Guest post by Qualcomm showed off a previously unannounced Eee PC running Android at Computex in Taipei. The machine has a 10-inch screen, built-in webcam, and a universal 3G radio that supports all UMTS and CDMA networks on all frequencies used around the world. The ‘smartbook’ runs on Qualcomm’s Snapdragon, a 1GHz ARM processor core that marks a shift away from Intel Atom x86-based netbooks. A second Android-based netbook — a prototype by contract hardware maker Compal Electronics — was also demoed at the show. Google, meanwhile, declined to discuss what steps it is taking to adapt the smartphone OS for laptops.
Snapdragon is an ARM chip and thus not x86 compatible. Therefore Windows (the desktop line – I know about and am ignoring CE) will not run on it (unless MS have been doing what Apple did and maintaining versions of Windows for different processors – just in case)
So if Asus/Eee are planning a new completely non-windows ARM based device, what is this all about http://www.osnews.com/story/21589/Asus_Microsoft_Launch_Anti-Linux_…
A case of the left hand not knowing what the right hand is doing?
An arm port of Windows wont be of much use without apps to run on it.
An arm port of Windows wont be of much use without apps to run on it.
Well, Microsoft could always integrate x86 emulation into Windows so that you could run x86 applications on ARM Windows.. You’d sacrifice helluva lot of performance then though, so it wouldn’t work for any intensive applications. It would probably work for any simpler apps like f.ex. office work.
ARM processors running at 1GHz in battery-powered mobile devices just don’t have enough processing grunt to make such a thing practical.
Is it possible? Yes. Is it useable at al? No. Sorry.
If you want an application to work on an ARM device of this capability … take the original source code and re-compile it for ARM. That is the ONLY practical way to run things on on an ARM device.
If you have an x86 binary-only copy of an application, you can forget about trying to run it satisfactorily on an ARM netbook.
Or the case of someone putting up a website which has no association with ASUS – I’ve seen many websites in the past trying to pass themselves off as Microsoft websites when in reality they’re nothing more than pathetic resellers trying to boost their sales through dodgy marketing tactics. Given that the website is registered to a ‘Michael Sharp’ to which the phone number provided on the whois cannot be verified and the address given doesn’t give a specific organisation given that the address has something like 14 business operating from it – I doubt it has any link back to ASUS.
As for ARM processors; they need to get Flash working, without flash and some decent CODEC support for music and video, its going to be ‘epic fail’. The problem that I see is that if ASUS does ship this, they’ll try to do it on the cheap and thus have missing key components to make the internet experience alot more pleasant.
As for ARM processors; they need to get Flash working, without flash and some decent CODEC support for music and video, its going to be ‘epic fail’. The problem that I see is that if ASUS does ship this, they’ll try to do it on the cheap and thus have missing key components to make the internet experience alot more pleasant.
Adobe Flash does exist for ARM processors already AFAIK, and they could always use Gnash as a backup. And FFMPEG supports almost every codec out there and what it doesn’t support can be found in other packages. I don’t have that Win32codecs package installed on my Linux at all yet everything I’ve thrown at my Linux has worked just peachy.
But yeah, given ASUS’s bad track record at actually doing any decent job at anything their product will most likely not be anything spectacular.
Even if they shipped FFMPEG, they would still have to pay royalties for using patented CODECS, and the win32codecs package is of questionable legality; so I don’t see either solution being used let alone Asus trying to get Gnash working considering that Gnash is still far from being in a stable and useful state (non-feature complete).
As for the ARM version, they Arm version is a cut down version of Flash so it isn’t equal to the full desktop version of it. What Arm needs is the fully fledged Flash that is available on the desktop and not some cut down mobile phone version which is missing key features which some developers are sure to use.
