Next pandora (competition model)

[gruso]

DisplayPort should be considered for digital AV out. It's royalty free, as compact as HDMI (and has a mini version), and can be converted to DVI or HDMI with a very cheap adapter.


I do tend to agree with Exophase though; a HDMI port is kind of expected these days.

 

[Prometheus]

^ That's what Apple uses. I knew it had a name, but I couldn't dare remember it.

 

[bismuthdrummer]

Perchance to dream!


Well, if Santa were to ever get a Pandora 2, I would suppose that he might want the screen to be 1024x600 resolution, seeing as how all the elves' netbooks are that resolution size, therefore websites and portable apps should support it as well. True that it would increase the pixel density, but edge up the screen's physical size as posited and this should mitigate.


I like the idea of being conservative with the form factor and controls. Barring any _simple_ improvements on the ergonomy that could be done, the existing case design is still very marketable with perhaps some gloss and/or a slimmer case. The Nintendo DS obviously did very well, as that case design has pretty much carried all the way to the 3DS. Shoot for a similar thinning as from the DS Heavy (original) to the DS Lite if at all possible.


Backlit keys, peripheral sensors, HDMI, all win if they're relatively easily accomplished. Of course upgrade the internals.


Does anyone have any concrete proposals to significantly lower the price? I am guessing I'll get the "Why do people want a castrated device?" answers, but I am focused primarily on keeping its general robustness while maybe axing certain expensive but not mission critical features. Please don't stone me for saying this, but Bluetooth comes to mind... would cutting Bluetooth not save quite a bit on the cost? Any other features that could be sacrificed?

 

[Prometheus]

^ I'm pretty sure that cutting Bluetooth wouldn't make a blind bit of difference, as it's on the same chip as the WiFi...


I seem to (vaguely) recall that the Pandora originally didn't have Bluetooth at all, but it didn't add to the cost when it turned out to be part of the same module.

 

[bismuthdrummer]

Did not know that. Is the wifi spendy, as in are there cheaper alternatives without Bluetooth?


I accept that the Bluetooth idea could very well be a dud, I just wanted to seed some brainstorming on the saving money side of this topic, if there is any saving to be done in the first place.

 

[Prometheus]

I honestly don't know. I just know that it probably doesn't make much difference in this particular case. Sorry about that.

 

[mcobit]

Getting a different, full working wifichip with good driversupport in there should be a priority.

 

[EvilDragon]

I'm interested to see how well the WiFi Chip performs with our boards from Germany.


The chip works in other devices, so I always wondered what the problem was here.


There are no real cheaper WiFi chips without Bluetooth out there, and as long as we don't have UMTS built-in, bluetooth is the best way to tether to your phone (uses WAY less power than WiFi).

 

[EvilDragon]

I know I saw a quote for a 5" 800x480 LCD for $15 from a Hong Kong supplier (that was for 5000). Much better response time than Pandora's, but used much more power. This was from a Taiwanese manufacturer, as far as I can remember Pandora's is as well. I'd still go for this without question vs Pandora's when there's such a big battery.

Well, without seeing this LCD in masses, it's hard to tell whether it is good or not.


I've very rarely seen dead pixels on Pandora LCDs (only know of about 5 or 6 LCDs so far with one dead pixel), it might be the cheap one has far more dead pixels on the LCDs.


GC has some connections to LCD manufacturers, so I'll poke them as well to get samples and get the best LCD for a reasonable price that is possible.

If TI gives you better accessibility to better chips or it has good pricing then great, but I wouldn't go with TI just because they're TI, or just because Pandora 1 uses it.

Sure, I'm not fixed on TI.


But it might make upgrading easier, as some things might be compatible.


NEON is pretty good and can make speed enhancements.


AFAIR, notaz mentioned he played with a recent Tegra a bit, but the missing NEON made it about as fast as our old Cortex A8


Most important thing for me is to stay as compatible as possible.


Ideally, most PNDs should run out of the box, but at least the OS should work on the new Pandoras without any issues.


