightness up, only to see that the text you were editing is now full of garbage ;)

So when the screen is off, I don't want the keyboard/nubs/touchscreen to work because it can only produce garbage if you can't see what you're doing. And there's a proven mechanism for that: close the lid! :)
 
Is that all that jumped out at you in the comparison post? Anything you want to say to ED about the looping brightness, the layout, something less trivial?
Trivial is in the eye of the beholder, I saw most of that spewing as trivial...

Although I could mention(again) that putting modifiers on the game controls(like the D-PAD) are a horrendous idea.
Euro is not as important to coders, which many of you are, as the symbols that are being moved to accomodate it's placement on "E". The only modifier that would change anything with the Dpad would be the keyboard Fn. I explained this and I also said it was optional. It also wouldn't affect any of the game controls. 

Maybe I don't drink enough of the kool-aid around here or maybe it takes a lackey's perspective to dismiss these design errors.  ;)
 
Expecly the Revo k101 plus gba which also only have one brignes button..
Those devices have no "Brightness Down" function, so looping it is the only way to make it work. They also offer very limited control of the brightness to begin with.
Since the Pyra will have a Brightness Down control I see no need to cripple it with such a limited selection of brightness settings, or to loop them, which is both annoying and in this case unnecessary.

Since it may not be obvious how the brightness down control works based on the key label I would recommend ensuring this is covered in the Quick Start Guide.

-Neelix
 
It sucks that the company couldn't deliver on the promise of getting the CPU PCB to you quickly.  It worked out for them since they got the business, and hopefully it will work out well for us and the prototype is of good quality.  At least it didn't cost extra after they failed to deliver on time.

I feel that there is benefit to be gained from 5GHz wifi, but I don't think it is worth doubling the cost to get it.  I really don't feel strongly either way, so I say skip 5GHz and keep it cheaper.  I would much rather have a chance of US compatible 4G if we have to pay a little extra.

I am pretty pleased with the keyboard layout.  I hope it is decent enough for most people (sorry we couldn't fix it for everyone).  The color choice for Fn labels is pretty good.  I think it might stand out a bit more when actually backlit if it is still the one Motorola used on their later keyboards. I like that I will be able to hit F12 pretty easily, and that it is in a loction that makes sense for how I use it most. The Usr key is a touch I really like.  Do you have any ideas what 2 things it will do by default? I do have one minor complaint about, but I will keep it to myself and adapt, unless someone else brings it up.  It is just a matter of convenience for me (when using a terminal) and some others, and I don't know if my proposed minor change would cause difficulties for others.

Thanks for all the news.
 
No, it has functionality issues to it. I've posted some needed changes here but I'll reiterate them anyways for the readers.
I agree with a couple of points you make here, but certainly not all. Since I don't think this should be another keyboard design thread, let's discuss it here.

<edit>Provided that the 5 Ghz chip doesn't rule out the 2.4 Ghz band (which would surprise me if it does) then I wouldn't mind paying a bit more to future-proof it. It is in the base PCB of the Pyra which should be with us for several iterations of the CPU board. If we can already identify an upgrade to the base PCB now, then we should not postpone that and just go for the most flexible solution.

</edit>
 
Last edited by a moderator:
Maybe I don't drink enough of the kool-aid around here or maybe it takes a lackey's perspective to dismiss these design errors.
I see plenty that is less ideal with the current layout, however I'm not dismissive about it, I acknowledge what it is, I also acknowledge that nobody has a solution for the perfect keyboard and what one may consider a good layout, for another is total garbage and flawed. To me the keyboard is fine enough, I see most of what I didn't like about the Pandora keyboard is solved. To me you seem to dismiss what others are trying to tell you, inventing some of the issues you are talking about and you're obsessive to the point of being abusive trying to prove you are right. But this is just words from a kool-aid drinking lackey...
 
Last edited by a moderator:
something like 3 levels (low - medium - high) with wraparound
You forgot "Screen Off"... :)
I never use that. Do you?
If I want the screen to be off, I close the lid (or wait until the automatic screen saver turns the screen off).

I don't see the point of having the screen off with the lid open. It's confusing: you can think the screen is off because of the screensaver, so you press a button or touch a nub to bring it back, but nothing happens. After a while you remember that you turned off the screen with brightness down, so you press brightness up, only to see that the text you were editing is now full of garbage ;)

So when the screen is off, I don't want the keyboard/nubs/touchscreen to work because it can only produce garbage if you can't see what you're doing. And there's a proven mechanism for that: close the lid! :)
I frequently do, when I'm using my Pandora as a music player and I'm alone (meaning I won't put earphones in). If you shut the lid, you're muffling the speakers inside. So I leave the lid open to hear properly, but don't need the screen on.
I don't have the "typing garbage" problem you seem to think I would have - but then with the screen off, I'm not actually using the Pandy, it's just doing it's own thing

Edit: In addition, if the screen is off for about 30 seconds prior to you hitting a key, it actually wakes the screen up, rather than types you blindly
 
Last edited by a moderator:
Although I could mention(again) that putting modifiers on the game controls(like the D-PAD) are a horrendous idea.
The only modifier that would change anything with the Dpad would be the keyboard Fn. I explained this and I also said it was optional. It also wouldn't affect any of the game controls.
Remember, the modifier used to control brightness is also a gaming button, unless shoulder Fn can't change brightness. Some people may prefer using the shoulder buttons.  I was originally behind the idea of using d-pad for brightness until I understood how annoying it would be to have it messing up the display while playing a game.
 
