Your Holy gaming Grail for Pandora 2..?


this linux 3.4 kernel is the solution to all it seems... native ability to dual boot both os's and support for android's power management...


only problem I see is that it will need a new version of android to support that kernel... well the future looks bright for both sides of this fence


EDIT: your average AOSP android build is 100-200mb also, so even a 512mb nand could potentially house both, but if it's cost effective I'd like 1GB nand, storing userdata for both OSs without putting it entirely on SD could get a little cramped
 
Last edited by a moderator:
10-20% of people returning the console because they can't use it.
Well now I think you're just insulting our intelligence.

Well, none of my customers ever returned a Pandora (except for repairs, but then he wanted to get it back as soon as possible), so I can't comment much on that.

In its stead, here's a thought: the linux distro we're running on Pandora will, unless Craig & co. do something really stupid like abandoning the ARM architecture, run on a Pandora 2 with almost no porting work required. Maybe some drivers, but you'll have that no matter what.

True. We're using OpenEmbedded, which is a system for all kinds of embedded devices. It runs on the old Zaurus, on some car information systems, on the BeagleBoard, the Pandaboard, etc.


We can continue to use our current OS without any issues.


Most of the work needed to be done are the drivers for the hardware. Once the kernel is running (and you need the same kernel drivers for the Android port as well), Linux should be working as well.
 
I struggle to see emulation development continuing for Linux, with other devices moving to Android (yes your Yinlips etc) and other linux devices dying off (Caanoo), the Pandora would stand alone if it were to just have Linux. Why would a developer develop a new emulator for it, they wouldn't, you would be waiting for New emulators developed for Android to be ported to Linux - not a great situation.

Umm... I have yet to see an emulator that derived from Android, to be honest.


I'm no Android user myself, but when I checked the internet, I couldn't find any.


Most emulators are ports from PC versions (mostly from the Linux versions), and I don't see any reasons why that should change anytime soon.


There are way more emulator gamers on the PC than there are on Android...


And any emulator ported from PC to Android can also be ported from PC to Pandora.


However, if there's an existing Linux port of the emulator (which is the case for most of the existing ones), a Linux port will be easier to make than an Android port.


So the future of emulation will probably still be a PC, and not a Pandora or any Android-Device.
 
Last edited by a moderator:
Im sure the future of emulation will remain with the PC for some time to come, don't disagree there. My opinion was with emulation for mobile devices though (Android vs Linux specifically).


On a different note and don't take this the wrong way, I'm not having a go at anyone, just would like to know why this is. Anyway, why is it ED that we are getting mixed(conflicting) messages between yourself and Craig regarding how easy it would be to have Linux running on future Pandora's. You and others in the community say it's simple yet Craig says it would be difficult. I'm confused.
 
Last edited by a moderator:
Android is certainly an option; we're carrying the existing load on a few peoples volunteer time; if we can get more hands at the rudder, maybe we can get some more stable Android work.


As I've always said, our biggest software problem has simply been lack of hands and given sufficient hands, coordination. You can only squeeze so much juice from a few oranges, as I often say :)


That said, every unit ever made has been sold, and every unit ever promised to be made ahs presold, so the unit has never had a problem selling; the software has never held it back (yet.) Its not perfect, but it does the job not half bad. But we aim for the sky, so if we can do more, and I mean we as a community, then we all will, right?


So yeah, Android, awesome. Whose going to work on it? ;)


(Hey, I ask that every month, so far, no takers ;) Maybe we shoudl get serious and really try to draft up a half dozen or dozen people to start cranking on it.. if we really drum hard, maybe we _can_ get a few people together and working on it. cpasjuste did a qwuick hack and got 'something' going back with Froyo I think; there was that fellow doing a port for his thesiss, got somethign going. So a single person can get 'somethign' going, but not usably; getting the network up, accelerated graphics, that sort of thing.. that takes work. But it can be done!


jeff
 
L2/R2 buttons you could just click it on when playing a game that requires them, and after just take them off!
 
On a different note and don't take this the wrong way, I'm not having a go at anyone, just would like to know why this is. Anyway, why is it ED that we are getting mixed(conflicting) messages between yourself and Craig regarding how easy it would be to have Linux running on future Pandora's. You and others in the community say it's simple yet Craig says it would be difficult. I'm confused.

It will never be simple, but it's not as hard as it was with Pandora 1.


The OS can be reused, kernel drivers are mostly the same for Android and Linux (we even might have the same kernel then), so there shouldn't be a big deal.


BTW: TI just told me that there are two products the have: The OMAP and the DaVinci.


We're currently using the OMAP. The DM3730 is a DaVinci.


The DaVinci is basically the OMAP series for the community. 99,9% of companies using the DaVinci are running Linux, whereas the OMAP is more aimed at the Smartphone Android market.


That's why you normally have a MOQ of 50,000 when you want to get OMAPs whereas you can order any number of DaVinci.


The DaVinci has official Linux support.
 
