Our New Machine, Pandora

[Javacat]

I'd want to have the keyboard with me most of the time and although there would be times when I wouldn't be using it (for playing console games), I can see myself using it quite a lot. To play a game with it, the keyboard would have to be taken off, flipped around and then 'locked' into place on the back, and then when I've finished playing a game I'd have to take it back off again and and plug it back in to make the clamshell design. Constantly plugging and unplugging the keyboard would be a pain and increases the chances of it breaking, or the connection becoming weak and the keyboard falling off (and thus making you lose it).

I've found that onscreen keyboards are a pain to use in landscape oriented devices (eg, DS). The onscreen keyboard is quite good on my AXIM. Also, as the screen is widescreen and (assumably landscape), placing the keyboard onscreen so that it isn't wasting screen space and is also easy to use will be quite hard to do. You could have all of the characters in a row across the bottom of the screen, but this would be difficult and slow to use. Alternatively you could have the characters groupd together, like a normal keyboard and locate it in the lower center of the screen, but this will leave big gaps down either side and severely limit the amount of space above it.

A built in (hideable) keyboard adds very little to the size of a device and means that software developed for it doesn't need to take into account different hardware configurations.

 

[Eolair]

Guys, if you don't care, I'll come back lurking in these forums in a year or so, it will be funny to see how real all of this really was back in 2007

K, thanks bye. Hopefully you'll shut up for a while now then :rolleyes:

It wouldn't satisfy me and I also know a number of people that wouldn't be buying this if it was designed like that <_<

I'm sure Craigix has his own idea by now. I'm really looking forward to seeing it. Maybe he has a really slick solution

 

[fatblueduck]

Javacat,

I suppose I'll have to agree to disagree with you. If a keyboard is physically integrated into the device there will be more opportunity for breakage there on the main device and if there is any breakage the entire device will need to be replaced. I'd rather just replace the keyboard if something happened to it. I can't imagine mechanics for hiding a keyboard that wouldn't be more prone to breakage than a plug-in device. A sliding mechanism or a small flipping out type of thing requires some very small parts design. DaveC's design involves far less complexity in its engineering demands and in its appearance.

Another bonus to DaveC's design is that the keyboard would have more space for larger keys, making it much more usable. I believe that your preferred implementation would leave us with smaller keys that would be harder to use and more difficult (and possibly more expensive) for production.

There are examples of electronic plug-ins that survive constant use over time without losing effectiveness or grip. sd cards, cartridge games, pcmcia cards, etc.etc. The connection just needs to be deep and the connection needs to 'catch' upon full insertion. If you have to un-plug and plug, its not a big deal - it'll be like your cup of coffee in the morning before starting with your day. if it locks/snaps together it would remain durable and usable over time.

My perspective is that there are design issues with small physically embedded keyboards. They aren't elegant and they are always somewhat of a compromise for the user since they are necessary but unpleasant to use, due to their small size and cheap feel. This may be the reason that apple is not selling i-phones with a small embedded keyboard. Apple takes design very seriously and they are aware of the un-pleasant design constraints with such a device.

Sure, a final design may have been determined already by people who are making those decisions, but maybe not. Its an interesting discussion.

Just wanted to voice my opinion. I hope other people will chime in and tell me if I'm way off on this or if they agree with me.

Chris

 

[mison]

Question:

I thought it had been decided that the SoC would be an MMSP2+, so why are we talking about alternatives?

Afaik, Craigix has never said that he's decided to use a particular SoC. It's been deducted that he's considered and possibly tested and evaluated the MMSP2+. But as far as we know nothing has been decided. He has only said he would let us know more later on.


a while ago, ED said in the german board that it's going to be MMSP2+ or imx31, both being tested at the moment.
( reply to the 4th quote )

 

[Javacat]

Constantly plugging in and unplugging the keyboard is going to wear the connections down more than sliding a keyboard in or out. In addition, Dave2's design would be quite difficult to use whilst moving around as the keyboard is too big to use as a thumbpad and the entire device will also be top heavy, putting wait on the hinge on the keyboard.

