Should The Pandora Become An Ubuntu Partner?

[viridior]

Not to detract from the general Ubuntu discussion, because it is interesting, but why not use Gentoo? Like Angstrom it can easily be configured to highly optimize for hardware, and like Ubuntu has a large library of software to choose from that are already pre-configured for installation.

I see this coming "But doesn't Gentoo compile everything for source, and doing that on the Pandora is crazy!", portage allows direct installation of pre-compiled binaries with just two or three extra lines in a config file. I am looking around for a, hopefully free, server that the Gentoo-arm-pandora project can use as a binary repository that will be available to anyone interested. The goal of the Gentoo-arm-pandora project to provide a pre-built image that will initially work from an SDHC (eventually on the NAND). Once you boot, you can configure the system to run whatever you want. KDE, Gnome, Xfce, e17, Firefox, etc, etc... If you are interested, please check us out.

I do like the idea of providing more OS options to the Pandora, and I think that any publicity for the Pandora will do nothing but help the platform, the devs, and the community. I hope that an Ubuntu arm system is successful and I am looking forward seeing it around.

 

[gruad]

Not to detract from the general Ubuntu discussion, because it is interesting, but why not use Gentoo? Like Angstrom it can easily be configured to highly optimize for hardware, and like Ubuntu has a large library of software to choose from that are already pre-configured for installation.

I see this coming "But doesn't Gentoo compile everything for source, and doing that on the Pandora is crazy!", portage allows direct installation of pre-compiled binaries with just two or three extra lines in a config file. I am looking around for a, hopefully free, server that the Gentoo-arm-pandora project can use as a binary repository that will be available to anyone interested. The goal of the Gentoo-arm-pandora project to provide a pre-built image that will initially work from an SDHC (eventually on the NAND). Once you boot, you can configure the system to run whatever you want. KDE, Gnome, Xfce, e17, Firefox, etc, etc... If you are interested, please check us out.

I do like the idea of providing more OS options to the Pandora, and I think that any publicity for the Pandora will do nothing but help the platform, the devs, and the community. I hope that an Ubuntu arm system is successful and I am looking forward seeing it around.

That's totally beside the point, since Gentoo is not looking for partnership. If you choose to set up a Gentoo repo for Pandora, that's a great thing, of course. It just won't give us the instant maintenance resources and publicity that a partnership with Ubuntu would bring.

There is nothing special about Gentoo or Angstrom that makes them more suitable for hardware optimizations. Last I checked, gcc on Ubuntu had compiler flags too.

The most important advantage of Ubuntu, to me, is that they offer proper releases as opposed to rolling updates.

 

[viridior]

Not to detract from the general Ubuntu discussion, because it is interesting, but why not use Gentoo? Like Angstrom it can easily be configured to highly optimize for hardware, and like Ubuntu has a large library of software to choose from that are already pre-configured for installation.

I see this coming "But doesn't Gentoo compile everything for source, and doing that on the Pandora is crazy!", portage allows direct installation of pre-compiled binaries with just two or three extra lines in a config file. I am looking around for a, hopefully free, server that the Gentoo-arm-pandora project can use as a binary repository that will be available to anyone interested. The goal of the Gentoo-arm-pandora project to provide a pre-built image that will initially work from an SDHC (eventually on the NAND). Once you boot, you can configure the system to run whatever you want. KDE, Gnome, Xfce, e17, Firefox, etc, etc... If you are interested, please check us out.

I do like the idea of providing more OS options to the Pandora, and I think that any publicity for the Pandora will do nothing but help the platform, the devs, and the community. I hope that an Ubuntu arm system is successful and I am looking forward seeing it around.

That's totally beside the point, since Gentoo is not looking for partnership. If you choose to set up a Gentoo repo for Pandora, that's a great thing, of course. It just won't give us the instant maintenance resources and publicity that a partnership with Ubuntu would bring.

There is nothing special about Gentoo or Angstrom that makes them more suitable for hardware optimizations. Last I checked, gcc on Ubuntu had compiler flags too.

The most important advantage of Ubuntu, to me, is that they offer proper releases as opposed to rolling updates.



You make a good point about Ubuntu and gcc, does Ubuntu have a streamlined way to include gcc changes into all installed packages without completely changing the software base? Will this require people to make these modifications by hand? I have 2 other issues with your comments.

