Fluendo To Sell Proprietary Codecs For Linux

Several readers wrote in to tell us that the open source media software development company Fluendo has announced plans to sell native Linux implementations of proprietary video codecs such as Windows Media, MPEG-2, and MPEG-4. (Press release here.) From the article: "Currently, many Linux video applications facilitate Windows Media video playback using Windows DLL files and Wine, which provides suboptimal performance, particularly with streaming video. Fluendo's codecs could potentially provide better integration for streaming Windows Media playback in Linux web browsers as well as through GStreamer-based desktop applications like Totem."

Comments

Hmmmmmmmmn,

[Whiney Mac Fanboy]

1) I don't believe (for xine at least) that wine is neccessary for asf (wmv) playback (the windows codec dlls are required, but used by xine without wine's help)

2) I guess a native binary blob is slightly better than a MS coded binary blob.... but frankly, it's still just a binary blob. You have no idea what its really doing.

Good luck to Fluendo however.

Re:Hmmmmmmmmn,

[John Betonschaar]

2) I guess a native binary blob is slightly better than a MS coded binary blob.... but frankly, it's still just a binary blob. You have no idea what its really doing.

I guess the vast majority of end-user couldn't care less what their video codec is doing, as long as it plays their damn video's. It's a bit like the NVidea Linux drivers: the free software purists see it as something awful to load a binary driver on Linux, but I for one am very grateful to have proper 3D accelerated drivers at all. Same goes for video playback... There will always be proprietary video codecs, just get over it. I don't see the problem anyway, if I'm want to run commercial software on Linux it is usually binary as well. Does that mean the software is useless or bad?

Re:Hmmmmmmmmn,

[Anonymous Coward]

1) I don't believe (for xine at least) that wine is neccessary for asf (wmv) playback (the windows codec dlls are required, but used by xine without wine's help)

Saying "wine" is an overstatement. Some code from Wine is used, the DLL loading stuff being one part, and the Windows functions that the DLL calls would also be needed. But it's far from the full wine that you would need to play WoW or run MS Office.

Re:Hmmmmmmmmn,

[Whiney Mac Fanboy]

It's a bit like the NVidea Linux drivers: the free software purists see it as something awful to load a binary driver on Linux,

Perhaps because of security fears?

"the NVIDIA Binary Graphics Driver for Linux is vulnerable to a buffer overflow that allows an attacker to run arbitrary code as root. This bug can be exploited both locally or remotely

Anyway, bringing nvidia into the discussion is a red herring, there is a huge difference between running a binary blob in ring 0 and userland. Let's discuss userland binary rather than kernel mode binary.

Re:Hmmmmmmmmn,

[timmarhy]

he is right. no one except people like RMS give a shit about the license of software, only that it does the job. besides, people can license their work how they like, why should anyone have the right to try brow beat them over it.

Re:Hmmmmmmmmn,

[Whiney Mac Fanboy]

Saying "wine" is an overstatement. Some code from Wine is used,

That makes sense. Thanks for the insight.

Re:Hmmmmmmmmn,

[Anonymous Coward]

Well, since such a huge amount of stuff is Open Source on Linux, of course it sucks to have some proprietary pieces here and there. At least for GNU being Open Source is more important than pleasing the end user, that is kind of the whole point of GNU(/Linux) you see? To make an Open Source operating system with Open Source software. If you don't care about the licenses or especially want many proprietary pieces, you could just use some other operating system.

Re:Hmmmmmmmmn,

[x2A]

I got hit by the file corruption bug in that's existed in recent 2.6 kernels due to race condition. Was burried pretty deep, took the experts a while to figure out what was going on. I had the source code, so did thousands of other people. The bug still caused lost files. The nvidia driver's a pretty complex piece of code, having to handle many slight differences and implement workarounds for many different cards and chipsets. Nvidia have paid people on the job, with the relevant experience. What makes people think that the oss community can do a better job than nvidia's own people, when they can't even keep their own codebases bugfree? Bugs happen, and with really complex code, it takes people with the most experience available to find and resolve the problem, properly.

Re:Hmmmmmmmmn,

[pipatron]

Because it's better for the public good if the drivers are open.

Re:Hmmmmmmmmn,

[Anonymous Coward]

binary blob

'blob' usually is an acronym for Binary Large OBject, so the 'binary' modifier is redundant.

Re:Hmmmmmmmmn,

[Anonymous Coward]

no one except people like RMS give a shit about the license of software

I give a shit and I'm not RMS. People who don't give a shit just have no concept of past or future in the software world.

Re:Hmmmmmmmmn,

[TheOrquithVagrant]

"Binary blobs" are only a problem in kernel space. Userspace binary blobs are acceptable. Fence them in with chroot and/or SElinux policies if you're feeling paranoid, and if you find them doing anything suspicious, just strace/ltrace/gdb the bastards. There are plenty of ways to "get an idea what it's doing".

Re:Hmmmmmmmmn,

[doob]

Perhaps because of security fears?

"the NVIDIA Binary Graphics Driver for Linux is vulnerable to a buffer overflow that allows an attacker to run arbitrary code as root. This bug can be exploited both locally or remotely

You say that as if it were a current problem. This has actually been fixed in the last 3(4?) driver revisions, including a bugfix only release to a previous branch of the drivers.

Re:Hmmmmmmmmn,

[Anonymous Coward]

ok, and what happens when you get 2 blobs, say a video driver and a video codec (handy example) that when used together under certain circumstances hose the system, being from different vendors and with neither accepting responsibility, your stuck, you have re-implement them both in F/OSS.

for this reason amongst others, open source is well worth being evangelical about, and it must be done NOW, with everything on the system.

a/c because im at work =(

Re:Hmmmmmmmmn,

[Nadir]

1) I don't believe (for xine at least) that wine is neccessary for asf (wmv) playback (the windows codec dlls are required, but used by xine without wine's help) All of the various linux players use a modified version of the wine dll loader.

Re:Hmmmmmmmmn,

[arivanov]

And this is the exact reason why codecs are currently the most popular vehicle for Trojan deployment and torjan browser helper DLLs are a thing of the past.

Re:Hmmmmmmmmn,

[Short Circuit]

Certain videos are usable under Linux which aren't under Windows. Why? Because commands are embedded in them that open links to websites and such.

I was surprised (and an embarrassed) when I tried to show an AMV to a friend, and it popped up a link to a porno website on his computer. (Note to self...don't grab AMVs off of Gnutella.)

This seems to be an issue with WMVs. I haven't noticed it with other container formats.

Re:Hmmmmmmmmn,

[Anonymous Coward]

He didn't say just RMS, he said people like RMS, ie people with integrity, and a strong sense of good judgement about the dangers of closed systems and information. ie people who are neither ignorant, nor idiotic.

Re:Hmmmmmmmmn,

[blackest_k]

There are two principle reasons to care.
first
Security, windows biggest problem is trusting binaries not to rip a hole in your security model, why would you want to reduce linux to that level.

Second
to do it better, you can't demand perfection from a commercial programmer but someone will want to tweak a CODEC for maximum performance on his or her hardware.

Re:Hmmmmmmmmn,

[Short Circuit]

I care. While I'm not likely to check the code for trojans and security flaws, I am likely to tinker with it.

Sometimes I need a feature in software that's not already there. Other times, I need to tinker with some hardcoded value or behavior. (Like the time I needed to modify wget to get around a broken robots.txt.) One time, I wanted to use xvidcap, but found that the latest version of the code was old enough as to not compile on a modern GCC. (I'd be happy to release my updated version, if anyone cares.)

Re:Hmmmmmmmmn,

[Anonymous Coward]

And people who live very nicely off grants from Colleges who look at the world through emerald colored glasses and have never had to work in the private sector.

Re:Hmmmmmmmmn,

[Anonymous Coward]

Nvidia have paid people on the job, with the relevant experience.

That's pure speculation. Few if any people outside Nvidia know exactly what the problem was, what was changed to fix it, which people did the job, or if they even have "relevant experience".

Re:Hmmmmmmmmn,

[uhmmmm]

Xine is probably like MPlayer, and every other linux media player I've heard of that can use windows DLLs for playback - it uses a DLL loader which was forked from avidec at some point, which was itself at some point forked from wine. So while it doesn't use wine directly, and the code's probably mutated along the way to the point where it may no longer resemble the original wine code, it is, in a sense, using wine to load the DLLs.

Re:Hmmmmmmmmn,

[Anonymous Coward]

a/c because im at work =(

a/c because you're at your MOM.