The size of the keyboard will be dependent on how it is attached. If it's located under the screen then it will be a bit too small and fiddley, but if it's basically the entire base of the device then it will be much larger. Also, if the keyboard is part of the base it'll mean that it's more weighted, spreading the weight more evenly.

The iPhone is aimed at a market where people want devices more minimalistic. As I said in my previous post, I like the onscreen keyboard of my AXIM. It lends itself well to the orientation and size of the device, much like the keyboard on the iPhone.



I didn't want to get into this. I was just voicing my opinion about the built in keyboard in the offchance that Craig does take into account what people have said in this thread and so that he knows that there are people that want a built in keyboard.

 

[realyst]

DaveC, that design rules!

Javacat: I don't think the rapid plugin-plugout would be an issue. With something like 16 connectors to the controller for a full QWERTY keyboard on a desktop(less if the controller is integrated into the keyboard, but I wouldn't advise it as it would probably make it thicker), that's 8 connectors on the top and bottom of the actual interface. Make it wide enough and you've got more copper there then your average game cart which gets much much more wear and tear then the keyboard will likely get.

And again, that's 16 connectors for an inefficient matrix for a desktop keyboard, you could probably do with much less for the actual mini keyboard.

Let's not forget Palm devices used horrible -and- fragile interfaces on some of their keyboards and they still worked good enough.

 

[jadon]

Question:

I thought it had been decided that the SoC would be an MMSP2+, so why are we talking about alternatives?

Afaik, Craigix has never said that he's decided to use a particular SoC. It's been deducted that he's considered and possibly tested and evaluated the MMSP2+. But as far as we know nothing has been decided. He has only said he would let us know more later on.


a while ago, ED said in the german board that it's going to be MMSP2+ or imx31, both being tested at the moment.
( reply to the 4th quote )


Maybe this belongs over on Craigix disclosure thread, but it fits to respond to the comments on the processors above.

I work for TI and I've moved into a position where I can support the OMAP3 devices. I'm quite reluctant to mention this on a forum like this, but I'm sure it is a lot better than being "outed" later. I can tell you that supporting something like the GP2X is quite outside the norm for TI, but I just happen to be a lover of retro games.

I'd love to help whoever is working on the next platform add an OMAP3 device to the above list of devices under evaluation. The OMAP3 would significantly out-perform both of those other devices in all major metrics, including instructions per second, power consumption, triangles per second, fill rate, video size and frame rate, and more, with a really complete peripheral set.

 

[Javacat]

DaveC, that design rules!

Javacat: I don't think the rapid plugin-plugout would be an issue. With something like 16 connectors to the controller for a full QWERTY keyboard on a desktop(less if the controller is integrated into the keyboard, but I wouldn't advise it as it would probably make it thicker), that's 8 connectors on the top and bottom of the actual interface. Make it wide enough and you've got more copper there then your average game cart which gets much much more wear and tear then the keyboard will likely get.

And again, that's 16 connectors for an inefficient matrix for a desktop keyboard, you could probably do with much less for the actual mini keyboard.

Let's not forget Palm devices used horrible -and- fragile interfaces on some of their keyboards and they still worked good enough.

I was talking about the things that clip the keyboard to the console, not the circuit.

 

[icurafu]

Maybe this belongs over on Craigix disclosure thread, but it fits to respond to the comments on the processors above.

I work for TI and I've moved into a position where I can support the OMAP3 devices. I'm quite reluctant to mention this on a forum like this, but I'm sure it is a lot better than being "outed" later. I can tell you that supporting something like the GP2X is quite outside the norm for TI, but I just happen to be a lover of retro games.