Of course, we will NOT pay the full price for those PCBs. We won't pay more than a non-express production would've cost us... but still, it's a bit annoying.
Good. Let's hope they accept this but such things are usualy cleared before within a contract. I guess their high prices are because of the fast delivery which was not done here.
Of course, Bluetooth is a no-brainer (it only adds 1 EUR to the price anyways), but the additional 5GHz doubles the cost (from 14 EUR to 28 EUR), so that's something to think about.

The different versions are all pin-compatible, so we can use any of them without a redesign. The question is whether 5GHz is worth the additional costs.

My experiences with 5GHz weren't too good, to be honest. I didn't really get more speed, but a less reliable network connection.

What do you think? Do you use 5GHz Wifi?
I'm for saving costs to keep the price at least a little bit attractive. I don't need 5GHz, even if my Router has it. The first Pyra batch could have the cheaper version to save you and us some costs, I see no problem with that.
F11 is fullscreen for almost any X11 program. It's one of the most used keys on my Pandora when I do office work with multitasking.
Surprisingly, F11 is also often used in Windows for full screen. ;)
That was actually a long discussion. It was on two keys, but we wanted to keep a user defined key as well.

Normal press increases brightness. Once it's on brightest setting, it loops back to lowest brightness.

With Shift, you can lower the brightness.

If you additionally press Fn, you will change the keyboard brightness.
I like this idea, one button for 3 functions. However, there should be an additional one-screen indicator that shows which brightnes level currently is set during the adjusting.
Though perhaps we could have the cheaper wifi chip in the devices without 3/4G, and the more expensive one in the devices that will be more expensive anyway? You could also offer all 4 combinations, but that is just going to complicate things. Most likely, people who are willing to pay for 3/4G will not mind paying also for the extra 5GHz band, while those with a tight budget will drop the 3/4G and they will certainly appreciate not paying for the extra 5GHz band. So if you're going to have two different main boards anyway, you can just as well make them differ in terms of the wifi chip too, right?
Yupp, makes sense. I don't need that 3G/4G actualy and I don't need 5GHz. Skipping these things could make that Pyra version a little bit cheaper which may also find more customers. :)
 
Last edited by a moderator:
5ghz would improve streaming ?
No. Well maybe, kind of.There is nothing intrinsic to 5Ghz that would improve streaming: Netflix's highest rating, UltraHD, requires 25Mb/s which is well within even 802.11g's theoretical throughput, although just barely. Using 802.11n and especially 802.11ac it can easily achieve that throughput. If you're the only house on the block with wifi then 2.4Ghz will have no problem achieving the required speed.

5Ghz theoretically has more throughput but it isn't needed. From basic numbers, 2.4Ghz networks can already get maximum performance from streaming, so switching to 5Ghz wouldn't improve things any.

However, there's more to networks than just the throughput. If you are surrounded by routers operating at 2.4Ghz you will get collisions which are going to cut into your theoretical throughput to the point that it may be worth it to switch to 5Ghz. Fewer collisions combined with a higher theoretical throughput means 5Ghz is more likely to actually allow enough data through. You may also have many devices connected to the network and all want to be using the internet at the same time to suck down lots of data, in which case the 5Ghz is an improvement over the 2.4Ghz.

On the other hand, there are drawbacks of 5Ghz: shorter range, more susceptible to interferance. I know I can't use 5Ghz in my house, the signal wouldn't reach from my bedroom where the router is into the basement where my entertainment centre is. So you might think you'll get better performance paying more for the 5Ghz only to actually get worse because you're too far from the router.

_wb_ makes an excellent case for his compromise, putting 2.4Ghz into the "wifi only" and 5Ghz into the cell modem versions.
 
I'm okay with 5GHz as long as it doesn't complicate manufacturing too much.
 
_wb_ makes an excellent case for his compromise, putting 2.4Ghz into the "wifi only" and 5Ghz into the cell modem versions.
I agree, if this has no significant negative impact on volume pricing of the chips, PCB diversity (having two types of PCB will drive up cost as well) and driver diversity (though I think we can handle supporting two different WiFi drivers).
 