But really, I agree that people should be evangelical about OSS. I'm still going to use binary drivers if they work better and OSS if I can. I just recently switched my studio to Windows, because keeping everything running under Linux was a pain. Different distros had their problems, and I just want a professional recording app to make it all better.

Still, I salute you.

Re:Hmmmmmmmmn,

[Ph33r th3 g(O)at]

Get Media Player Classic, and you can avoid most of the spyware properties associated with WMVs. Otherwise, you can turn those features off in WMP setup, but that would also involve trusting Microsoft to unconditionally honor those settings, which I would not.

Re:Hmmmmmmmmn,

[FictionPimp]

I'd pay for fully supported video codec's open source or not. I just want my crap to work, preferably in 64bit. The only thing currently keeping me in 32bit is the lack of video codecs and flash in 64bit. I love open source, but I have no problem with binary blobs. Some of my favorite applications are closed source. I would prefer open source, but I am not going to let my productivity suffer in exchange for my ideals.

Re:Hmmmmmmmmn,

[Just Some Guy]

So hundreds of developers (and being kernel hackers, probably paid) took a long time to find a file corruption bug that would presumably affect a lot of people. And somehow you think that NVidia's little graphics driver team should be able to magically resolve your problems overnight?

Re:Hmmmmmmmmn,

[Just Some Guy]

You say that as if it were a current problem. This has actually been fixed in the last 3(4?) driver revisions, including a bugfix only release to a previous branch of the drivers.

Yeah. If only the other changes included in that revision didn't cause my FreeBSD machine to hang so that I could use the fixed driver, I'd be set.

Re:Hmmmmmmmmn,

[Kjella]

Well, normal users are always more interested in what the code can do today, rather than what it could do tomorrow. Strangely there's a very good corrolation between what the normal users thinks of it tomorrow, and what the coders are able to do in the meantime. We've already tried a system full of binary blobs, it's called Windows or OS X. A few like nVidia are tolerated because a partially free machine is better than a useless machine since there's no real choice but in the long run fewer, not more blobs are the answer.

Re:Hmmmmmmmmn,

[MartinG]

2) I guess a native binary blob is slightly better than a MS coded binary blob.... but frankly, it's still just a binary blob. You have no idea what its really doing.

A few things:

a) It is infinitely better than a windows dll because it works (my machine is x86_64. windows blobs won't work there)
b) I don't have windows. How can I get windows binary blobs without buying windows or breaking the law?
c) True, I don't know what it's really doing, but it comes down to trust. I have personally met spoken to some of the fluendo/gstreamer folks and I trust them a hell of a lot more than some unknown devs at MS who I'll never even know the names of let alone meet and talk to.

Re:Hmmmmmmmmn,

[m50d]

It's a bit like the NVidea Linux drivers: the free software purists see it as something awful to load a binary driver on Linux, but I for one am very grateful to have proper 3D accelerated drivers at all. Same goes for video playback... There will always be proprietary video codecs, just get over it. I don't see the problem anyway, if I'm want to run commercial software on Linux it is usually binary as well. Does that mean the software is useless or bad?

But what's the advantage over the current approach? The free software people won't prefer it because it's still just a binary blob. The whatever works people won't care because the current approach works and is cheaper. So I really don't think there's a market (if they release it in binary form, I'll bet we won't even get non-x86/amd64 architecture support).

Re:Hmmmmmmmmn,

[tolan-b]

> 1) I don't believe (for xine at least) that wine is neccessary for asf (wmv) playback (the windows codec dlls are
> required, but used by xine without wine's help)
>
> 2) I guess a native binary blob is slightly better than a MS coded binary blob.... but frankly, it's still just a
> binary blob. You have no idea what its really doing.

The MS coded blob is illegal to distribute in the format it usually is (w32 codecs pack) in many (most?) countries. So while this Fluendo deal isn't an ideal solution, it is better than the current one for open source operating systems, at least it means everyone can view these formats without breaking the law or having to buy a Windows license. ;)

I believe ffmpeg have experimental wmv3 support though (xine doesn't support them without the windows dlls loaded through lib wine).

Re:Hmmmmmmmmn,

[Short Circuit]

Meh. The problem's not on my computer. I only run Linux. :)

Re:Hmmmmmmmmn,

[Ph33r th3 g(O)at]

Noted -- don't know if WMP runs under WINE :).

Re:Hmmmmmmmmn,

[RossyB]

> 1) I don't believe (for xine at least) that wine is neccessary for asf (wmv) playback (the windows codec > dlls are required, but used by xine without wine's help)

Aside from the fact that distributing the DLLs is legally dubious at best, good luck using the Windows DLLs on a x86-64, PPC, Sparc, ARM, or MIPS Linux system...

Note how some of the fluendo plugins come in x86-32, x86-64, sparc and ppc versions.

Re:Hmmmmmmmmn,

[piranha(jpl)]

NVidia sat on their hands for years with that security problem, didn't they?

Re:Hmmmmmmmmn,

[x2A]

Noooo, missed my point - there may be those hundreds of developers that work on the kernel, but it took the very select few, that had the proper experience and knowledge of the code and design, to track it down and fix it. The team of kernel developers could have been 5000, or it could have been 5. Sometimes, it's not the hands and the eyes, it's the right hands and the right eyes. Experience is very important in a complex codebase.

Re:Hmmmmmmmmn,

[x2A]

Well... if they fixed it, they must have had the relevent experience.

Re:Hmmmmmmmmn,

[Anonymous Coward]

You don't know what the hell they did to it. You don't know whether they effectively fixed it, or broke something else worse.

Re:Hmmmmmmmmn,

[Ded Bob]

Does this patch in the FreeBSD nVidia forum fix it for you? It helped me on a 6-STABLE system with the 9631 driver even if the patch said it was for 7-CURRENT and the 9746 driver. I recommend it.

Re:Hmmmmmmmmn,

[x2A]

OMG!!! They could have been the one's who put it there deliberately in the first place!!! We all know that nvidia works with the government... they've probably put backdoors in the chipsets too! Damn I gonna rip out my nvidia graphics cards and not put them in until I have personally verified each line of code, and every single transister on the chip schematics, and every single transistor and wire and line of code that controls the machines that make them... THEN I'll know I'm safe. It's a good job I have as much experience with GPU design as nvidia and ati, as that means I'll definitely find whatever's there.

Re:Hmmmmmmmmn,

[Just Some Guy]

Thanks for the tip. I'll give that a shot later tonight.

Re:Hmmmmmmmmn,

[h2g2bob]

FFMpeg does MPEG1/2, MPEG4 (aac, divx, etc), WMV1/2/3, H.263 (flv6), VP6 (flv8), and frankly god knows what else.

Who in there right mind would think that releasing proprietory codecs for exactly the same to linux to do the same thing is in any way a good business plan!? If it was encoding these video formats, there may be some merit (eg: better encoding techniques), but playback!? The only possible advantage is it can be used in non-GPL'd programs.

Re:Hmmmmmmmmn,

[Lussarn]

no one except people like RMS give a shit about the license of software

Maybe you can explain why there are licenses on software if nobody except RMS cares about them?

Re:Hmmmmmmmmn,

[Anonymous Coward]

No, they didn't. There was some confusion about what issue was fixed with what fix. People concluded it took years, while in fact they were confusing two different issues.

Re:Hmmmmmmmmn,

[k8to]

Agreed, ffmpeg has the ability to decode these same formats, and on alternate arches to boot. The difference is that the use of ffmpeg may not be legal in your jurisdiction due to the lack of patent licenses. This is most likely to matter to, say, broadcasters.

Personally, I'm not very excited about paying money for essentially patent licenses. I supposed I'm resigned to being a patent license transgressor rather than monetarily supporting the patent holders. (Of course I also have the option of eschewing the content entirely, and mostly (but not entirely) do.)

Re:Hmmmmmmmmn,

[frank_adrian314159]

... there is a huge difference between running a binary blob in ring 0 and userland. Let's discuss userland binary rather than kernel mode binary.

Which is why, if Linux and Windows were microkernel-based, there would not be as large a problem. Don't blame the market for doing what it has to do because of technological deficiencies.

Re:Hmmmmmmmmn,

[Anonymous Coward]

I also suppose you dislike the new automated atm machines that use personal pin numbers huh?

Re:Hmmmmmmmmn,

[caudron]

I guess the vast majority of end-user couldn't care less what their video codec is doing

In reading the parent post, I didn't get the impresion he was posting as the "majority of end-users" but rather offering his perspective, which might different substantively from the majority opinion. You're right that the majority won't care, but we needn't squash the minority opinion either. Disclaimer: I happen to agree with him in principle, though I do, somewhat hypocritically, run binary video drivers on my system.

