Porting Various Operating Systems To Pandora?

[randomhack]

XScale is based on ARMv5TE ISA according to wikipedia except for floating point instructions.

I know that :/


ohk yeah i expected that but i got confused by your last post.

 

[chad78]

Maemo & Android - There is certainly some decent software there, but do you really need the whole OS? Maemo is (and Android likely will be) compiled for ARM, so a compatibility layer would be much more desirable. Do you really want to have to reboot into a different OS every time you want to use a different app? In any case, I'm sure somebody will at least try both of them and I wouldn't be shocked if they succeed.

Well, if "a compatibility layer" will allow me to run all the Maemo (or Android) applications at full speed - then that would be perfect.

QUOTE
Ubuntu & FreeDOS - You run into a problem with program compatibility here. Even if you port the OS, none of the existing software is compiled to run on ARM processors. Not really worth the trouble.


Well, since you've found out Ubuntu has been compiled for ARM, (and running on a N810), do you think it's still a bad idea?

QUOTE
PSP Custom Firmware - The PSP runs on different (MIPS) hardware. You'd have to actually emulate it (not going to happen) to run any PSP software natively.


That sucks. I guess I'll have to hang on to my PSP in order to play PSP games. So much for one gizmo to rule them all and in the darkness bind them.

 

[Vorporeal]

QUOTE
Ubuntu & FreeDOS - You run into a problem with program compatibility here. Even if you port the OS, none of the existing software is compiled to run on ARM processors. Not really worth the trouble.

Well, since you've found out Ubuntu has been compiled for ARM, (and running on a N810), do you think it's still a bad idea?

There is a version of Ubuntu called Ubuntu Mobile, which looks very nice and is an enticing piece of software to try to get running on the Pandora. The only issue is that the current version is written for x86 devices, and it has a high minimum RAM requirement (256MB/512MB recommended). Both of these issues probably could be solved, but it would require finding someone interested enough to invest a sizeable chunk of time to doing this.

 

[chad78]

There is a version of Ubuntu called Ubuntu Mobile, which looks very nice and is an enticing piece of software to try to get running on the Pandora. The only issue is that the current version is written for x86 devices, and it has a high minimum RAM requirement (256MB/512MB recommended). Both of these issues probably could be solved, but it would require finding someone interested enough to invest a sizeable chunk of time to doing this.

Nokia is paying for Ubuntu to be completely ported to ARM - specifically for their Internet Tablet devices (the N-Series N800, N810, N810 WiMax Edition)

These Internet Tablets are very similar to the Pandora - both are ARM processors, (although the Pandora is more powerful), both have the same resolution touchscreens, both have wifi, (with the exception of the N800) both have thumb keyboards, etc. It seems to me that the hard work will be done. Just some minor tweaking, if anything, to get it to work on the Pandora. My question now is will Chip's proposed "compatiability layer" give me close to full powered performance on close to all the programs that the full version of Ubuntu could run.

 

[Sphinxter]

The native compiling is interesting.
http://www.linuxdevices.com/news/NS2097004728.html

 

[ldesnogu]

So you're saying XScale can not be considered an ARM platform? I meant by ARM CPUs, not by.. ARM themselves. ^^

I was simply kidding

 

[Codemonkey]

I would like to see an optimized gentoo with a small window manager. That should work very good and you can switch in the terminal mode where you can do everything with the integrated keyboard.
The only thing is the resolution, but I think its okay.

 

[Svartalf]

I would like to see an optimized gentoo with a small window manager. That should work very good and you can switch in the terminal mode where you can do everything with the integrated keyboard.
The only thing is the resolution, but I think its okay.

Considering that Gentoo is largely source installed, as opposed to pre-built package installed, I don't think that this would be practical for several different reasons.

While Gentoo provides the "ultimate" in performance out of a machine, it's not something I'd inflict (and that would be the word, unfortunately...) on an unsuspecting and ignorant end-user. Most of the people are gamers- they're not going to dink with the innards of things like portage to get their machine up and running and to put new stuff on their machine. Moreover, I don't think they're going to have access all the time to download all the sources through the ebuild process. Now, having said this, the word is that you could put it on a SDHC card and go to town on your own. To me, that sounds tempting, in and of itself.