Asus when it comes to making hardware is ok – I have a Eee PC and as happy as larry, and when it comes to installing turn key solutions that require them to make no effort in terms of customisation they seem to do an ok job at it; btw, I am doing a report on the Eeepc which I bought, lots of praise for it btw when it comes to using it as a netbook. I just hope that maybe in a couple of years some of the rough edges of Linux will be smoothed – but lets be honest though, the focus on the desktop by a single vendor with some dollars.
ffmpeg is free software (and freedom software). It is more capable than w32codecs, and it is not just a bunch of x86 binary-only codecs designed for Windows. Unlike w32codecs, ffmpeg is not a copy of anyone else’s code. ffmpeg violates no copyrights.
As for patents … extremely dubious. There is no “physical machine” involved in the function of a codec, it is pure mathematics. As such, in almost every country on the planet, a codec is not patentable. Even in the US, in view of the in re Bilski decision, a codec is probably not patentable.
So in using w32codecs you could fall foul of copyright law, if it can be shown that you are distributiong a copy of someone else’s code. they are x86 only anyway, so of no use to ARM.
ffmpeg/libavcodec OTOH is not a copy of proprietary code. There should be no problem distributing it.
Adobe and open source:
http://opensource.adobe.com/wiki/display/site/Home
http://opensource.adobe.com/wiki/display/site/Projects
http://stlab.adobe.com/
Edited 2009-06-02 00:11 UTC
More on patentability questions in the US:
http://www.patentlyo.com/patent/2009/06/bilski.html
http://yro.slashdot.org/article.pl?sid=09/06/01/2158224
Edited 2009-06-02 00:30 UTC
So what, software patents exist as they are right now – until software patents are over turned – companies have to operate in this place called reality; where the sky is blue, grass is green, the US has a 2 trillion budget deficit and there are software patents.
Edited 2009-06-02 04:51 UTC
But something tells me that there are some companies that are already not taking US into account. That something is reality and history, and it comes from Samsung and a bunch of Japanese and Chinese companies NOT pleanning to release some of their products outside their respective countries. So get you head out of that US centric ass and think a bit.
Why the f*cking hell would I have a ‘US Centric’ way of thinking when I am not even from the US! Your lack of location in your profile has killed any possible credibility you thought that you might have had here.
Software patents as a concept are ridiculous, but that doesn’t justify pretending they don’t exist. Bliski is going up for review with SCOTUS, which is hopefully a good thing in terms of reinforcing the level of ridiculousness, but for now the reality is that patents are a minefield for any commercial company distributing products in a jurisdiction where they are held valid. Just ask SanDisk, who had their MP3 players seized under an injunction by a German court, for not paying licensing fees. Philosophical arguments on the part of free software are legitimate, but they don’t negate reality.
Considering the w32codecs are lifted from Windows and therefore MS code, you’re correct, they will generally fall foul of copyright law in any jurisdiction.
Clean room coding doesn’t magically protect one from patent infringement. OSS implementations of patented codecs are still patent infringements. Note, as I said above, I think software patents are BS, but they are a legitimate obstacle for any commercial organization hoping to leverage OSS in a business model. At the end of the day, multi-billion dollar organizations like Asus can’t simply dismiss the risk of patent infringement because the community decides they’re invalid. Microsoft just got hit with a $200M judgement because of some freaking company that claims a patent on interpreting XML in documents. It’s pathetic, but any commercial organization has to take that risk/liability into concern.
The mainstream vendors simply can’t ship a netbook with ffmpeg as a solution for codec compatibility, the exposure/liability is to high. I only bring this up because I find that, too often, the free software community is too willing to point to solutions like this and say “problem solved”. It’s not solved. It won’t happen. The community is free to download “illegal” codecs at will, but manufacturers can’t include them, and that simply adds an extra roadblock.
Asus et al. will have to shell out for licenses, much as fluendo does, if they want to legitimize the popular proprietary codecs on a free platform.
The market has become too addicted to YouTube et al., with videos on demand. As long as the content providers and purveyors support proprietary codecs, it will hinder widespread FLOSS adoption.
It would be better for the community to only accept standards like OGG. Sure, it sucks compared to the alternatives, but sacrifice is required if the community truly wants open standards. Dirac may help, if they can get it straight as well. And if users aren’t willing to sacrifice in order to accept free codecs, then the proprietary ones have pretty much justified their existence.