Does that mean the software is useless or bad?

Useless? Maybe. Bad? We'll never know. It's closed source. We just take it largely on faith in the company from which we receive the binary that the driver doesn't do anything bad.

just get over it.

We don't have to get over our opinion any more than you do. We can keep our opinion and voice it and if it is worth hearing, then at least on forums like slashdot, there is a system in place for making sure it gets heard. That's what happened with the parent post. Enough people found it worth hearing that it got modded up, like your post...and probably not like this post here. ;-)

Don't take this reply as combative, rather as clarifying. I don't think your point is invalid or even wrong, just that it's not the final word either. When you say things like "get over it" you come across (perhaps wrongly) as appearing like you can't understand the opposing view, which is deadly to honest discussion.

Tom Caudron
http://tom.digitalelite.com/

Re:Hmmmmmmmmn,

[xenocide2]

I think you over estimate the number of people involved in the Linux / BSD nvidia driver. If five people in the kernel can find and fix bugs, that's probably doubling nvidia's bug hunt task force. Not to mention the other valuable tasks that can be done with it, like fixing suspend / resume, or automated source code analysis tools like Coverity. At this point, nVidia shouldn't be worried that if they open the drivers, nobody will help and find bugs. They should be worried about what happens when people DO find bugs. Apparently bugs in nvidia drivers go back for years; I imagine that at least one vendor or render farm has negotiated some level of support that requires backporting security fixes. That sort of work is not easily outsourced to the community.

Re:Hmmmmmmmmn,

[x2A]

"I think you over estimate the number of people involved in the Linux / BSD nvidia driver"

No I wasn't trying to make any assertions as to number of people involved, I was just using the numbers 5000 and 5 in regards to the kernel to demonstrate the idea that having loads of manpower involved in something doesn't mean that they can all solve these various problems, there may only be 1 in 1000 that have the experience and knowledge of internal design etc required to fix it. The linux kernel is huge, and bugs tends to be fixed by people specific to the particular subsystem... Linus may fix an issue with the VM subsystem, but usually won't be the person to fix a problem with ext2 - that's better handled by the ext2 team. In the same way, what I'm saying, is that the nvidia driver developers are going to be the people best for fixing problems with the nvidia driver, and openning it up to thousands of other people is as likely to change that, as the thousands of people who have access to the linux kernel has changed that Linus* is the best person for fixing problems deep routed in his VM subsys.

(*no, it may not be linus who's the best person, and there maybe a few others who are on equal enough footing, I haven't followed it enough to say, but it works as an example to demonstrate my point nonetheless)

Re:Hmmmmmmmmn,

[level_headed_midwest]

I don't care where is comes from, as long as:

1. It works well.
2. It will install into my kernel.
3. It is updated and maintained to work with reasonably current libraries/kernels/etc.
4. It is always offered as a download somewhere- the driver can't just dissapear from the net someday, leaving me with a paperweight.

Of course, making binary drivers is not preferable to real open-sourced ones as the OSS ones are in the kernel tree and can't be taken down some day. but when faced with no Linux support or a binary blob...I'll take the blob any day. I have no less than 3 blobs in my system as we speak: one is my ATi display drivers (can't complain much, except for the known AMD64 XVideo X11 segfault bug) another is VMware's kernel modules (with the any-any-update, they work fine) and the last is HighPoint's binary drivers for my RocketRAID 2310 drive controller (actually patches into the kernel menuconfig, even.) I can't complain too much.

Re:Hmmmmmmmmn,

[10101001 10101001]

Nvidia have paid people on the job, with the relevant experience. What makes people think that the oss community can do a better job than nvidia's own people, when they can't even keep their own codebases bugfree? Bugs happen, and with really complex code, it takes people with the most experience available to find and resolve the problem, properly.

You highlight one of the reasons why I'm a member of the Free Software movement instead of the Open Source Software movement. That is, the OSS movement tries to present an unrealistic and unprovable supposition: that open source software is invariable "better" (ie, more bug free) than closed source software. It reminds me a lot of capitalists who try to equate an idealized model (free market) with the real world as a basis to ignore obvious examples where the idealized model doesn't work.

Back to the discussion, the real reason why those of us in the Free Software movement want access to the source code, beyond any moral reasons, are a few pragmatic ones. NVidia may not be around forever. Even if they are, they won't necessarily release updates in a timely fashion or for the hardware/software that I use. Once a bug is found, cutting off the ease of anyone being able to fix it, by making any fix a result of binary blob hacking, does in no way improve the odds of it being fixed. Further, the fact that the source is closed is a prime reason *why* NVidia employees are the only ones so well familiar with their driver. Given enough people in the world, especially those willing to hire others to write software for them, it stands to reason that the source being available would very likely increase the number of people who are familiar with what were "binary blobs".

Will this make source code magically less buggy? No more than making source closed has made it less buggy. But the source code being Free at least offers the potential for people themselves or to pay others to fix bugs and for the advantage to pass to everyone. And personally I like being in a position where I have a reasonable chance of fixing known problems even if some company or other specific person isn't interested. Certainly, software being Free only helps.

Re:Hmmmmmmmmn,

[ChrisJones]

It's also very important to distros. You won't find ffmpeg with all the patent-problem codecs in any sane distro, but you can drop the fluendo codecs into just about any distro and every gstreamer application can use them. Ok so many of the distros still won't be able to ship them out of the box, but the commercial ones will have the choice and users of the Free-er ones can choose to bolt them on later.
I don't at all think it is ideal, but the simple facts of reality are that those formats do exist, users do want to view the media that's encoded in them and not all of those users are prepared to do legally questionable things to do that.
I have already bought the full codec pack from Fluendo for two reasons - firstly because I want to have decent, supported codecs (read the README for the win32codec gstreamer plugin - pitfdll - the author confesses that he has no idea how the dll loader works, so I really hope there aren't any bugs in there), secondly I really like the work Fluendo are doing and I want to help support them improve gstreamer. Just like I buy Crossover licences every few years to say thanks to Codeweavers for the great work they have done on WINE. On the other hand I don't give my money to Transgaming because they don't tend to share their work very much. Fluendo release a fair amount of code (and obviously they can't release the codec source for some of these, but look at their mp3 plugin - it's free as in beer and you can have the source, you just can't distribute binaries without an mp3 licence).

Re:Hmmmmmmmmn,

[x2A]

There can be huge advantages to having the source code, there's no denying that. If you're building your business on a software package that manages your accounts, stock etc, having access to the source code means your business isn't dependent on the people who wrote it sticking around and maintaining it. But 3D accelleration driver for a high end graphics card? It just doesn't cut it in the same way. Are you going to have a business tied into the functions provided on that nvidia graphics card? Are you going to be upgrading the rest of the systems to the extent where current drivers no longer work, while keeping the same old graphics card? There really isn't the same level of tie in. Yes there could be the annoying aspect of it, if your operating system requires updates that will stop the driver working (linux, I'm looking at you) - but a fair amount of this can be solved in the code we do have access to (the kernel, the module stub/shim layer). If NVidia die within a year of buying a new graphics card, that's pretty unlucky, but anything over that, you'll be able to replace your graphics card with a similarly powerful one without having to fork out too much at all. A pain, but nothing too tragic. And with Windows backward compatability what it is, you can be pretty confident that todays drivers will see you through the next 5 years. Windows 2003 will still load drivers designed for Win98 (and pos 95 (VxDs) but not sure). I just can't see, in this case, it being that big a problem.

But on the other hand, look from NVidia's point of view. They're state of the art leaders, in constant battle with ATI for the best and the fastest GPUs, they gotta keep some things close to their chest. They also gotta protect their brand. If a driver fork doesn't perform quite as well (which can be for many reasons, eg, if they've licensed a compiler that can optimise better than the standard GCC everyone'll be using), benchmark figures could reflect badly on their product. And, at on a smaller level, they get to do things like flash their latest and greatest at people who come to their website's front page looking for drivers.

If things change, then we can reevaluate, but as things are at the moment, we have little to gain, but they have a lot more to lose, they keep on top of kernel changes fast enough, and the drivers seem to do their job just fine.

Re:Hmmmmmmmmn,

[10101001 10101001]

But 3D accelleration driver for a high end graphics card?

What makes you think I have a high-end graphics card? Or that I should only concern myself with the costs of something when it involves running a company?

If NVidia die within a year of buying a new graphics card, that's pretty unlucky, but anything over that, you'll be able to replace your graphics card with a similarly powerful one without having to fork out too much at all. A pain, but nothing too tragic. And with Windows backward compatability what it is, you can be pretty confident that todays drivers will see you through the next 5 years.