 

[Svartalf]

I'm getting concerned. Let's say I want to do a NetBSD port, or perhaps even something more exotic like L4. I have learned from this discussion not everything is open source or has open documentation available. For instance, 3D accel.

What else would be crippled or broken if I were to abandon Linux, starting from the boot loader and up?

2D/3D accel for starters. DSP access for sound and similar processing. Possibly the WiFi- but I think MWeston's probably going to pick one that works nicely without issues in that space.

Blame the vendors, not Linux- it's not our fault that they're being anal retentive (though that may change over time- many of the OEMs (the people that're buying these OMAP and other devices...) have taken to insisting on more and more open information or outright open source support in addition to the tech data for the devices. I know, I've seen it myself (heh...gotten contracts as a result of it... > ) for a while now. It's part of the reason that AMD and Intel have been so forthcoming with programming info of late. It's also happening with vendors like Broadcom and ImgTec, but to a lesser extent right at the moment- they're not caving yet...

 

[Karel Jansens]

Considering that Gentoo is largely source installed, as opposed to pre-built package installed, I don't think that this would be practical for several different reasons.

Since the Pandoras share the same hardware, what would be against pre-compiling an optimized Gentoo distro, coupled with a precompiled package repository? If done properly, it would make for one kick-ass fast Linux distro, taylored to the Pandora. No?

 

[proflogic]

Since the Pandoras share the same hardware, what would be against pre-compiling an optimized Gentoo distro, coupled with a precompiled package repository? If done properly, it would make for one kick-ass fast Linux distro, taylored to the Pandora. No?

Already happening?

I mean... I would assume whatever is being distributed on the internal NAND is going to be optimized for the Pandora . And whoever maintains the third-party packages for the Pandora is probably going to make sure they are optimized for the hardware as well.

The environment just might not support USE flags and whatever else Gentoo does to make it easier to make a system of custom binaries. But you're not even using that if you want pre-compiled packages... Besides, who wants to compile all their software on the Pandora itself? Unless your PC is quite old, a cross development toolchain is probably going to give you faster results.

 

[Karel Jansens]

Already happening?

I mean... I would assume whatever is being distributed on the internal NAND is going to be optimized for the Pandora . And whoever maintains the third-party packages for the Pandora is probably going to make sure they are optimized for the hardware as well.


Well... yeah. Sorry 'bout that.

QUOTE
The environment just might not support USE flags and whatever else Gentoo does to make it easier to make a system of custom binaries. But you're not even using that if you want pre-compiled packages...

Besides, who wants to compile all their software on the Pandora itself? Unless your PC is quite old, a cross development toolchain is probably going to give you faster results.



People who use the Pandora as their only computer?

...

Don't laugh. My Psion 3a was my only computer for the better part of two years. Admittedly, that was in the Second Aera of the World, when 4 MB was considered a really big memory card and connectivity meant a dialup modem to a BBS, with faxing capability as a nice extra, but still...

 

[tmcnair]

QUOTE

No. It was because some people take things way too seriously. It was because some people don't believe in god, but have a psychological need to worship something, so they worship a licensing agreement. It was because some people are ineffective and impotent in real life, so they compensate by being stupendous assholes on the internet.


If I appear to be less than respectful of the GPL and it's adherents, it is because that disrespect was hard earned.



Licenses are licenses - they need to be honoured. But GPH, and now OpenPandora, would never dream of violating a proprietary license or NDA would they? That is because they'd get their **s sued in a heartbeat. The FSF crowd is an enforcement mechanism for GPL. Either way the results are the same - violate a license, pay a price. Violate a proprietary license and get sued, violate the GPL and have your reputation besmirched. I'm not sure why one would make you any less respectful than the other.

 

[Chip]

Licenses are licenses - they need to be honoured. But GPH, and now OpenPandora, would never dream of violating a proprietary license or NDA would they? That is because they'd get their **s sued in a heartbeat. The FSF crowd is an enforcement mechanism for GPL. Either way the results are the same - violate a license, pay a price. Violate a proprietary license and get sued, violate the GPL and have your reputation besmirched. I'm not sure why one would make you any less respectful than the other.

First of all, this thread is two months old.

Second, nobody is going to respect a community that is so blatantly disrespectful of others - and a lot of the GPL / FSF people are exactly that.