But in the mean time, it helps nothing to pretend that the patents are meaningless just because they are vacuous in some jurisdictions and not others. The vendors will not adopt the OSS work-arounds. This is an issue that needs to be addressed as aggressively as the ODF issue was, at gov’t levels.
They exist in the US, insofar as the USPTO has issued them. AFAIK, none of them have survived an actual court challenge. The general idea of codecs has reams of prior art. Specific codecs differ from others only in respect of the mathematics they use, and even in the US, despite USPTO’s behaviour, you cannot patent mathematics.
The reality being that one cannot without risk simply distribute codecs (even if you have rights to the source code) with a product you wish to sell in the US.
Sadly, agreed. I feel somewhat for US residents in this respect.
Agreed. there is no proper role for w3codecs on a Linux system IMO.
It does in all countries in which the codecs are not patented. That is most of them.
Not in most countries.
Agreed. This means only that the codecs package cannot be shipped fully. Only some codecs can be shipped in the default installation. However, most distributions will tell you what packages are missing when you try to play a file which requires a codec.
For example, Amarok will tell you that you need to install the libxine-ffmpeg package if you try to play an .mp3 file. If you then try to install package libxine-ffmpeg, the package manager will tell you that you need to enable the restricted-extras repository (or something like that). Enable the repository, and then libxine-ffmpeg is installable. With libxine-ffmpeg installed, Amarok can then play .mp3 files.
Its not rocket science.
Yet the Kubuntu distribution was not shipped with a .mp3 codec.
They are not illegal at all in most countries. It is not at all difficult to install them post delivery.
If I were Asus, I would NOT distribute them (because of the risk). But I would have the software state that the required codec is not installed, and name the package that included the required codec (that is legal in most countries).
Meanwhile, for the US only, there are legal codecs available for purchase. Google for “fluendo”.
That is a bad example. Adobe have opened up the .flv specification, and given permission for anyone to write software to play such files. Gnash and swfdec, either one of which will allow you to play youtube videos, are perfectly legal codecs everywhere. The .flv codec which is part of ffmpeg (which allows for example VLC to play downloaded flash videos) is likewise perfectly legal.
… and .swf and .flv. Theora, Vorbis, Dirac, FLAC … there are actually quite a number available. You don’t even need OGG, you can use Matroska instead.
Vorbis is better than .mp3. Theora is catching up, and recently almost gained parity with H264.
Why? All it means to most of the people on the planet is that they have to follow a few extra steps to get multimedia to play. This is no different to Windows.
Yes. It shouldn’t be an issue, and the US seriously needs to fix this problem. But in most places, it isn’t really a problem, but more like an annoyance.
Edited 2009-06-02 06:38 UTC
Who the hell said it was a copy of anything or used win32 binaries or any of the other crap you mentioned in that paragraph? you really are getting to the end of my tether right now – you have addresses NOTHING and started a rant about issues I NEVER raised.
Who gives a shit about that – as it stands right now there are software patents, love or loath it. You may live in lalala land where courts are a voluntary thing to abide by but everywhere else in the world there are patents and if you ship patented technology with your product – you’re going to get sued; end of story.
Which doesn’t change the fact that there are patents and they violate those patents. Until software patents suddenly evaporate (which I wish they did), we have to play in this world called reality.
So what – I don’t see the full source code for Flash Plugin – so no soup for you!
I didn’t rant, I just stated some relevant facts that you omitted.
The facts are relevant because, for most of the people on the planet, the codecs are not patented, and consequently the only legal protection that proprietary binary codecs have is copyright. ffmpeg does not violate copyright, so for the majority of people, their use is not illegal at all.
Ergo, multimedia codecs are not a problem on Linux for most people.
For the small percentage of people living under repressive regimes where there is a problem, there is still a legal alternative in the Fluendo codecs where one can pay the burdonsome tax that codecs attract in those countries.
Do you know the difference between patents and copyright? if not – may I suggest you shut up until such time that you know the difference.