No, it's not "too tragic" to have to fork out $100 every 5 years just to keep a computer system running. But the question is, will I actually see drivers that will see me through the next 5 years? And just how smart is it to feed into a system of consumerism that suggests that I just buy a new board every so often to work around design flaws of a driver or hardware?

Windows 2003 will still load drivers designed for Win98 (and pos 95 (VxDs) but not sure). I just can't see, in this case, it being that big a problem.

If you hadn't noticed, the original discussion was about an nvidia driver that had a locally/remotely exploitable security flaw in it. So, being able to run *known* insecure drivers isn't exactly a great example of good security. That was the major reason for the complaint.

But on the other hand, look from NVidia's point of view. They're state of the art leaders, in constant battle with ATI for the best and the fastest GPUs, they gotta keep some things close to their chest. They also gotta protect their brand. If a driver fork doesn't perform quite as well (which can be for many reasons, eg, if they've licensed a compiler that can optimise better than the standard GCC everyone'll be using), benchmark figures could reflect badly on their product. And, at on a smaller level, they get to do things like flash their latest and greatest at people who come to their website's front page looking for drivers.

Ever heard of a magical word called Trademark? There's nothing stopping NVidia from trademarking their driver (both ATI and NVidia already seem to have) to prevent most of the badness you make out. More to the point, there's nothing stopping someone *right now* from making a crappy nvidia driver and pushing it into Linux distros precisely because it's the only legal option at the moment. If NVidia's driver is so good, do you really think a crappy reverse engineered one will "smear" their name less than a badly optimized one that someone else compiled from the official source? And I agree, it is at a very small level that nvidia or ati could use their websites as a marketing tool when a person is looking for security-update drivers.

If things change, then we can reevaluate, but as things are at the moment, we have little to gain, but they have a lot more to lose, they keep on top of kernel changes fast enough, and the drivers seem to do their job just fine.

I already reevaluated, after having more than I desired in crashes with my NVidia card. This was especially the case when it became clear I couldn't get help from Linux or Xorg developers for problems I was having and I'd have to rely solely on NVidia to fix my problems. Given the probable complex interactions involved and the simple fact that NVidia was clearly not so invested in Linux (again, they haven't gone out of their way to ensure that their driver can be legally included with distros), it was something of a futile condition to be in. I've already made my decision. The point of my position was more to express my reasoning than as some sort of mantra that no one should buy from NVidia. But certainly, I can't fully reason on the logic behind why I should choose NVidia over ATI or the reverse, at least based on company actions. I'll just have to go with the reverse-engineered drivers that exist for ATI instead of having to deal with such petty squabbles.

Re:Hmmmmmmmmn,

[k8to]

I agree with you on the distro point, although fluendo is _not_ making these available in a distro-friendly way. They're binaries which will incur breakages over the life of a distribution. If Fluendo were to say, partner with a company, such as Linspire, then this could be addressed, but currently it is not addressed.

As for comparisons with win32codec solutions, sure. Win32codec solutions are hacks, aren't portable, and are scary. I was talking about ffmpeg which has already created native implementations of these codecs in a clean-room way. At least, that's what they claim.

Re:Hmmmmmmmmn,

[evilviper]

1) I don't believe (for xine at least) that wine is neccessary for asf (wmv) playback (the windows codec dlls are required, but used by xine without wine's help)

You're wrong.

Linux programs aren't magically able to use Win32 DLLs. They are able to use them (AT ALL) because of Wine. Every time you load a Win32 codec, you're using wine code.

Quite some time ago, just the necessary pieces of Wine (needed to load DLLs) were copied, and integrated with MPlayer, Xine, etc. In MPlayer, this is in the loader/ folder, and it's easy to tell that it is wine code (with significant modifications).

You don't need to have the full Wine package installed (seperatly), but you're still using it.

Re:Hmmmmmmmmn,

[evilviper]

(my machine is x86_64. windows blobs won't work there)

Yes they will.

b) I don't have windows. How can I get windows binary blobs without buying windows or breaking the law?

You can freely download the codecs from Microsoft.com (that's how they were aquired in the first place). You may be violating their EULA, but whether that license is legally binding or not is a legal grey area. Claiming it's illegal is just ignorant.

Re:Hmmmmmmmmn,

[MartinG]

Claiming it's illegal is just ignorant.

I never claimed it was illegal. I implied I didn't know how to get the codecs without breaking the law. Perhaps you have partly answered my question. Can you show me where to freely download them from Microsoft in a format that will work with gstreamer on x86_64 linux?

Re:Hmmmmmmmmn,

[ChrisJones]

I fully expect to see Linspire licence these for inclusion if it is at all possible - they already include a bunch of codecs.
Binary breakage will only happen if gstreamer changes its ABI during the 0.10 phase (which is entirely possible and I don't think they have made any hard commitments to keep it completely ABI stable)
I don't dispute that ffmpeg's implementation is a clean-room reverse engineering, but that doesn't necessarily make its IP legality different, I think.

Re:Hmmmmmmmmn,

[Anonymous Coward]

Perhaps because of security fears? You say that as if it were a current problem. [snip] No, he didn't. You interpreted what he said as such. Fact remains: With the binary blob driver, it's much harder for you to check for security problems, and much harder for you to fix them, should you be so inclined. Whether you have the inclination or knowledge to do so, is an entirely separate issue.

Re:Hmmmmmmmmn,

[Anonymous Coward]

emerge amd64codecs

(n.b. I actually use mplayer but I assume they work with gstreamer)

Re:Hmmmmmmmmn,

[JohnnyLongbaugh]

"There will always be proprietary video codecs, just get over it."

Yes, and there will always be corrupt governments and unethical, immoral big businesses (i.e. M$) that are only in it for the monetary gain and not for the common good of the people. Does that mean that we should just get over it as far as they're concerned?
    I realize that this isn't exactly in the same ballpark with your comment, but that very attitude is a part of what is bringing down our society and what we used to value as our way of life. It is the complacency and the attitude that "Well, it's going to happen anyway. We should just get used to the idea." that's hurting us.
    That kind of thinking is what eventually lead to our Revolution;
we got sick and tired of the out-of-touch, big-headed, greedy, self-righteous, arrogant douche-bags telling us what to do and how to do it.
    This country and our way of life has been steadily heading right into the proverbial crapper because far too many people ignorantly and/or lazily, simply go-along with what they're told by the media and the big-businesses that control what we do, how we think and what the government's next step should be, is the best way to act/think/behave.
    One only needs to examine the various versions of our American dictionary (and for those of you that believe we should add more Spanish to the book that helps define our language and the way we communicate, just remember that all that that book holds originally comes from the Queen's English) to realize that our language continues to degrade simply due to laziness and the unwillingness of the masses to conform to what we used to think of as "proper". Have you read any books that were written over a hundred years ago? Can't you see what a huge difference there is between then and now in the spelling of common, every-day words and the changes our grammar has taken? Have you read of the numerous cases where many among us believe that Ebonics should be taught along-side proper English, because they believe that this (Ebonics) lackadaisical approach to our language should actually be considered an official dialect and that teaching young African-Americans English alone is confusing for them? Ebonics is nothing more that a form of slang, or "jive" and stems from those too lazy to teach their children proper English.
    But pardon me, I digress. The entire reason for this rant is that your attitude of "...just get over it" simply will not do if our culture is to survive. If however, you are willing to eschew the ideals this once-great country was founded on, if you will allow those that want to see us fall do as they wish, if you no longer care about the very beliefs that made this country what it was, then by all means, allow M$ to continue in it's quest for total dominance.

Re:Hmmmmmmmmn,

[xenocide2]

"Linus may fix an issue with the VM subsystem, but usually won't be the person to fix a problem with ext2"

It's kinda interesting that you bring this up, becuase he recently went through and fixed a data loss problem caused by interaction between VM and ext2. Just because someone else is the maintainer of it doesn't mean he can't go through and debug it like anyone else. But I think that there are plenty of people who would like to go through the nvidia source code. They may not all be experts, but I have a sneaking suspcion that the nvidia team isn't either. At first, you'll probably see a few code critiques, during which the module maintainers may be exposed to "kernel best practices" rants. But for all their internal knowledge, there's likely enough documentation in the drivers alone to more than double the efforts at improving nvidia drivers after 6 to 12 months. For example, the nouveau project is already making progress towards an open source driver for nvidia 3d chipsets, without anything more than the binary drivers, cards, and helpful users. Even if a tragically small percentage of OSS developers are able to help nouveau, it's still more than nvidia has dedicated to it.