1. The discussion a long time ago evolved into advantages of using Ubuntu vs. Angstrom, both of which have valid points, and the beginning topic was the idea of a potential partnership. ANY publicity for the Pandora project will be good, maybe I have been on a different planet for the last few years, but I didn't know that Ubuntu had taken over and had single-handedly cured cancer. It is a powerful and great OS, but last time I checked every other Linux OS since and before Ubuntu used the same software base.

2. I'm not sure what the definition of what a "proper" release is. I may need some help. Is that a 'stable' version of software that has been tested, vetted, and just works? I've used that on many systems, Gentoo included. Gentoo, and systems like that, don't require the latest sources, they can easily be used in "traditional" releases (twice a year for Gentoo). Regardless, I'm not advocating the required use of bleeding-edge technology on any platform, but as far as using Ubuntu, or Angstrom, or Gentoo, just the fact that the platform will be using ARM means that it will need a lot of people to work on the issues of using software that probably had been designed for a x86 in mind when developed, and so I would argue that your "proper" releases argument is mostly moot.

 

[Svartalf]

It's quite amusing really. 'Back in the day' you'd install the 'full' OS on your PC's hard disk and load up all the drivers and TSRs you would need for a full day's productivity in your config.sys and autoexec.bat. You'd create a boot-disk for a more streamlined OS to get the maximum out of your computer. The Pandora is a mirror of the old times, you'll have a simple system installed on the unit to get the most out of the hardware, and you'll need a boot-disk to load up a full OS

You're quite mistaken on the "full OS" line... Think of it more along the lines of having DOS vs Win3.11 (Which is a very bad, bad analogy, but it fits in many people's minds...). Angstrom is nearly as "full" as Ubuntu. It's just designed tighter with a few different design tradeoffs made to allow it to fit into tighter spaces, and we're coupling it with a lightweight application launcher instead of a more bulky window manager/app framework...

 

[Chip]

The discussion a long time ago evolved into advantages of using Ubuntu vs. Angstrom, both of which have valid points, and the beginning topic was the idea of a potential partnership.
Which is kind of odd, since we're definitely going to have both anyway.

Just because we're offering assistance to Canonical, doesn't mean we're abandoning Angstrom. Angstrom development for the Pandora continues as we speak.

 

[viridior]

The discussion a long time ago evolved into advantages of using Ubuntu vs. Angstrom, both of which have valid points, and the beginning topic was the idea of a potential partnership.
Which is kind of odd, since we're definitely going to have both anyway.

Just because we're offering assistance to Canonical, doesn't mean we're abandoning Angstrom. Angstrom development for the Pandora continues as we speak.

That is excellent, maybe we can start talking about using and developing for these systems, and their advantages as options to users, instead of why one is better then the other. I think the people that will be using these platforms are intelligent enough to figure out what best fits their needs.

Chip: (Probably a development corner question, but it seems appropriate here.) I've checked the pandora wiki and searched the forums, is there a method/place that we, users/non-main developers, can use to see what software is being used/tested? It might go a long way at allowing people to start work on porting/using what software that want to use prior to shipment.

 

[Chip]

I've checked the pandora wiki and searched the forums, is there a method/place that we, users/non-main developers, can use to see what software is being used/tested? It might go a long way at allowing people to start work on porting/using what software that want to use prior to shipment.
Not at the moment, but that may change soon. Right now the lead devs are working on very boring, very low-level stuff. They'd prefer to work on things without too many distractions from the peanut gallery. As we get closer to the Pandora shipping and the software gets closer to being finished, we'll probably open things up for inspection / comment.

 

[viridior]

I've checked the pandora wiki and searched the forums, is there a method/place that we, users/non-main developers, can use to see what software is being used/tested? It might go a long way at allowing people to start work on porting/using what software that want to use prior to shipment.
Not at the moment, but that may change soon. Right now the lead devs are working on very boring, very low-level stuff. They'd prefer to work on things without too many distractions from the peanut gallery. As we get closer to the Pandora shipping and the software gets closer to being finished, we'll probably open things up for inspection / comment.

Completely understandable, we are having a great time over at gentoo-arm-pandora working with what I'm sure are similar issues, hopefully you have hardware to test on!