The whole point of that story was to show that there is no reason to get all excited about GPL stuff when:

1. The developers involved are well aware of the rules
2. They are clearly doing everything they can to obey them
3. It will obviously be resolved in the near future

So any of this "ZOMG MUST OBAY TEH GPL!!!!!111!!ONE!!" is just self righteous douchebaggery.

 

[tmcnair]

Licenses are licenses - they need to be honoured. But GPH, and now OpenPandora, would never dream of violating a proprietary license or NDA would they? That is because they'd get their **s sued in a heartbeat. The FSF crowd is an enforcement mechanism for GPL. Either way the results are the same - violate a license, pay a price. Violate a proprietary license and get sued, violate the GPL and have your reputation besmirched. I'm not sure why one would make you any less respectful than the other.

First of all, this thread is two months old.

Second, nobody is going to respect a community that is so blatantly disrespectful of others - and a lot of the GPL / FSF people are exactly that

The whole point of that story was to show that there is no reason to get all excited about GPL stuff when:

1. The developers involved are well aware of the rules
2. They are clearly doing everything they can to obey them
3. It will obviously be resolved in the near future

So any of this "ZOMG MUST OBAY TEH GPL!!!!!111!!ONE!!" is just self righteous douchebaggery.


Yeah I must admit I don't have much tolerance for that kind of incivility, for example getting called a douchebag.

The typical MO of someone trying to protect GPL'd code is first to ask nicely, then if that is ignored start making noise, and once in a blue moon the threat of civil action. As with every group there are a few bad apples who are just jerks. I've been lurking in these forums longer than my ID suggests - I remember the GPH/GPL fiasco. I don't remember it being all that incendiary. Yeah accusations were made - but they were true weren't they? GPH should have responded with source code in the "ask nicely" stage - that would be the honorable (and legal) thing to do.

Full disclosure: I have code in the linux kernel - so I am, in your eyes, a GPL zealot I suppose. My code doesn't get used in any of these consumer devices so I'm not directly harmed when someone uses the code and doesn't pay the "royalty" (i.e. release their modified code), but I can certainly sympathize with someone getting upset about their code being stolen.

 

[AireTamStorm]

I'm not sure, but it looks like I'm in the minority regarding the GPU.

Isn't it highly disadvantageous to use a closed-source module? The SGX running on OMAP isn't even "proven" really. Basically, this means all GPU-related bugs are completely outside community control. If the Pandora releases with a Class-A show stopper for a GPU bug, we are officially screwed. I'm sure the development community has considered this, but if ImgTec suddenly decides that the SGX on this particular board isn't profitable enough to warrant continued development... These just being examples, but there are MANY potential problems with using a binary-only drivers, especially on an embedded device like this. I use nVIDIA cards on nearly every Linux machine I own, but there was at one point a waiting period of six months for nVIDIA to fix a well documented compiz bug (GPU onboard memory running dry resulted in all the windows turning black instead of the driver simply running off into system memory). A whole ton of potential problems can arise from the SGX being closed-source.

@Chip, regarding a post very early on in this thread:

GPH's initial GPL violation isn't excusable by any means. They broke the license, they're at fault. You cannot just decide not to release source code because the software is in beta. If you release "GP2X OS Version .8", you cannot wait until 1.0 and give them 1.0 source code. Anybody that has a copy of 0.8 is entitled to the source code for Version 0.8. Call them zealots, fanatics, or lunatics all you want, it doesn't change the fact that some of these people have actually written the code in question and were entitled to get it back.

Lets say you were to call your software "beta", and compile some bit of code into the kernel that you didn't want the community to see (DRM for example). So when you released the source code, you removed the DRM code from the source package and called it final. The people that have a copy of your kernel are entitled to all of the code in that kernel (including all the changes you applied), so you have effectively violated the GPL.

The GPL is a non-negotiable license that protects the OSS community's ability to continue their development. It is one of the only staples that open source software has had to combat commercial takeover or corruption of open standards. It is in every way to your benefit that the organizations/individuals that utilize GPL'd code release their source, so I do not understand your complaints.