_wb_ makes an excellent case for his compromise, putting 2.4Ghz into the "wifi only" and 5Ghz into the cell modem versions.
I agree, if this has no significant negative impact on volume pricing of the chips, PCB diversity (having two types of PCB will drive up cost as well) and driver diversity (though I think we can handle supporting two different WiFi drivers).
There will be two types of the PCB anyway: one with cell modem and one without. I'm just saying that the difference between those two can be made bigger.

In terms of drivers I don't think there's an issue, most likely the same driver can handle both variants of the chip, and even if not, then having one more kernel module in the default OS barely makes a difference.

Volume pricing of the chips could be an issue, as well as estimating how many of each kind are needed -- but the latter will already have to be done for the cell modem anyway, so if we tie them together, it makes no difference in that respect.

What was the plan for the first batch? 1000 without cell modem? 500 with and 500 without? What would be the price difference per wifi chip between ordering 500 and 1000? My guess would be: not much.
 
There will be two types of the PCB anyway: one with cell modem and one without. I'm just saying that the difference between those two can be made bigger.
Technically it'd all be the same PCB, it's just a matter of whether the cell modem is populated or not.And ED's said that the wifi module they're using is pin compatible, so still no need for different PCBs, just swap one for the other in the pick-and-place.
 
I've been following development for a long time, but I've never gotten around to posting until now! And it looks like I missed thr kryboard poll... Oops :)


Anyway, I wanted to say that I never understood the point of 5GHz WiFi until I moved to my current apartment, where the 2.4GHz spectrum is so saturated that it barely works much of the time. I would gladly pay a little more for 5GHz to make the WiFi usable for me, though I understand most people don't live in such a congested radio environment.


Anyway, thanks for all the hard work, and I'm very excited about the Pyra!
 
http://www.speedguide.net/faq/is-5ghz-wireless-better-than-24ghz-340

So, essentially 5Ghz can theoretically be faster and has more channels so is less likely to suffer from interference, but it has a shorter range: at the same power, the distance you get is smaller at higher frequencies, and also solid objects are bigger obstacles.

2.4GHz is more likely to suffer from interference (obviously this depends on how many 2.4GHz wifi/BT devices you and your neighbors are already using simultaneously), but it has a better range.

In my experience, range is the important thing. Speed will not matter when you're using wifi for internet; if you're using it for file transfers on your home network (e.g. streaming stuff from a NAS) then speed does matter, but you'll never beat the speed of something wired anyway.

Of course, if we pay extra for the 5GHz then we still have the 2.4GHz bands, so I guess you could get the best of both worlds.

In practice, I would probably be using the 2.4GHz bands most of the time though, because for me, range is nearly always a bigger problem than interference or speed. Not much point in having a fast and non-congested 5GHz band when you can't see the access point...

Should we pay 14€ extra for 5GHz? In my opinion, it's not really worth it. I'd rather get a bigger/faster eMMC or something for those 14€. You can still be future-proof by upgrading your wifi router to a dual-band one and making sure that all or most of your other devices (laptop, phone, smartTV, etc) use the 5GHz bands, which means the 2.4GHz bands get less congested and will work better again.

Though perhaps we could have the cheaper wifi chip in the devices without 3/4G, and the more expensive one in the devices that will be more expensive anyway? You could also offer all 4 combinations, but that is just going to complicate things. Most likely, people who are willing to pay for 3/4G will not mind paying also for the extra 5GHz band, while those with a tight budget will drop the 3/4G and they will certainly appreciate not paying for the extra 5GHz band. So if you're going to have two different main boards anyway, you can just as well make them differ in terms of the wifi chip too, right?
It has been said before, but I want to emphasise the primary reason for using the 5GHz band is the extra channel space, allowing for less interference in dense urban areas. I live in an apartment building where the 2.4GHz band is completely packed with multiple networks colliding on all channels, and sometimes interference lowers throughput so badly that streaming HD video over the air to my desktop becomes untenable (although fortunately, usually only for a short time). On portable devices like my old-ish smartphone, an old netbook of mine and the Pandora, even the interference from my desktop can completely cripple throughput to as low as double digit kbit/s, although I know that modern laptops don't have this issue. Ideally I would use cable for my desktop and the 5GHz band for portable devices.

The 2.4GHz band is more suitable for virtually every other circumstance, so if it has to be a binary choice, I think it might be helped by some demographics. Are members of the Pandora community city dwellers, or mostly based in more sparsely populated areas? I have a feeling on-the-go uses are less relevant; I suspect range is more important than available channel space in typical hotspot scenarios. Also, in hotspot scenarios throughput might be bottlenecked further up the chain anyway, although, who knows what the future will bring.

I guess it should also be noted that dual band USB wifi dongles are available and are probably very acceptable for home use where the 5GHz band is important, so it wouldn't be the end of the world for urban users.

All that said, if 5GHz was an option such as being available on the system with the cell network chip like you suggested, I think that would be great.
 
Back
Top