I imagine that the pandora wiki will be used for most of the documentation, so I will continue to check in there, and I'll just continue to slow roast in the mean-time. (I'm a southerner, so I don't know if you would know much about boiled peanuts...)

BTW, if any of the devs are in the Washington, D.C. area, send me a PM and I'll buy a few tall ones for all of you, thanks for all the great work! I promise I'm good for it.

 

[Svartalf]

Just because we're offering assistance to Canonical, doesn't mean we're abandoning Angstrom. Angstrom development for the Pandora continues as we speak.

Yep. That it is. (There's a reason why I was commenting as I was earlier in this thread... )

 

[Svartalf]

Right now the lead devs are working on very boring, very low-level stuff.

Yep. That they are...

QUOTE

They'd prefer to work on things without too many distractions from the peanut gallery. As we get closer to the Pandora shipping and the software gets closer to being finished, we'll probably open things up for inspection / comment.



As a suggestion to people concerned about their stuff working/not working...

Do things in SDL (Even if you don't "like" the library and want to code to the bare metal, I'd suggest NOT doing that to be able to provide your title to a broader audience...) Doing it in SDL, you can code to a Windows/Mac/Linux target audience AND to the Pandora, Beagleboard, and Wiz- largely all at the same time as long as you fit within the performance envelope for the targets in question. It's also worth noting that if you do things in SDL as long as the lib's not busted on the target, it's going to work largely the same on all targets.

Do things against the OpenGL ES 1.1 or 2.0 wrapper libs. We're going to pretty much have those two APIs in hand shortly. If you use the Windows/Linux/MacOS dev wrapper libs, you can fix any OpenGL 1.3 or 2.0 code to work right- and you can expect the title to largely work out of box once you compile to ARM for the Pandora. If you fit in the memory profile for the Beagleboard, it'll work there too. If you're an ES 1.1 app and can fit within the memory and CPU performance envelope for the Wiz you can expect it to largely work on all three targets. For the SDL part of things, there should be a version showing shortly that copes with ES 1.1 and 2.0 that's derived from the iPhone efforts and an experiment someone did for the N800 and ES 1.0.

If you're not doing 3D, you can still probably score a Beagleboard for a verification target and the world there is going to be similar (but not identical, mind...) to what you'll find on a Pandora. If you want to gear up all the way and test things initially or want to compile up something for someone with a dev board to test for you here shortly go look at what all you need to develop for the Beagle.

 

[mali]

Damn, nobody reads in off topic, anyway. It fits a litte bit here, so I post a link if you don't mind:
http://www.gp32x.de/board/index.php?showtopic=45349
I wonder with which distro it ships, there's only stated, that it runs linux.

 

[Svartalf]

Damn, nobody reads in off topic, anyway. It fits a litte bit here, so I post a link if you don't mind:
http://www.gp32x.de/board/index.php?showtopic=45349
I wonder with which distro it ships, there's only stated, that it runs linux.


The big takeaway on that device is that it's using an ARM11 CPU in it's SoC...

The ARM11 is in the same class as the Wiz and the Nokia N800. It's nice. Not anywhere near as nice enough compared to the Pandora and a mobilized Beagleboard.

 

[mali]

Damn, nobody reads in off topic, anyway. It fits a litte bit here, so I post a link if you don't mind:
http://www.gp32x.de/board/index.php?showtopic=45349
I wonder with which distro it ships, there's only stated, that it runs linux.


The big takeaway on that device is that it's using an ARM11 CPU in it's SoC...

The ARM11 is in the same class as the Wiz and the Nokia N800. It's nice. Not anywhere near as nice enough compared to the Pandora and a mobilized Beagleboard.

And looking at the battery-life, this thing is pretty useless
Is the SoC so bad, or is it the LCD, that needs so much?

 

[Svartalf]

And looking at the battery-life, this thing is pretty useless
Is the SoC so bad, or is it the LCD, that needs so much?

I'm not quite sure about that. The N800 with an OMAP2 will go 3-4 hours on a 1500mAh battery. They're using the same class of battery that the Pandora's using and to get 1/2 or less of the life we're seeing...

I could finger the backlight as being part of the culprit there, but I have to wonder what else is eating juice like that.