The DaVinci has official Linux support.
And when (not if) both Android and Linux are running on whatever future Pandora 2 might come out, will you make an official endorsement of both? Advertise it as being able to run either?
 
The DaVinci has official Linux support.
And when (not if) both Android and Linux are running on whatever future Pandora 2 might come out, will you make an official endorsement of both? Advertise it as being able to run either?

My personal primary goal will be Linux, of course, but it would be nice having Android as well.
 
My personal primary goal will be Linux, of course, but it would be nice having Android as well.
Obviously, but it doesn't answer the question. :p


When Android is running suitably on the Pandora or Pandora 2 or whatever, will you say "this is the Android we endorse" as part of the marketing for the device, and whenever anyone in the media says they want to do an article will you mention that it can run Android in addition to Linux?


And vice versa for Linux, assuming Craig is correct on the difficulty of getting Linux and the ease with which Android will run?
 
My personal primary goal will be Linux, of course, but it would be nice having Android as well.
Obviously, but it doesn't answer the question. :p


When Android is running suitably on the Pandora or Pandora 2 or whatever, will you say "this is the Android we endorse" as part of the marketing for the device, and whenever anyone in the media says they want to do an article will you mention that it can run Android in addition to Linux?


And vice versa for Linux, assuming Craig is correct on the difficulty of getting Linux and the ease with which Android will run?

Of course. If multiple OSes run mostly flawlessly, I'd see no reason to NOT let anyone know they can do that.


I also want to add Debian, Slackware, ArchLinux, etc. onto the official Pandora homepage as easy-to-use images :)
 
ED, can you comment on if you envision Pandora 2 coming pre-loaded with Android and Linux with dual boot capability? Possibly with a first-time-boot options informing the user and letting him/her choose a default?
 
If I'm still around doing firmware, I'll be mostly focusing my efforts on Linux.


Personally I don't see much ground in Craig's claims of Linux lacking behind in device support, it's the same kernel/drivers after all. Maybe he has the UI in mind or something. I'd say it's a lot harder to get something like keypad and nubs (pandora's distinctive features) to be useful on Android than on a Linux distro. On Linux they work out of the box and all programs expect a mouse and a keyboard, on Android everyone thinks you can only do multitouch and will just not care, so you'll have them just wasting space, plus screen overlay controls that you don't really need.


I also find Android to be painful to develop for, I had to do some of it at work. There is basically very stripped down rootfs, almost no tools there, not even a frigging gdb, have to f*ck around with gdbserver+adb or use GUI things like Eclipse that I don't like at all. Native code must be in PIC shared libraries, meaning unavoidable unneeded performance hit, plus weird address space layout hampering use of dynarecs, plus a forced JVM sitting in it. No fun there at all.
 
Last edited by a moderator:
edit; I think my post is going to make peopel mad, so just going to cut it out :p
 
Last edited by a moderator:
I'd say it's a lot harder to get something like keypad and nubs (pandora's distinctive features) to be useful on Android than on a Linux distro. On Linux they work out of the box and all programs expect a mouse and a keyboard, on Android everyone thinks you can only do multitouch and will just not care, so you'll have them just wasting space, plus screen overlay controls that you don't really need.
My thoughts exactly. Unless the Pandora 2 will provide a multi-touch screen and form factor more similar to a smartphone1, then I don't expect Android to bring the insane increase in usability that some people are advocating. I.e. interaction wise, the Pandora is almost a polar opposite of smartphones.


(1) Which would be inferior to a powerful smartphone combined with an iControlpad in most aspects.
 
I'd say it's a lot harder to get something like keypad and nubs (pandora's distinctive features) to be useful on Android than on a Linux distro. On Linux they work out of the box and all programs expect a mouse and a keyboard, on Android everyone thinks you can only do multitouch and will just not care, so you'll have them just wasting space, plus screen overlay controls that you don't really need.
My thoughts exactly. Unless the Pandora 2 will provide a multi-touch screen and form factor more similar to a smartphone1, then I don't expect Android to bring the insane increase in usability that some people are advocating. I.e. interaction wise, the Pandora is almost a polar opposite of smartphones.


(1) Which would be inferior to a powerful smartphone combined with an iControlpad in most aspects.
The Nubs are something that I can see that could be hard to support for every app on Android as well.. will they have to be mapped to key presses like you have to do with the ICP? that would render the Nubs as kind of useless.


sure maybe some emulators could come out with Analog support... but what about the rest of the apps.
 
Last edited by a moderator:
The Nubs are something that I can see that could be hard to support for every app on Android as well.. will they have to be mapped to key presses like you have to do with the ICP? that would render the Nubs as kind of useless.


sure maybe some emulators could come out with Analog support... but what about the rest of the apps.
Yes. Craig mentioned that something akin to Blutrol could be utilized to map the nubs and buttons to onscreen controls, but as far as I understand it such mappings indeed have to be made per application.
 
Back
Top