I'd love to help whoever is working on the next platform add an OMAP3 device to the above list of devices under evaluation. The OMAP3 would significantly out-perform both of those other devices in all major metrics, including instructions per second, power consumption, triangles per second, fill rate, video size and frame rate, and more, with a really complete peripheral set.



I would love to see a cortex in Craigx's device, but what about cost? cortex A8 was brought up before, and it was determined that orders were expensive and could only be purchased in ridiculous batch amounts. Seemed off limits for anyone but Samsung and Nokia.

I'd quite a bit more for a device with that SoC.

BTW, what devices currently use the OMAP3, it looks very powerful.

 

[Eolair]

Question:

I thought it had been decided that the SoC would be an MMSP2+, so why are we talking about alternatives?

Afaik, Craigix has never said that he's decided to use a particular SoC. It's been deducted that he's considered and possibly tested and evaluated the MMSP2+. But as far as we know nothing has been decided. He has only said he would let us know more later on.


a while ago, ED said in the german board that it's going to be MMSP2+ or imx31, both being tested at the moment.
( reply to the 4th quote )


Maybe this belongs over on Craigix disclosure thread, but it fits to respond to the comments on the processors above.

I work for TI and I've moved into a position where I can support the OMAP3 devices. I'm quite reluctant to mention this on a forum like this, but I'm sure it is a lot better than being "outed" later. I can tell you that supporting something like the GP2X is quite outside the norm for TI, but I just happen to be a lover of retro games.

I'd love to help whoever is working on the next platform add an OMAP3 device to the above list of devices under evaluation. The OMAP3 would significantly out-perform both of those other devices in all major metrics, including instructions per second, power consumption, triangles per second, fill rate, video size and frame rate, and more, with a really complete peripheral set.

Interesting... your flagship the OMAP3430 would certainly be killer hardware in a device like this (beats anything hands down). I think a big problem might be the quantity and price point though. Currently i.MX31 can be bought in batches of 152 units for ~$25 per unit. The MMSP2+ probably has a similar minimum quantity and price point.

Would the OMAP3430 even come remotely close? You're better off talking privately to Craigix about this though. He's the guy who's running the show. Try send him a PM

 

[Orkie]

Guys, if you don't care, I'll come back lurking in these forums in a year or so

Please do so, you are being more annoying than DaveC.

 

[jadon]

Interesting... your flagship the OMAP3430 would certainly be killer hardware in a device like this (beats anything hands down). I think a big problem might be the quantity and price point though. Currently i.MX31 can be bought in batches of 152 units for ~$25 per unit. The MMSP2+ probably has a similar minimum quantity and price point.

Would the OMAP3430 even come remotely close? You're better off talking privately to Craigix about this though. He's the guy who's running the show. Try send him a PM



His mailbox was full.

I think the pricing would have to consider what the GP2X development community can do for the device. There's no way that a GP2X-class device will ship the kind of volume that the OMAP3430 would typically address, but if the effort generates software that the big guys can use, then I think that gives me a story inside the company.

 

[craigix]

Sorry about the mailbox being full, it happens all too often since the new system was announced. I'm going to delete some messages now.

PS chip prices:

MMSP2+ is around $12 (including TV out)
IMX.31 is around $30 + TV out chip which has to be bought separate and added to the PCB, I guess the max cost of both would not be above $40.

I'm interested in chatting about the OMAP3430 although we are pretty far along the road now so it would have to be very soon

 

[Tobriand]

If the OMAP3430 could become a contender, I would be a very very happpy bunny indeed. Not going to gget my hopes sky high yet, but well, it'll be very interesting to see what happens....

 

[Rivroner]

I like very much the 532 mhz of the i.mx31

http://www.logicpd.com/eps/som/freescale/i.MX31/


Omap3430 speed is? I can´t find it.

I think is better a single core at 532 that the double 360-300 mmsp2+ core.