As we're using dynamic libraries, this shouldn't be a big problem though.


Add drivers and update the kernel (sounds simpler than it is ), and we're basically set.

 

[Foxgod]

I know I saw a quote for a 5" 800x480 LCD for $15 from a Hong Kong supplier (that was for 5000). Much better response time than Pandora's, but used much more power. This was from a Taiwanese manufacturer, as far as I can remember Pandora's is as well. I'd still go for this without question vs Pandora's when there's such a big battery.

Well, without seeing this LCD in masses, it's hard to tell whether it is good or not.


I've very rarely seen dead pixels on Pandora LCDs (only know of about 5 or 6 LCDs so far with one dead pixel), it might be the cheap one has far more dead pixels on the LCDs.


GC has some connections to LCD manufacturers, so I'll poke them as well to get samples and get the best LCD for a reasonable price that is possible.

If TI gives you better accessibility to better chips or it has good pricing then great, but I wouldn't go with TI just because they're TI, or just because Pandora 1 uses it.

Sure, I'm not fixed on TI.


But it might make upgrading easier, as some things might be compatible.


NEON is pretty good and can make speed enhancements.


AFAIR, notaz mentioned he played with a recent Tegra a bit, but the missing NEON made it about as fast as our old Cortex A8


Most important thing for me is to stay as compatible as possible.


Ideally, most PNDs should run out of the box, but at least the OS should work on the new Pandoras without any issues.


As we're using dynamic libraries, this shouldn't be a big problem though.


Add drivers and update the kernel (sounds simpler than it is ), and we're basically set.

True, it would be such a waste if most, if not all of the software would have to be redone for the panda 2.

 

[monty]

A lot of things I've been reading here would double to triple the production costs

Here's one idea that shouldn't. The Pandora has a very responsive capacitive resistive* screen, sometimes just as responsive as a Smartphone and therefore well able to be navigated with a finger.


The problem is that, what with everything getting tucked into corners on the desktop there are certain tasks that can't be done easily without a stylus or especially sharpened thumbnail. Minimising for one.


Solution: bring the screen up so its flush with rest of the lid. What do you think?


* Meant resistive, of course, excuse me.

 

[slaeshjag]

The precision will still be rubbish, making it mostly worthless in the desktop enviroment. In minimenu? Would probably work.

 

[monty]

The precision will still be rubbish, making it mostly worthless in the desktop enviroment. In minimenu? Would probably work.

Thin fingers perhaps, but I manage well enough with even text selection in the desktop.


Another reason would be that it would save ~2mm of space and look a bit nicer.

 

[Binky]

The Pandora has a very responsive capacitive screen,

I'd be fine with bringing the screen forward, but I'm pretty sure the P1's screen is resistive, which I prefer!


<rant>If you were comparing styli, 1cm diameter and squishy, with a little visor which needs continuously trimming, would be a would be a class failure. So why create capacitive touchscreens that depend on them?!</rant>

 

[Sugar_Kane]

I would like there to be a medium-sized, silken, vibrating pouch attachment

 

[Emnasut]

I would like an aluminium case.


Also UMTS would be really nice; I like my gadgets sturdy and want my phone to be primarily a phone, so I'm using a Sonim XP3 Quest (http://www.sonimtech.com) instead of one of those smartphone thingys. Now using a MIFI works great with the Pandora, but built-in UMTS would be more comfortable. Also my provider gave me a second card for free with the same data plan, so it doesn't add any costs (also wouldn't if I'd use a smartphone).

 

[Exophase]

Well, without seeing this LCD in masses, it's hard to tell whether it is good or not.
I've very rarely seen dead pixels on Pandora LCDs (only know of about 5 or 6 LCDs so far with one dead pixel), it might be the cheap one has far more dead pixels on the LCDs.


GC has some connections to LCD manufacturers, so I'll poke them as well to get samples and get the best LCD for a reasonable price that is possible.