*makes popcorn and pops open a beer*

[Anonymous Coward]

and settles back for a nice round of productive but gentlemanly discussion over the need for massive deployment of closed source modules in Linux.

Re:*makes popcorn and pops open a beer*

[RotateLeftByte]

At 07:48 In the morning? Can I have some of what you are smoking!

Oh, well, back to bed. :-)

Re:*makes popcorn and pops open a beer*

[x2A]

Vmware's always done pretty well...

Re:*makes popcorn and pops open a beer*

[ray-auch]

Also available in other timezones...

Re:*makes popcorn and pops open a beer*

[x2A]

Not offtopic - vmware's a very successful, closed source solution, that runs on linux. There is now a free version, but before that it was sold. It's probably the best tool in its class, and really bought virtualisation forward and out into the public. It inserts binary blobs into your kernel ala nvidia. Yet it's probably the most accepted closed-source-on-linux software out there, yet so forgotten about during these discussions. Inconvenient to have an example of closed-source-on-linux that... worked?

Does it not prove that it can?

Correction:

[geminidomino]

"Fluendo to TRY to sell proprietary codecs for Linux."

Look at all the flak NVidia's binary-only drivers take from the GNU-types, and those are FREE.

Re:Correction:

[Whiney Mac Fanboy]

Look at all the flak NVidia's binary-only drivers take from the GNU-types, and those are FREE.

And why did they take all that flak? Perhaps it was because of the security implications of running a binary kernel module? Not an unrealistic fear either:

The NVIDIA Binary Graphics Driver for Linux is vulnerable to a buffer overflow that allows an attacker to run arbitrary code as root. This bug can be exploited both locally or remotely

PS. Love the way you capitalized FREE! That'll get the "GNU-types" worked up. Seriously - nice trolling.

Re:Correction:

[replicant108]

Look at all the flak NVidia's binary-only drivers take from the GNU-types, and those are FREE.

They might be FREE, but they aren't Free.

In other words - you will pay for using them, just not with money.

How much you pay depends on how much you value your freedom.

Re:Correction:

[Anonymous Coward]

"Fluendo to TRY to sell proprietary codecs for Linux."

Look at all the flak NVidia's binary-only drivers take from the GNU-types, and those are FREE.

Yeah, we really don't want more binary kernel modules to make the system crash and impossible to debug.

Wait, who says that these video codecs are going to be kernel modules like the nVidia drivers, and not userspace programs like the Quake3, Doom3 and UT200X that we all love?

Re:Correction:

[x2A]

I can choose between open and close source drivers for nvidia... feels pretty free to me.

I can write software and choose to release it open or closed source... that feels pretty free to me.

Erm... nvidia can too.

"Freedom to disagree" anyone? Oh no... it's YOUR way only, that's freedom!

Re:Correction:

[x2A]

It's a good job there's never been any security flaws in any open source software, otherwise that'd make your argument look really stupid!

Re:Correction:

[arivanov]

You are comparing apples and oranges. Wrong comparison.

Nvidia is taking flak because the sole reason for the "binary" its own asinine behaviour (same as ATI with the newer Radeon ATI-supplied drivers).

Fluendo intends to provide a service by implementing specs for which the originating party requires a licensing fee.

So the right comparison is not to Nvidia (or ATI for that matter), but to Digium. Digium provides a machine tied (oh what a sacrilege) closed source (oh what a crime) implementation of the g729a codec (oh...). The requirement for licensing it is set by the parties who wrote the spec initially and as most ITU specs it is not royalty free. You have to pay and Digium is the only linux company providing that service. And guess what - people buy it. In large quantities actually as it is essential for interoperability with other systems out there.

So unless they f*** up their business model it will work and will be considered a valuable service to the linux/bsd running public (nothing like Nvidia/ATI asinine in-kernel graphic stunts).

Re:Correction:

[Short Circuit]

Everyone else is irritated that NVidia's drivers are closed source. I'm just happy advanced native Linux drivers exist. Back when I had a 3D accelerator, ATI's and NVidia's drivers were considered a blessing by most.

Funny how people tend to complain once they learn to expect something.

Re:Correction:

[davek]

BINGO. The whole point of the operation is freedom, IMHO. That includes the freedom to choose to greedily keep my knowledge to myself in vain hopes that it will somehow help me in the future.

Now, when they find it impossible to support a moving target without interaction with OSS communities, and their binary blobs stop working after 2 months, THEN I'll say "I told you so" and hopefully they'll open their minds a bit.

-dave

Re:Correction:

[agurkan]

OK, am I "free" to have slaves? Would that improve "freedom"? Not that software and human rights are equally important of course, but your reasoning is not logical.

Re:Correction:

[x2A]

"am I "free" to have slaves?"

Should you be allowed to force people to do things against their will, through threat of violence and intimidation? Are you really asking me that?

"Not that software and human rights are equally important of course"

Are you saying that nvidia not giving you something that they wrote themselves, is in some way comparable, to the previously mentioned violence and intimidation to force people into work? Because dude, not only are they not on the same page, they're not even in the same book.

"but your reasoning is not logical"

You made some irrelevant question about slavery, then use it in a sentence about software, as if you were trying to make some comparison, and then question my logic?!!

Wow.

Re:Correction:

[msuarezalvarez]

Apart from trivial applications, mostly all software is going to have security flaws, as perfection is way too hard and way too expensive to achieve. The difference is how you react to flaws.

Of course, you knew that.

Re:Correction:

[x2A]

No I didn't know that, I, from reading slashdot, was under the impression that nvidia had this unique thing called "security flaw" that doesn't exist in the OSS world, but now you're telling me that what actually matters is whether it gets fixed, compiled, and rolled out, or fixed, rolled out, then compiled?

I think what all this really comes down to is something called "being a control freak". How many people here have even looked at 1% of the source that's behind a standard distribution? Would you have gone through the nvidia driver source code, found, and repaired the bug? Yeah I know, it's whether you would do if you had the source, it's that you can't because you don't... just don't like being told "no, you can't have that", it's unfair, and goddamnit, you're gonna sulk, and keep sulking all over the place, loudly, until you get your own way.

Wow. What a grown up place this is.

Re:Correction:

[Hamoohead]

[soapbox]

Your confusing the GPL/FOSS definition of "free" with the "I didn't have to pay for this" definition. The former refers to distribution of source code, as in "free to modify and distribute". Just because you didn't have to pay for it does not make it free in the FOSS (Free Open Source Software) world.

[/soapbox]

Re:Correction:

[msuarezalvarez]

Well, I do tend to go look in the source of things that misbehave when I use them, and fix them, and post patches. I thus ended up maintaining, comaintaining and helping out with pieces of rather high visibility OSS software, which essentially all distributions carry and put in their main desktop menus. I would probably not be of much help in fixing a video driver, as I have absolutely none of the required knowledge; but I know for a fact that there are people that have it and would certainly be of much help. People have been able to come up with sort-of-working drivers -without- access to specs, source code or even a mildly encouraging look from nVidia and friends, so I am not being bold at all at thinking that they would pretty much be of use in more favourable contexts.

In any case, if that is the impression you got from reading Slashdot, I suggest you be more careful in selecting your sources of information. Even within Slashdot some judicious filtering would enourmously help with your understanding.

Re:Correction:

[x2A]

"if that is the impression you got from reading Slashdot..."

It's the impression I get of the majority opinion on here, I, as you can tell, have my own differing opinion (and sarcasm). My opinion is that people should be entitled to keep their source code to themselves if that is their decision, without having to deal with this constant whining across the internet by spoilt brats who think they have a god given right to somebody elses work, just because they want it. If you have any young children, just as they're developing their rebelious streak, you'll know exactly what I mean. They'll say anything to try and justify why they should be getting something they want, taking any reason they can find or make up to fit what they've already decided should happen. All it's doing is creating an image of the OSS community and its users, of a bunch of spoilt stroppy adolescents who can't deal with things in the real world. Like it or not, there are very good reasons why ATI and NVidia won't open up their code. So what if Intel have done? They're not the one in the fierce competition pushing the state of the art as ATI and NVidia are, they don't have the same thing to protect. And NVidia isn't going to want to open up drivers, and potentially risk damage to their name and brand due to dodgy patched drivers that hurt performance of their products.

So nvidia drivers had a bug, but you can throw 100 additional people at it, and you'll still get the occasional bug get through, you always do. But the drivers do actually work just fine, I've never personally had problems with them on multiple machines running linux over the past few years, upgrading kernels from linus' tree regularly enough. There comes a point when people are just looking for problems, making up hypothetical ones, to justify their attitude of snobbery and arrogance towards anyone who doesn't do things their way. It's getting really old, it's causing real alienation and harm to any cause people might have in the furthering of freedom and openness. This kind of "pushing my beliefs on others" is exactly why so many people started becoming disenchanted with religion, whatever the message may have been that day. It's one of the best ways to lose support.