 

[javaJake]

P.S. I don't understand all the nerd-love for Ubuntu. It's a perfectly fine Linux distro, I recommend it for the desktop. But why does it have such a cult? Why does the mere mention of its name get people so hot and bothered?

I'm a bit late to this party, but I haven't seen anyone else make this pro-Ubuntu comment yet, and it is on topic.

The primary reason for Ubuntu is it's so easy to use. GNOME (and its simple, clutter-free GUI), coupled with excellent hardware compatibility (Ubuntu 8.10 now works instantly on all my trouble PC's and laptops), a beautiful interface, and an excellent community to prop up any shaky "newbies", Ubuntu is one of the best distributions available. And it shows in its rankings. The fact that it has an actual business behind it makes it easier to enter the "real world", namely on Dell computers.

That's the reason why my family's computers all run Ubuntu. That's why I used it as my primary operating system for the longest time while I got acclimated to Linux (and before I switched to Gentoo).

That reason isn't good enough for the folks using the Pandora however. It's assumed that most users here can at least learn how to cope with unexpected surprises when venturing out into untested territory. Ubuntu is designed for the newly-converted Windows audience. The Pandora is not.

So, that's my 2 cents. Take it or leave it.

 

[Brockway]

I'm fairly sure the Pandora's default Angrstrom OS will be easier to use for a windows convert than full blown Ubuntu. Its designed to be simple, right? The icons are just there, it'll have most of the same apps like Firefox already, and it doesn't look like you can mess it up as easy. It'll probably be much faster too.

But, I will more than likely install Ubuntu on a SD card anyways, just because I can. Its easier to pass it off as a netbook with a full desktop OS running on it, and really, thats why I preordered a Pandora. To replace my slow N800 with something that actually fits my needs, an ultraportable computer that I can take anywhere and do most anything my laptop could.

 

[PoisonedV]

P.S. I don't understand all the nerd-love for Ubuntu. It's a perfectly fine Linux distro, I recommend it for the desktop. But why does it have such a cult? Why does the mere mention of its name get people so hot and bothered?

I'm a bit late to this party, but I haven't seen anyone else make this pro-Ubuntu comment yet, and it is on topic.

The primary reason for Ubuntu is it's so easy to use. GNOME (and its simple, clutter-free GUI), coupled with excellent hardware compatibility (Ubuntu 8.10 now works instantly on all my trouble PC's and laptops), a beautiful interface, and an excellent community to prop up any shaky "newbies", Ubuntu is one of the best distributions available. And it shows in its rankings. The fact that it has an actual business behind it makes it easier to enter the "real world", namely on Dell computers.

That's the reason why my family's computers all run Ubuntu. That's why I used it as my primary operating system for the longest time while I got acclimated to Linux (and before I switched to Gentoo).

That reason isn't good enough for the folks using the Pandora however. It's assumed that most users here can at least learn how to cope with unexpected surprises when venturing out into untested territory. Ubuntu is designed for the newly-converted Windows audience. The Pandora is not.

So, that's my 2 cents. Take it or leave it.

it takes me about 2 minutes to make debian, or really any recent linux distro, look exactly like ubuntu

 

[javaJake]

it takes me about 2 minutes to make debian, or really any recent linux distro, look exactly like ubuntu
Well, I'm sure you understand that it isn't all about looks, obviously.

 

[PoisonedV]

it takes me about 2 minutes to make debian, or really any recent linux distro, look exactly like ubuntu
Well, I'm sure you understand that it isn't all about looks, obviously.

then what would be the other practical uses of it?

 

[BenTolputt]

QUOTE
it takes me about 2 minutes to make debian, or really any recent linux distro, look exactly like ubuntu


Wonderful. How long does it take for you to get that as a pre-install option on Dell machines? How long to develop a helpful community & Google-indexed wealth of information to help new users? Remember, communities where people get slapped down for not RTFM are not helpful for new non-technical users!

Ubuntu is not a great because it is optimised, memory efficient, or guaranteed more user friendly than any other distro out there. It is great because of the community and helpful, step-by-step instructions that have been developed around it. There is ALOT of redundant information in the instructions for doing something in Ubuntu, but for people not familiar with Linux (some haven't ever used a command prompt) - this is not a Bad Thing.