Yes, I know the difference. It is not a secret, after all.
http://en.wikipedia.org/wiki/Patent
http://en.wikipedia.org/wiki/Copyright
Now, did YOU know that copyrights are in most countries automatic? That means if you publish something, then you automatically have copyright on it if you are the author … all over the world.
Patents, OTOH, you have to apply for. You get awarded a patent in the country that you applied for it. Not everywhere. Patents have limited jurisdiction. If you want patent coverage in multiple countries … you have to apply for the patent in each country.
The US is trying to get patents recognised across country boundaries in a type of “you recognise ours and we’ll recognise yours” type of arrangement, but so far this doesn’t hold for the most part. This is largely because the US is way, way to liberal in giving out patents for the silliest things. What other country would agree to abide by that, and suddenly owe some US company royalty money every time they wanted to sneeze?
So, US patents are recognised in only a few other countries. Most codecs have US patents only.
So … for most of the people of the world, US patents on codecs do not apply.
Way to go about moving the goal posts from patents to copyright.
Regarding patents; there are international agreements where patents are automatically inherited by all those who are signed up to certain treaties. But hey, you keep ignoring those international agreements – maybe your dreams will come true.
Edited 2009-06-03 09:42 UTC
Excuse me?
Software is weird, for some reason it has become accepted (mostly thankfully just in the US) that software can be protected both by patent and by copyright.
Copyright protection – no-one has any qualms with it. You shouldn’t be able to profit from the work of others without their permission.
In Linux, the fact that win32codecs are copyrighted is not a problem, because ffmpeg codecs are not the same code … yes they are work-alike, but they are not a copy. ffmpeg codecs do not violate copyrights held by corporations.
Patents are different. It is possible to violate a valid patent even if the code is NOT a literal copy of the original invention. Even a work-alike codec can violate a patent. Patent protection in that sense is more extensive than copyright protection.
BTW I thought you were supposed to be the one that knew how this stuff all worked? Why am I having to explain it to you?
Anyway, patent protection is stronger protection in that sense, but at the same time it is weaker because of its limited jurisdiction. This makes sense because patent protection is SUPPOSED to apply to a physical invention. Normally sold in a single market. If you want patent protection in multiple markets, then you have to get multiple patents for the same invention, one for each jurisdiction.
There are a few of these types of patent treaties, yes, that is true. I even alluded to them. The US is trying to do more of these all the time.
That does not in any way change the fact that a US patent (for example) simply does not hold for most people on the planet.
Ergo, for most people, the fact that a given codec may have a US patent granted is simply not a problem, because most people live outside the jurisdiction of that patent regime.
You really do have a hard time admitting when you are wrong, don’t you?
Edited 2009-06-03 09:57 UTC
I have no problems admitting when I make a mistake. I do have a problem when you make an assumption with out first asking me the basis of my argument; the basis of my original argument is that if they were to ship an ASUS machine loaded with CODECs with patents it would exclude them from a number of markets unless they had agreements with the relevant patent holders. Some have gotten around this, however, by including a download site for customers. In the case of Acer they have included downloads of VLC on their ftp site – instead of bundling it with the computer itself.
That was the crux of the argument – you on the other have taken the discussion so way out of context that you’re taking about completely irrelevant issues that bear nothing to my original statement which sparked off the whole discussion thread. The first question was relating to the need to bundle CODECs, to which I explained the problems – on the basis that Asus were not willing to narrow their potential market.
This thread is about the Qualcomm Smartbook (being an ARM architecture device), running Android, and not running Windows. Your point about CODECS I guess was meant to try to raise an issue where the running of Linux (in the form of Android) would somehow present a problem.
It doesn’t present a problem at all.
If an OEM wants to bundle CODECS, then they are available for Linux. All that needs to be done is to not sell that variant of the machine in jurisdictions which honour patents on CODECS.
If an OEM wants to sell just one version of Linux machines in all jurisdictions, then they simply need to ship them without CODECS. Users in most countries can simply download and install ffmpeg codecs after purchase. Users in the jurisdictions that are afflicted by patents on CODECs can buy their own copy of the Fluendo CODECS. (This is the most common approach).