 

[WizardStan]

GPH's initial GPL violation isn't excusable by any means. They broke the license, they're at fault. You cannot just decide not to release source code because the software is in beta. If you release "GP2X OS Version .8", you cannot wait until 1.0 and give them 1.0 source code. Anybody that has a copy of 0.8 is entitled to the source code for Version 0.8. Call them zealots, fanatics, or lunatics all you want, it doesn't change the fact that some of these people have actually written the code in question and were entitled to get it back.

Totally off topic, but just a quick question, since you're obviously more knowledgeable about such things.
Suppose someone chooses to sell their software under GPL. They collect their money, and hand over a binary and the source, as required. Someone, through less than legal means, obtains a direct copy of the binary that is being sold. Is this person who has effectively stolen the binary still entitled to the source code? Or does that right only extend to those that have obtained the software legally?

Reasonably on topic:
Perhaps an agreement can be worked out with the SGX drives as it is for the wifi, that the Pandora team can sign some NDA to get the source to fix any bugs themselves. A reasonable compromise, everyone wins. TI can put less effort into driver support, and bugs get fixed faster. Not likely to happen though.
On the other hand, since these are entirely new systems, I can kind of envision them wanting these to work out better than, say, a PC. They need to prove that the SoC is reliable, so might actually expend more effort fixing driver bugs there than their desktop counterparts. Pure speculation, pay it no mind if you so wish.

 

[Megamixman]

GPH's initial GPL violation isn't excusable by any means. They broke the license, they're at fault. You cannot just decide not to release source code because the software is in beta. If you release "GP2X OS Version .8", you cannot wait until 1.0 and give them 1.0 source code. Anybody that has a copy of 0.8 is entitled to the source code for Version 0.8. Call them zealots, fanatics, or lunatics all you want, it doesn't change the fact that some of these people have actually written the code in question and were entitled to get it back.

Totally off topic, but just a quick question, since you're obviously more knowledgeable about such things.
Suppose someone chooses to sell their software under GPL. They collect their money, and hand over a binary and the source, as required. Someone, through less than legal means, obtains a direct copy of the binary that is being sold. Is this person who has effectively stolen the binary still entitled to the source code? Or does that right only extend to those that have obtained the software legally?

Reasonably on topic:
Perhaps an agreement can be worked out with the SGX drives as it is for the wifi, that the Pandora team can sign some NDA to get the source to fix any bugs themselves. A reasonable compromise, everyone wins. TI can put less effort into driver support, and bugs get fixed faster. Not likely to happen though.
On the other hand, since these are entirely new systems, I can kind of envision them wanting these to work out better than, say, a PC. They need to prove that the SoC is reliable, so might actually expend more effort fixing driver bugs there than their desktop counterparts. Pure speculation, pay it no mind if you so wish.



License don't apply when the means of procurement itself is illegal. That's like stealing a laptop, and expecting warranty to cover it if it breaks.

That situation is the situation with the Wifi Chip as of now. The main problem is that ImgTec is wildly protective of their IP. That being said, TI likely has an agreement with ImgTec for the lifetime of the 3XXX series, which is a long time keep in mind. So it is unlikely that they could just say "Hey, this chip is getting boring, let's move on".

 

[AireTamStorm]

Totally off topic, but just a quick question, since you're obviously more knowledgeable about such things.
Suppose someone chooses to sell their software under GPL. They collect their money, and hand over a binary and the source, as required. Someone, through less than legal means, obtains a direct copy of the binary that is being sold. Is this person who has effectively stolen the binary still entitled to the source code? Or does that right only extend to those that have obtained the software legally?

That is a bit of a grey area so to speak, but since the binary wasn't legally obtained I would assume that they are not entitled to the source code for that binary. If Company A compiles GPL'd Software A and Customer B buys Software A, then gets the source code, modifies and releases it for free, Then Customer C could get the binary without cost and therefore be entitled to the source code. If Customer C illegally obtains Software A, not only does he have no right to the binary, but the source as well.

Reasonably on topic:
Perhaps an agreement can be worked out with the SGX drives as it is for the wifi, that the Pandora team can sign some NDA to get the source to fix any bugs themselves. A reasonable compromise, everyone wins. TI can put less effort into driver support, and bugs get fixed faster. Not likely to happen though.
On the other hand, since these are entirely new systems, I can kind of envision them wanting these to work out better than, say, a PC. They need to prove that the SoC is reliable, so might actually expend more effort fixing driver bugs there than their desktop counterparts. Pure speculation, pay it no mind if you so wish.