Right, I hadn't seen a sample so I don't know the quality. But things like this are definitely worth looking into. Just from what I was poking around, the prices seem to go (from lowest to highest):


Chinese, Taiwanese, Korean, Japanese, American


That applies for pretty much all ICs too, like memory. But of course if you want PoP your options are pretty limited, especially w/LPDDR2 which is probably what you'd push for in Pandora 2. If you can somehow arrange the space for not needing PoP DDR3 (possibly DDR3L if the SoC supports it) is probably cheaper, although higher power of course.

Sure, I'm not fixed on TI.
But it might make upgrading easier, as some things might be compatible.


NEON is pretty good and can make speed enhancements.


AFAIR, notaz mentioned he played with a recent Tegra a bit, but the missing NEON made it about as fast as our old Cortex A8


Most important thing for me is to stay as compatible as possible.


Ideally, most PNDs should run out of the box, but at least the OS should work on the new Pandoras without any issues.


As we're using dynamic libraries, this shouldn't be a big problem though.


Add drivers and update the kernel (sounds simpler than it is ), and we're basically set.

Tegra 2 is the only Cortex-A9 SoC without NEON (every Cortex-A8 one has it too). Even Tegra 3 will have it. For compatibility purposes going with NEON is a must, but it's not used enough to really say that lacking it makes things a dual core 1+GHz Cortex-A9 as fast as a 600MHz Cortex-A8, not across the board. Other than NEON there won't be big compatibility issues outside of the OS.. I really doubt anything significant (if anything at all) is really going to be using that DSP.


I really have no idea what kind of availability there'll be but if you get the chance you should sample SoCs from Samsung, Freescale, maybe Qualcomm.. Shouldn't have a problem running all PNDs out of the box on those. Angstrom Linux is a given for pretty much anything, with some porting effort of course.

 

[Mr_Loon]

Here are my thoughts, in no particular order of importance, on what I'd like to see for Pandora 2 :


1) Improved hinge design allowing multiple locking positions (available off the shelf)


2) Built in USB hub allowing connection of USB 1.1 devices


3) A more ergonomic case design, specifically with greater rounding on the corners below the FN and comma keys, thus allowing for more comfortable extended sessions with Pandora. (but not in the way Sugar Kane seemed to be suggesting )


4) HDMI out, it is appreciated that a license fee would have to be paid but the cost per unit for this would be far less than the cost of a display port to HDMI adapter. If present trends continue come 2015 or whenever Pandora 2 sees the light of day HDMI will probably be pretty much the standard video connector.


5) Larger Higher resolution screen


6) Shoulder buttons used for left & right mouse clicks out of the box


7) Nubs that equal the quality of the 3Ds circle pad


8) ABXY buttons that equal the quality of an Xbox 360 pad


And finally, drum roll please, my 'big' idea :


The ABXY buttons should have nothing printed upon them but instead should have a grid of tiny LEDs embedded into each button, software could then tell the buttons what to display, for example when emulating the PS2 (we can but dream) the buttons could display the Square, Triangle, Cross & Circle of a Sony controller etc. It is appreciated that this would be no easy feat to pull off but then the Pandora project was never about putting ease ahead of excellence and innovation.


This is an idea I've had kicking about for 20 years or so in relation to how to improve the PC keyboard, it would be great to see this implemented on the Pandora.

 

[Emnasut]

The ABXY buttons should have nothing printed upon them but instead should have a grid of tiny LEDs embedded into each button, software could then tell the buttons what to display, for example when emulating the PS2 (we can but dream) the buttons could display the Square, Triangle, Cross & Circle of a Sony controller etc. It is appreciated that this would be no easy feat to pull off but then the Pandora project was never about putting ease ahead of excellence and innovation.


This is an idea I've had kicking about for 20 years or so in relation to how to improve the PC keyboard, it would be great to see this implemented on the Pandora.

Like this?


http://www.artlebedev.com/everything/optimus/

 

[WizardStan]

1) Improved hinge design allowing multiple locking positions (available off the shelf)

You'll note how those all say "DS Lite" on them. Nintendo isn't in the habit of licensing their property for use in machines that don't belong to them.