Thank you very much for the info Craigix

 

[atomicthumbs]

I like very much the 532 mhz of the i.mx31

http://www.logicpd.com/eps/som/freescale/i.MX31/


Omap3430 speed is? I can´t find it.

I think is better a single core at 532 that the double 360-300 mmsp2+ core.


Thank you very much for the info Craigix



Methinks that dual core would be better. What're the specs for the OMAP thingy's 3D processor?

 

[Eolair]

The OMAP3430 is a monster. ARM11 Cortex-A8D at 550MHz, PowerVR SGX 3D graphics accelerator , video and audio processors... I'm pretty sure it would be awesome but do we need that kind of power? It's so far above the i.MX31 and the MMSP2+ in performance that it's crazy.

ARM11 Cortex-A8D performance:
"The superscalar processor core is said to generate almost double the performance per MHz compared to previous generation versions, featuring deeper pipelines, higher frequency and dedicated Level 2 cache"

PowerVR SGX Performance:
Effective fillrate up to 4000Mpix/sec (PSP: 664Mpix/sec)
Polygon througput up to 100M polygons/sec (PSP 33M polygons/sec)
http://www.imgtec.com/powervr/products/gra...ndex.asp?Page=1

Edit: Link added

 

[Rivroner]

The OMAP3430 is a monster. ARM11 Cortex-A8D at 550MHz, PowerVR SGX 3D graphics accelerator , video and audio processors... I'm pretty sure it would be awesome but do we need that kind of power? It's so far above the i.MX31 and the MMSP2+ in performance that it's crazy.

PowerVR SGX Performance:
Effective fillrate up to 4000Mpix/sec (PSP: 664Mpix/sec)
Polygon througput up to 100M polygons/sec (PSP 33M polygons/sec)
http://www.imgtec.com/powervr/products/gra...ndex.asp?Page=1

Edit: Link added



We need it

i.mx31 performance is like equal isn´t it? i.mx31 has like 532 mhz.

Interesting link:

http://www.analog.com/en/press/0,2890,3%25...F121923,00.html

PRICES of BlackFin BF533 , really chip and better performance than i.mx31 :

http://www.analog.com/en/prod/0,2877,ADSP%252DBF533,00.html

 

[fusion_power]

Sorry about the mailbox being full, it happens all too often since the new system was announced. I'm going to delete some messages now.

PS chip prices:

MMSP2+ is around $12 (including TV out)
IMX.31 is around $30 + TV out chip which has to be bought separate and added to the PCB, I guess the max cost of both would not be above $40.

I'm interested in chatting about the OMAP3430 although we are pretty far along the road now so it would have to be very soon

Hm, for me, the first question is: which chipset give me the highest battery/Akku times? Maybe I'm alone with this opinion but for me, A Handheld should be designed not only by CPU Power but also designed to have a long battery life. One of the reasons, why Nintendos Handhelds so successful is, you can play a very long time with one battery charge. So for me, CPU Power is not the most important point and CPU Clockrates like 500 or 600MHz scares me when I look and compare at the GP2X and his 200MHz and his poor 5-6 Hours of Playtime...

Oh, somebody here sayed, this IMX.31 costs 23$, was this price without TV Chip?
And I really would like to know the building costs of the new device - if this is available at its time. 30$ for an entire Chipset isn't much, maybe 40$ for CPU 20$ for RAM, 20% for *Idontknow* and 10$ Case... 150-200$ (or maybe 150-200€ )pure manufacturing costs are not much I think so I hope, the end-price for users will not be that high

 

[Eolair]

We need it

i.mx31 performance is like equal isn´t it? i.mx31 has like 532 mhz.

Haha...no it's not even remotely close. The i.MX31 is using the PowerVR MBX Lite which provides fill rates of 135Mpix/sec and throughput of 1.7M polygons/sec. Compare that to the numbers I mentioned above :lol:
http://www.imgtec.com/PowerVR/products/Gra...ndex.asp?Page=4

Edit:Linky again