I'm pretty sure that is the case- but we can all agree on there being potential problems. I wish companies would license out the source to community projects like this... it would only be good PR at this point. What IPs do they really have to protect in comparison to nVIDIA / ATI / Intel / VIA? I don't see binary-only being practical, especially so regarding driver development -or- IP protection at this point.

 

[Chip]

I'm not sure, but it looks like I'm in the minority regarding the GPU.

Isn't it highly disadvantageous to use a closed-source module? The SGX running on OMAP isn't even "proven" really.

Unfortunately that's true. However, what's the alternative? What other SoCs are available with similar capabilities that do have fully open source video drivers? The OMAP3530 is the best chip available. It's arguably the only chip available in it's class. Either you take what you can get with the SGX or you have nothing. You don't have to be happy about it, but Imgtec isn't going to change its policies over an itty bitty product like the Pandora. So either you make due with binary blog drivers or you cut off your nose to spite your face and cancel the whole project. It's nice to have ideals, but in the real world you have to make compromises.

GPH's initial GPL violation isn't excusable by any means. They broke the license, they're at fault. You cannot just decide not to release source code because the software is in beta. If you release "GP2X OS Version .8", you cannot wait until 1.0 and give them 1.0 source code. Anybody that has a copy of 0.8 is entitled to the source code for Version 0.8. Call them zealots, fanatics, or lunatics all you want, it doesn't change the fact that some of these people have actually written the code in question and were entitled to get it back.

And they got it back, as was intended all along. The actual issue was that GPH themselves didn't write the firmware. The contracted an outside company to do it, and that company screwed up. They included non-GPLable code in with the linux-based firmware. When GPH received the firmware with the source they realized they couldn't release it, but they had already taken pre-orders and people were clamoring for the hardware. They had a choice of delaying shipment for another couple months while the firmware was sorted out or shipping what they had and fixing it later.

They did the right thing and shipped what they had. They then removed the offending bits of code and sent out the new firmware and all relevant source code a couple months later.

Did they break the letter of the license? Yes, I guess they did. Did they violate the spirit of it? Not from where I'm standing. They were in a shitty situation and they made the best of it. As soon as it was possible to release a GPL compliant version of the firmware they did it. There was never any intention to "steal" code or "rip off" programmers. No harm, no foul.

My problem with the GPL devout is the same now as it was then - All they care about is the rules. Whether the rules may or may not be appropriate doesn't seem to matter. It's not like somebody is taking a piece of GPL software, changing 2 lines, then charging a boatload of money for it and not giving out the source. This wasn't the case with the GP2X and it isn't the case with Pandora.

The GPL is a non-negotiable license that protects the OSS community's ability to continue their development. It is one of the only staples that open source software has had to combat commercial takeover or corruption of open standards. It is in every way to your benefit that the organizations/individuals that utilize GPL'd code release their source, so I do not understand your complaints.

My complaint is that the GPL brigade rode into town, screaming like madmen and shooting up the place, despite the fact that nobody was ever in any danger of the things you're talking about. Nobody was ripped off, corrupted, or taken over. No harm was ever intended and none was suffered. It was a complete non-issue from the beginning but because the letter of the law had been broken, a huge deal was made of it.

Now we have the Pandora with its closed source video and wifi drivers. Are they GPL compliant?!? Are they against the rules?!? Shouldn't we do something?!?

Here's a better question - Should it matter?

Is this whole "derivative work" clause really necessary? Are you really going to claim that drivers (which do not contain any of the original GPL code) must be open source because they work with a GPL OS? Do you really think you can tell people "Sorry, you can't have proper drivers for the hardware you purchased because of a licensing issue" and expect them to take you seriously? Ask Creative Labs how well that worked out for them.

Fact: If your license makes unreasonable demands, it will be ignored. Look at DVDs (ripped constantly), the iPhone (majority jailbroken), or the PSP (do you know anybody who has one that isn't hacked?). Those are all instances where the user license is unnecessarily restrictive, and is therefore ignored by a large portion of users. So how do you expect me to feel about the GPL, which made unreasonable demands in the case of the GP2X and which is threatening to do the same in the case of the Pandora?