If an OEM wants to bundle CODECS, and sell to a jurisdiction where patents on said codecs apply, then these too are available to the OEM for Linux. The OEM needs simply to buy the Fluendo CODECs, add the cost to the purchase price of the machines being shipped to the relevant jurisdictions, and all is well.
None of this is any different for Windows. OEMs still need to assure that Windows machines have paid the patent royalties if said machines are going to be sold in jurisdictions where patents on CODECS apply.
Whats the difference?
Where is there any problem for Linux on ARM that you are alluding to?
I just don’t see your case, at all. Please explain it.
Edited 2009-06-03 10:28 UTC
Flash 9 runs on the N800 – ARM LINUX based.
http://en.wikipedia.org/wiki/Gnash
http://en.wikipedia.org/wiki/Swfdec
http://en.wikipedia.org/wiki/MTASC
http://en.wikipedia.org/wiki/Flash_for_Linux
All available as source code.
Even certain parts of Adobe’s own Flash code are available as open source code for free.
http://www.theregister.co.uk/2009/06/01/adobe_flash_builder_catalys…
Flash is not anyhwere near the problem that you seem to think it may be.
As for codecs … open source has a zillion of them.
http://en.wikipedia.org/wiki/Libavcodec
This is the codec library used by the best media player available, also open source:
http://en.wikipedia.org/wiki/VLC_media_player
Also not a problem. Out-of-the-box codec support on open source systems is better than that on proprietary systems.
Edited 2009-06-01 23:37 UTC
To which you ignore the issues raised by the first reply to my post – but hey, you keep replying because you like the sound of your voice – or more correctly, the appearance of your own sentences.
Which is useless because the plugin is still closed source.
As I said in response to the first paragraph, you ignore the reply to the first reply in favour of wanting to see your own text. I addressed a number of the raised – and yet you ignore that post in favour of wanting to post something already addressed.
No cookie for you, try again.
Edited 2009-06-02 04:43 UTC
Actually, with regard to CPU processing grunt, there seems to be a dual-core 1.5 GHz ARM chip now available.
http://news.prnewswire.com/ViewContent.aspx?ACCT=109&STORY=/www/sto…
http://www.theinquirer.net/inquirer/news/1050062/qualcomm-reveals-i…
Imagine combining this with some other emerging technologies:
http://www.liliputing.com/2009/05/pixel-qi-displays-impact-on-batte…
http://www.liliputing.com/2009/06/sandisk-launches-ssd-line-optimiz…
http://www.h-online.com/open/Kernel-Log-What-s-coming-in-2-6-30-Dri…
One would suddenly enable a powerful platform with more speed and capability than current Atom netbooks using only a fraction of the battery power.
Ultra-portability with no compromise on functionality.
Edited 2009-06-02 00:52 UTC
http://www.computerworld.com/action/article.do?command=viewArticleB…
Oh, and BTW, more on a (re-)emerging technology (that will be available by next year) that can really help here:
http://www.freesoftwaremagazine.com/columns/the_new_xorg_features
Going ARM for low price and long battery life mostly means going Linux. XP or W7 are absolutely out of the question, and Win CE offers no advantages, while it costs money. Other alternative exist, but are too far fetched.
So it is Linux, OK. But why Android? Android might improve fast from being just fine for smartphones to being insanely great. But is still not insanely great, and there is almost no software for it! Why not go for a full Linux distro? Why have only a web browser and a PIM (with the possibility of paying $5 for any of 1000 almost identical color matching games), when you can have that and thousands more programs for free?
Ubuntu (as a meaningful example) has an ARM version with tons of software available for download. They are even working on integrating an Android runtime package so you will even be able to run the $5 color matching games! And all for the price of a 4G flash chip, or less.
I’m sure some highly payed exec at Asus or Acer must have thought about this already. I wish I knew what do they think they are doing, because I feel they are going to crash and burn, hurting Linux and Android in the accident.
because android have gameloft games..