Oh, but well done to you for the work you've put back into the community, the work of the OSS developers is truely invaluable, it should be appreciated for what it is, not for what other stuff isn't.

I'll stop now :-p

Re:Correction:

[Anonymous Coward]

What in gods name are you talking about?

(your pick, Lady Luck, Lady Lag, or Jeff the god of Biscuits.)

Re:Correction:

[geminidomino]

I didn't confuse anything. My point is that those who would rather sulk about Nvidia not open-sourcing their drivers aren't going to be likely to actually PAY for closed source codecs. It has nothing to do with GNUs "Free as in I-said-so" crap.

DRM, codecs... same thing

[Anonymous Coward]

Funny this comes right after an article that slams DRM...

"DRM's sole purpose is to maximize revenues by minimizing your rights so that they can sell them back to you..."

There should be no proprietary codecs. Ever.

but?

[tomstdenis]

Aren't there native OSS libs for mpeg4 playback already? Might need a binary for wmv but not for mpeg4.

Tom

Enlighten me...

[Anonymous Coward]

I haven't found a media file yet I couldn't already play. How much overhead can I expect to save from switching over to these paid binaries?

Not worth it for WMV

[Anonymous Coward]

All the WMV videos I have seen look like shit; why would I pay for a codec?

Re:Not worth it for WMV

[El Lobo]

Really? What are you using to play your WMV? Your toaster?

Re:Not worth it for WMV

[Shawn is an Asshole]

Many WMVs I've seen look like shit. It's not the fault of the codec, it's just many idiots feel the need to re-encode videos over and over again. Same problem with Google Video and YouTube. So many things there have been re-encoded so many times all you can really see are blury objects behind the encoding artifacts.

Re:Not worth it for WMV

[Lissajous]

All the WMV videos I have seen look like shit; why would I pay for a codec?

Ah...now see, you're probably trying to view them on a linux system using wine to wrap up the native windows codecs. That could well be the reason why your WMV experience is up-til-now sub-optimal. I read somewhere about a company that was going to sell some native linux codecs. I *think* they were called Fluendo or something like that. Anyway, that should sort you out.

Hope this helps, /Lissajous

Re:Not worth it for WMV

[Ash-Fox]

Ah...now see, you're probably trying to view them on a linux system using wine to wrap up the native windows codecs.

Eh? I just use VLC which uses ffmpeg -- opensource codec that can decode wmv9 (and earlier) videos with the same quality that I see under windows.

That could well be the reason why your WMV experience is up-til-now sub-optimal.

Nope.

I read somewhere about a company that was going to sell some native linux codecs. I *think* they were called Fluendo or something like that.

Yeah.. Good luck with that..

Re:Not worth it for WMV

[Anonymous Coward]

Correct. So the reason the WMVs look like shit to me is that they are encoded by windows users who don't care about quality. On the other hand if I see a file encoded with xvid it's more likely to have been encoded by somebody who cares about quality and more likely to be watchable.

So again, given that the WMV videos are lower quality then xvid or divx or mpeg-4, why would I bother to buy a codec for it? No, I'm not watching them on a toaster, I'm using mplayer or vlc.

Good luck with that

[Cheesey]

I suppose the market is Linux distributors who can't bundle MPlayer for legal reasons. Can't see anyone buying this directly, though.

They'd probably be legally unable to be as good as MPlayer, (a universal video player, home page, debs), as licensing some codecs will require signing up to agreements to play nicely with DRM. MPlayer is good because there's none of that nonsense: it just works, for every video that I've tried.

Re:Good luck with that

[Dr. Manhattan]

They'd probably be legally unable to be as good as MPlayer

That doesn't matter - that's not what this is intended for. It just has to be not much worse than the common alternatives on Windows. Linux has plenty of other advantages that make it a good choice - maintenance alone is far easier for Linux than Windows, for example.

I have much more time to visit with my parents when I'm over now that I've got them switched to Linux. I don't have to keep Windows running anymore. But I couldn't have done it (there or with my family at home) if they couldn't watch viral videos on YouTube and email. Setting that up was possible thanks to the quasi-legal packages, but not easy. Some repositories were down when I tried - twice - to use EasyUbuntu or Automatix and if I didn't know what was going on behind the scenes I couldn't have done it.

I'm not convinced that ESR has the timing right, but the general outline - that the transition from 32- to 64-bit represents a major opportunity for Linux, and being able to play (note: not edit, just play) multimedia stuff easily and legally is important - I think is spot on. See here for the oft-argued-about details.

Users with more advanced needs or less full pocketbooks (or less ethics, depending on the exact circumstances) could use the 'other' packages. But a good out-of-the-box multimedia experience is worth a lot for Linux promotion. Since the problem isn't technical, it's legal, a legal approach is unfortunately needed.

Re:Good luck with that

[Pecisk]

I don't need player who "just works". I need nice grapgical interface (and no, no mplayer, no gxine/xine, no vlc can provide that, they are totally geekish apps. I am geek, but for videos, I need something more "normal"), I need nice integration with my enviroment, etc. etc.

Totem/Rhythmbox provides me that. Yeah, I know, anyone who would say that mplayer isn't for him must be stupid or luser, it is stock answer to my requirements. But it won't change a bit what I said.

Mplayer is legal nightmare and isn't even developed fully anymore. Yes, it is nice to do some crazy things like playing movie in framebuffer, but that's it.

Buyers could be Gstreamer/GNOME/KDE users who would want to be legit for some reason - companies, shops, public terminals, etc. So this offer provides rather elegant way to do this. For me, I will stick with gstreamer bad/ugly plugins.

Re:Good luck with that

[xenocide2]

Of course, MPlayer is also legally unable to be as good as MPlayer. The w32codecs package they create is quite large but given out for free. It seems Fluendo aims to resolve this by making available many of the libraries previously used via w32codecs. As a result, they're probably going to pursue MPlayer on the subject. Many of the codecs are probably freely distributable, and I hope that doesn't change.

Of course, there is another reason one might buy this directly: non x86 platforms. Powerbook G4 users frequently find themselves shafted by binary only Linux projects, to the point where it might make sense to run OS X instead. MPlayer of course runs on OS X as well, but the codecs still don't work. I guess itunes / quicktime / whatever the hell apple gives people does the job as well, so people try linux, realize they're a second class citizen, and return to the land of Apple dissapointed. Fluendo may be able to help fix this. Currently, however, your choices are x86 and x86_64. Shame on them!

Re:Good luck with that

[Bert64]

Well, if you want to edit, you can transcode to a more open format and then edit it... It would serve to discourage the use of proprietary formats too, which is great.

Re:Good luck with that

[Anonymous Coward]

Uh, what's wrong with VLC? What would be geekish about it? I've used VLC when I ran XP, in 2003. I used VLC the last two years on my Mac.

A few days ago I switched back to Linux (SO much better than that Mac OS) and I can't see what makes Totem less geeky than VLC. Ok, it's native Gnome, no WX-Windows stuff, but other than that I'd use VLC whenever Totem doesn't work for me.

IMHO VLC isn't geeky, but it's a player that works. On all major platforms. Can't say that about QT Player or WMP. And compared to Winamp VLC seemed a lot more user-friendly to me.

Re:Good luck with that

[simcop2387]

as far as interfaces go, i've seen that kmplayer (a kde frontend to mplayer) and totem seem to be fairly decent (though i haven't touched it in a while, i used it back when gstreamer was fairly new and didn't support all that much). now that gstreamer has matured alot more i'm not too suprised about this, and hope that things like this happen much more.

Re:Good luck with that

[Ash-Fox]

Fluendo aims to resolve this by making available many of the libraries previously used via w32codecs. As a result, they're probably going to pursue MPlayer on the subject.

MPlayer also uses ffmpeg, which supports MPEG, VC1, WMV (all the way upto version 9 apparently) etc.

ffmpeg is opensource and free. The only issue with it, is that in some countries, they haven't paid licensing fees for certain patents.

Re:Good luck with that

[Pecisk]

Maybe basic interface is quite ok, but I have very big problems with VLC Settings interface. Totem at least have few settings clearly spelled out.

But I can agree abut VLC on Mac and Windows - there it is several steps ahead of other players.

Re:Good luck with that

[Pecisk]

Why does everyone says "in some countries"? Germany, US, Canada, Australia is "some country" now? :)

Honestly, FFMPEG is illegal in MAJORITY of IT world. Why it is so hard to say that? It is patent minefield.

