AMD Joins Intel's MeeGo OS Effort 117
angry tapir writes "In an effort to expand software compatibility for its upcoming Fusion chips, AMD has joined rival Intel's efforts to develop the open-source MeeGo OS. AMD 'will provide engineering expertise intended to help establish the technical foundations for next-generation mobile platforms and embedded devices,' the company said in a blog post on its website."
Re:Desktop dinosaurs realize mobile cannibalizatio (Score:1, Interesting)
Sorry, too much marketing speak.
Maybe AMD is getting involved because one, it can. It's pretty much the only real open source project for mobile that is being worked on. They don't really have to work spend a lot of cash on contractual obligations to get involved. And two, because err...well, related to number one. It works on netbooks too. It's not tied to mobile chips. They're definitely targeting mobile, but the stuff will work on just about anything. Netbooks, embedded, cell phones, etc. There's nothing stopping them from including x86. That might be what brought AMD to the table.
I was completely with you until the end... (Score:5, Interesting)
3. In 2 years when the 'tablet computing' fad has largely blown off, Intel and AMD will realize that its just not that important to keep pushing down their marginal revenues until its just not worth the investment to keep with it.
I just don't agree that 'tablet computing' is a fad. Bill Gates was definitely correct that this form factor held lots of promise but Microsoft just could not untether themselves from the Windows gravy-train and thus constantly missed the mark with their tablet OS (not to mention pricing put it well out of the general consumer market). Intel got lucky in an emergency conspiracy with Microsoft to kill the netbook market but the tablet market is out of their control (with upcoming wave: Google/Android, HP/Palm, and RIM/BlackberryOS will put a permanent dent in PC sales). I bet both Intel and AMD are pretty disappointed in Microsoft's lack of real competitive OS here and have no "in" with the tablet players of today.
Re:Desktop dinosaurs realize mobile cannibalizatio (Score:3, Interesting)
I don't know. I think it's more bandwagon jumping than "joining forces."
I think this is a case of AMD knowing what side it's bread is buttered on. If they can drum up more business through direct sales because linux now works as well or better than a competitor(like say, intel, or ARM)'s product, and the cost is a few dozen engineers on a crazy fun short term amount of time spent kernel hacking, I'd say it's worth it.
From what it sounds like, MeeGo might be a less than stellar MID or phone OS compared to say, Android, but, MeeGo's probably my first choice if I had to design a DVR set top box, or a kiosk that would be shipped to thousands and thousands of locations.
Are you sh*tting me? (Score:5, Interesting)
*puts on anti-Nokia troll hat*
I don't know about hats, but it sure sounds like you're trolling!
What's stopping Google from using code in the MeeGo base?
MeeGo uses a (somewhat) stock kernel, I believe. Android puts in all kinds of special sauce not in mainline like wake-locks.
Drivers written for Android aren't necessarily going to just work in MeeGo, unless you add all that additional stuff (cruft?) to the kernel, etc...
Like most popular distros, MeeGo uses the standard GNU userland; Android uses their own, non-GPL userland.
we'll probably see a lot of cross contamination between Android and MeeGo kernel code.
Sure, anything that's from upstream.
(I suspect that since Android's source is completely open, there would be no...
Android, like MeeGo is largely open. But there are certain things that are not released under a FOSS license (e.g. some drivers, particularly power-related and graphics-related drivers).
On Android, I believe that all of Google's core applications are completely closed-source. What's more, it's non-trivial to set up the phone to sync with non-Google servers.
Nokia would steal Google's UI code and Dalvik
I don't think that Android's UI is particularly better or worse than MeeGo's.
And Dalvik? That's like asking the MeeGo folks to go stick their hand in a beehive filled with thousands of tiny little Larry Ellisons with stingers. Surely, you must be joking!
Google would steal their best threading, i/o, and whatever other code is probably superior in MeeGo that isn't probably going to wind up in the base kernel trunk
Part of the whole point with MeeGo is to try to get as much stuff pushed up into upstream projects as is possible. If there's some good threading or i/o improvements to be made to the kernel, it seems reasonable that the MeeGo kernel devs will work hard to get it into mainline. From my perspective, Android has an "after the fact" attitude towards kernel development, whereas MeeGo has more of a "let's cooperate with upstream" attitude.
I don't know if ARM is working directly on the Linux kernel/Android or not
Sure. One of the projects they sponsor is Linaro. Linaro is a projects tasked with making it easier to deploy Linux-based systems on top of ARM: http://www.linaro.org/ [linaro.org]
Re:Desktop dinosaurs realize mobile cannibalizatio (Score:3, Interesting)
You really need to get your hands on one of the tablet flops coming out with Android to understand just how bad android can be. Even established makers like Toshiba have a 80%+ return rate. It just goes to show that you need proper UE and proper testing and proper development to create something even if you start with a "ready" OS.
In any case, MeeGo is geared not just towards the phone and Tablet market. All those STBs, media SoCs in TVs using bespoke builds are ripe for the picking and make a much easier target. It is also a market where Intel has some clout as their media processor + Atom combination delivers same or better performance and price/performance than the incumbent (predominantly MIPS) offerings from Pace and the like.
Re:Desktop dinosaurs realize mobile cannibalizatio (Score:2, Interesting)
The one piece I've noticed a lot of geeks miss about OSes is that UI is a really, *REALLY* important piece, particularly when you're targeting Apple, in the mobile sphere. It's not about just good, it's also about consistent.
If Nokia wants me to build Qt framework applications, they need to get UIs consistent and in order, particularly if they want me to be able to develop from anything from a netbook, to a MID to an STB.
Re:Recompiling for different mobile processors (Score:1, Interesting)
I don't see your problem. It works for just about any established Linux distro; e.g. for each of my systems, a i7-920 desktop Ubuntu64, a 32bit Atom-based Server, a ARM Kirkwood-based plugcomputer or my ARM Cortex based smartphone, I can issue one and the same command to get the app I want, for instance "apt-get install lynx" and it will download the correct binary for any of the platforms.
An appstore could function just like the other Linux repositories.
Re:I was completely with you until the end... (Score:2, Interesting)
It's a fad. Not that it doesn't have its place, but considering that Apple and their brain dead followers in the mainstream press have pushed it as a "revolution" that would "change" (oh, where have I heard that before?) "everything", while being less capable and twice as expensive as a netbook, and that the only influence it has had is that news outlets can push their wares in the form of paid "apps" through iTunes instead of the usual web content, it just isn't worth the attention it has received. The tablets are pushed for everything, and are unsuitable for most: poor ebook readers, poor computers.