Re:Good luck with that

[Anonymous Coward]

Germany, US, Canada, Australia is "some country" now?
Yes.

Re:Good luck with that

[Anonymous Coward]

I am a evil law breaker. I use linux and Mplayer to Violently break the law, help the terrorists win, and destroy the economy.

Mplayer + firefox mplayer plugin + the codecs bundle makes my online experience FAR better than it is with windows. Everything plays, Hell I can click to save the video file if I want to (the HORROR!) and have complete control over the safety of my PC (THE HORROR!!! OMG!!!)

Just because some rich assholes that run this country think they need to force their ideas on everyone else does not mean that the rest of us will simply roll over and play dead for them.

I am one of those dangerous people. I watch video from websites online on linux. Be Warned! I may be living next door to YOU! sitting there within feet planning my next evil plot, like daring to use a mp3 player or worse, converting a DVD to a mp4 for my daughters ipod....

Run RUN FOR YOUR LIVES!!!!!!!

Personally I say, fuck the law and all those that enforce it. They are useless to society.

Re:Good luck with that

[Anonymous Coward]

They'd probably be legally unable to be as good as MPlayer, (a universal video player, home page, debs), as licensing some codecs will require signing up to agreements to play nicely with DRM. Huh? Such as what?

This is a CODEC ffs - it can tell the player 'macrovision on' but the codec doesn't talk to the device that enforces macrovision, that's up to the player. Which is open source and you can hack it.

I doubt the codec handles DVD IFOs so it won't control no-skip sequences. If it allows you to play DRMed WMV content then that's no loss, is it? So is there an actual problem here or just FUD?

Re:Good luck with that

[Lumpy]

Really... Funny how my wife and 15 year old child are way more capable than you using MplayerGUI. It's easier than Windows Media Player.

I strongly suggest you get out of using a computer and specifically linux, it's way to complicated for you if you cant handle mplayer.

Hell you can even get a mozilla/firefox mplayer plugin that makes it so everything on the net plays nicely without effort.

apt get mplayer-mozilla tends to install everything, but I do admit that that is incredibly difficult to do so you can click on "add remove software under ubuntu and search for mplayer and click on everything that shows up.

Re:Good luck with that

[spellraiser]

MPlayer is good because there's none of that nonsense: it just works, for every video that I've tried.

MPlayer works well, except for streaming videos, especially Microsoft's ASF (surprise, surprise). For instance, there is no support for skipping through streams, which pretty much makes them unusable in my opinion. This is the only gripe that I still have against Linux. All other issues and limitations that I've encountered so far I've been able to use.

Mind you, I still use Linux in spite of this minor con. If I really want to watch a specific video stream I just sit through all of it, or watch it at work, where Windows is the norm, more's the pity. It's just that it would be really nice to resolve this lingering annoyance one way or the other.

Re:Good luck with that

[MikeBabcock]

On Fedora with an extra repo enabled, I can easily yum install mplayer mplayer-plugin mplayer-gui and it works beautifully. On top of that, it works with the win32 codec DLLs as well, giving me plenty of proprietary playback capabilities.

The only problem I've ever had with mplayer was voice/video syncing, but that's gone now in my experience.

Re:Good luck with that

[tepples]

Germany, US, Canada, Australia is "some country" now? Yes. Then which countries were you talking about that have a bigger IT market than those four?

Re:Good luck with that

[Anonymous Coward]

There are no software patents in Germany (nowhere in the EU actually). The fact that either clueless or malicious patent agency employees grant them sometimes does not make them valid - though in the case of patents it unfortunately is irrelevant if they are valid or not.

Re:Good luck with that

[ajs318]

Actually, in at least the EU and the UK, and probably the British Commonwealth, software patents are not legally enforcible.

Nor will they ever be. If the law is changed to allow software patents, then it will be recognised that they were never valid before; so any software patent previously but falsely granted in the EU or UK will be recognised as bogus. The holders will have to re-apply for them, but will be blocked by reason of Prior Art and/or obviety (since there will be code out there to do the same things as what they're trying to patent).

Re:Good luck with that

[jZnat]

If you're editting video, you probably have raw streams for the video, so I don't think patents cover that (how could you patent a format where each frame of video is represented by a bunch of individual YUV values?).

Re:Good luck with that

[ChrisJones]

just because they are able to use it doesn't mean it's in anyway a good solution.
mplayer's GUI is a pretty horrible affair, especially when something goes wrong and it spams you with multiple, incomprehensible errors.

it's interesting that things like mplayer and ffmpeg have always focussed purely on playing media with user considerations a long way behind, while gstreamer is building a framework and a community of media applications based on it. Ultimately though people will go where the best support for their media is, so if gstreamer isn't able to reach parity with mplayer/ffmpeg in terms of codec support then it won't be able to replace those projects. That would be a shame because they aren't flexible frameworks and they aren't in any way pretty ;)

Re:Good luck with that

[loftwyr]

Wow. Your entire post was a fud filled ad for this product.

Mplayer does just work. If you want it pretty, it's called gmplayer and it nicely integrates with Gnome, KDE, Fluxbox, whatever. It can nicely integrate with almost any theme based environment due to it's rather nice theme engine. I'm not sure what you're talking about by calling it a "geekish" app.

It is in full development, they're just trying to stabilize for version 1 (after years of beta, it's nice to see). As for the legalities, you just can't distribute it. Nobody said you can't download and install it. It will also compile on a vast variety of hardware and includes a fantastic number of codecs either built-in or through ffmpeg.

You can stick to gstreamer if you like, it's good software but don't for a minute think that anything else you posted was true.

Re:Good luck with that

[Anonymous Coward]

It's just that it would be really nice to resolve this lingering annoyance one way or the other.

You can start by complaining to the provider of the content. Tell them you are being left out in the cold my them providing it in the format they are providing it in. They probably won't do anything about it but at least they'll know there is at least one person out there who is pissed off about it.

Re:Good luck with that

[evilviper]

I don't need player who "just works". I need nice grapgical interface (and no, no mplayer, no gxine/xine, no vlc can provide that, they are totally geekish apps.

I have no idea how Play, Pause, Stop, Open, Forward, Reverse, etc. buttons can possibly be "geekish".

In any case, if you don't like the default/official GUIs for MPlayer or Xine, there are NUMEROUS other 3rd party front-ends to chose from.

Mplayer [...] isn't even developed fully anymore.

Now that's just complete and total bullshit. MPlayer is being more actively developed, and by more people, than it ever was before, and that continues to steadily increase as well.

Re:Good luck with that

[evilviper]

It is in full development, they're just trying to stabilize for version 1 (after years of beta, it's nice to see).

Every MPlayer release has always been a stable version.

The numbering scheme is being switched because many people got that mistaken impression.

Re:Good luck with that

[evilviper]

as licensing some codecs will require signing up to agreements to play nicely with DRM.

Which ones?

WMP's main format, VC-1 (aka. WMV3/WMV9) is an open standard. Microsoft can't impose any terms on anyone. WMA is in a lot of DAP devices, so that must be rather easy to license.

Quicktime uses entirely open formats, h.264, AAC, and mp4.

RealMedia, well, we already have binary codecs from them on multiple 32 and 64-bit Linux and other Unix platforms.

Re:Good luck with that

[evilviper]

The holders will have to re-apply for them, but will be blocked by reason of Prior Art and/or obviety (since there will be code out there to do the same things as what they're trying to patent).

I don't know UK patent law, but that sounds unlikely. Since the "prior art" in this case will be their own work (made public), I wouldn't expect it to be used as justification for refusing a patent.

Or is there really a secrecy/time requirement for patents in the UK?

Re:Good luck with that

[Anonymous Coward]

Please look up the meaning of the word "some".

Don't troll.

[Anonymous Coward]

NVidia's lamented product is a binary only driver with a binary only kernel component. A codec is application software. The nvidia software runs on the bare metal, interacting with your machine in god knows what ways. The codecs are fairly self-contained. It's not really compariable.

Re:Don't troll.

[Anonymous Coward]

And Fluendo's "binary only" codecs require "binary only" gstreamer libraries from Fluendo.

Re:Don't troll.

[msuarezalvarez]

The codecs are gstreamer plugins, exactly like all the other plugins you can get for gstreamer. Unless by "require binary only gstreamer libraries" you mean that you need the .so file that the plugin comes in (like all plugins...) you do not make much sense.

Re:Don't troll.

[Anonymous Coward]

The DRMed plugins will only work with fluendo's binaries. Recompile the gstreamer source, and it will not work.

Re:Don't troll.

[msuarezalvarez]

The plugins are not DRMed and do not play DRMed content.

From asking on #fluendo on FreeNode, it becomes apparent that your second statement is not true.

Yes, they're part of ffmpeg

[Anonymous Coward]

The three major video players for linux: mplayer, xine and vlc all use the ffmpeg library by default for playing mpeg4 video. Recently, ffmpeg added support for Windows Media 9 (WMV3) as a native codec, so you don't even need the windows binaries for that anymore (although it's probably in "alpha" status).

Re:Yes, they're part of ffmpeg

[BlenderFX]

WMV3 works perfectly with ffmpeg (AMD64 here). I don't really think it's "alpha"

Re:Yes, they're part of ffmpeg

[Pecisk]

But ffmpeg can't be distributed legally in all countries. These codecs are for sale for those people who want be legit - usually companies - common crowd will still stick with gstreamer apps/plugins or mplayer/vlc/ffmpeg combo. I see this Fluendo step as very contributing to choice I can make on Linux system - I can say boss that these codecs can be easily installed in public terminal which aim to provide video serices for example. Just buy a license and vola, you are set.

And kudos to ffmpeg team. I use Totem with gstreamer bad/ugly/ffmpeg combo and I can say - hats off to you guys. Quality is very good.

Re:Yes, they're part of ffmpeg

[Shawn is an Asshole]

The FFMPEG devs have always done an excellent job at bringing Free video support to Linux. Thanks to them most video can be played on Linux without propreitary codecs. There is a downside, though: patents. FFMPEG isn't licensed, so it's not legal for distributions to distribute FFMPEG in countries dumb enough to allow software patents (USA, Japan, others?).

The Fluendo stuff could be a good thing if distros would ship with it. Then video would finally work "out of the box". For those like myself who avoid binary blobs and try to only use things that are truely Free will still have the option of using FFMPEG.

Re:Yes, they're part of ffmpeg

[fireboy1919]

That severely limits the market, doesn't it? It seems like about the only thing you'd be looking at is for mp3 players or embedded devices - and then only for ones that are using codecs developed for x86 or whatever they're making this for.

I mean, what other Linux companies require mpeg4 for anything at all?

Re:Yes, they're part of ffmpeg

[AaronW]

And ffmpeg is not perfect. When it comes to WMV files it still has problems. It cannot decode J-frames and some forms of WMA are not supported.

It also tends to be a bit buggy, with various bugs popping in and out depending on exactly when you check out the code. It also does not have regular released versions like other software.

I just recently had to convert several hundred gigabytes of various videos people have uploaded with varying degrees of success. WMV caused problems, and I cannot get 3gp audio files to decode (possibly due to the fact that I'm running 64-bit.

It also is flakey when dealing with AC-3.

-Aaron

Correction: Leaked codecs.

[Anonymous Coward]

""Fluendo to TRY to sell proprietary codecs for Linux.""

The "your codecs want to be free" crowd will take care of that problem.

Sounds great. If...

[jitterysquid]

Rah rah. I like people trying to sell commercial things on Linux. This will only work if they are johnny-on-the-spot when it comes to updates. I would hate my purchased codecs to keep me from updating gstreamer, the kernel, or whatever. In fact, I should not even have to *think* about my purchased codecs when I run a yum, apt-get, or up2date.

I'll just wait here for the Free Software fire-breathing demons of zealotry. It's quite cold right now and my furnace needs a break.

libavcodec

[Anonymous Coward]

Why pay for something that can be played natively by lavc, and by extension mplayer et al?

I have a better idea

[Anonymous Coward]

Stop encoding media using proprietary codecs!

Re:I have a better idea

[dave420]

But for 90% of the market, whether a codec is proprietary or not is of little consequence. Closed and open work just as well on Windows, and until that changes, that's the way it's going to be. Linux users don't have enough clout to change that, unfortunately.

Re:I have a better idea

[PingSpike]

Hell, I'd settle for every other video I download not requiring a codec I don't have and won't automatically download. And I'm on windows! I think that auto download codec thing for WMP has worked all of one time(s) for me.

Re:I have a better idea

[Rob T Firefly]

Stop encoding media using proprietary codecs!

Wait a sec.. You mean you want us to abandon the possibility of implementing DRM, allow customers to fairly use our media, remove our ability to exploit the living crap out of our content providers, and fail to benefit the industry types who give us huge sacks of cash money to keep telling our bosses that their proprietary codecs rock everyone's socks six ways?

Yeahhhh, we'll get right on that.

Love,
the **AA

Re:I have a better idea

[fossa]

The most recent Windows Video, RealVideo, and Sorenson video (Used by QuickTime prior to H.264) are what I might call proprietary in that there is no public standard. They are likely covered by patents as well. MPEG-4's AVC or H.264 may be publically documented but is covered by patents which put free software implementations in the same legal mud as any proprietary codec. On the audio side, MP3 and AAC are in similar situations: publically available but patented.

For publically available and unpatented codecs, as far as I know we've got MPEG-1 video and MPEG-1 Layer II audio (MP2), and this only because they are so old that patents have expired. Vorbis is a modern unpatented audio codec, and Theora is covered by patents which have been freely licensed to the public making it effectively patent free.

MPEG-1 suffers because it simply isn't as good in a quality vs. bitrate sense as the modern video codecs. It's possible that bandwidth and disk size increases could this help somewhat. MPEG-1 is also only good at resolutions near or below 352x240, even from higher resolution source material. However, MPEG-1 remains the most likely to be playable on a given computer. The one additional advantage it has is that decoding may be less CPU intensive than, say, H.264.

Ogg Theora with Vorbis audio is nice, and competitive with though inferior to other modern codecs in terms of quality vs. bitrate. Like all the Xiph codecs, Theora suffers from obscurity. Version 1 is not officially released, but there are limited tools for playback but little for encoding Theora. Xiph's QuickTime Components recently added Theora playback to the QuickTime system (Mac and Windows). Some Theora filters for DirectShow (Windows Media) support playback and, for the intrepid, encoding.Due to differences between the Ogg container format and other established containers, it has had some trouble properly interfacing with DirectShow for example, but playback does work fairly well. Real's Helix system (Helix Player for Linux, RealPlayer 10 on Windows) also has plugins for Theora playback and encoding.

As an occasional dabbler in filmmaking, I release my videos in Ogg Theora+Vorbis and MPEG-1 Video+MP2 audio. I have released versions using MPEG-4 ASP video (DivX, XviD, ffmpeg's mpeg4) and MP3 audio, yet I lack the necessary patent licenses for that and am somewhat reluctant to do so. Commercial software such as DivX typically extends its patent license to its customers for various uses (I believe Microsoft does not allow commercial use of Windows Media Video by default).

Re:I have a better idea

[shish]

Why not just use VLC / MPlayer; both of which have support for every audio and video format I've ever come across built in?

Re:I have a better idea

[PingSpike]

I actually just started using VLC and am overall pretty happy with it. The only thing I don't like is some videos tend to temporarily get distorted when fast seeking through a video.

Re:I have a better idea

[ardor]

There is no viable alternative. Theora is just too obscure to be useful. There are DirectShow codecs for Theora, but they don't work well, jump over frames, almost always mess up the A/V sync and so on. ffdshow theora is not enough, you need Ogg support too (since Ogg is the container, and theora is the video bitstream). Also, this involves installing codecs in 90% of all cases, since nobody has theora, as previously stated.

So. Alternatives? Dirac? Snow? Powerful and very advanced, but a) beta b) even less known.

So, for ensuring that everybody can watch a video, you have to go with WMV or Quicktime, or avi with mpeg4 video (divx, xvid). A very bad situation, indeed. But this is reality.

Re:I have a better idea

[rhizome]

I actually just started using VLC and am overall pretty happy with it. The only thing I don't like is some videos tend to temporarily get distorted when fast seeking through a video.

Those are probably WMV's or another of the Microsoft codecs which seem to have bad support for seeking and random access.

Re:I have a better idea

[joeljkp]

I've given this a bit of thought over the past year or so. The best alternative is probably Ogg/Theora/Vorbis, so your main problem is simply to boost the accessibility of those codecs and make them popular to use. There are a few usable Theora encoders for *nix and Windows, but I don't know of a real (legal; ffmpeg2theora, you're out) drag-and-drop converter.

Either way, the demand simply isn't there. This could be fixed by creating a site like YouTube called OggTube or something, and using the Cortado plugin, but good luck with the bandwidth bills. Plus, YouTube has essentially cornered the market on user-created video. Another alternative is to create a site whose sole job it is to amass video from across the web and convert it to Theora for people to download (like a FilePlanet clone, but doing your own conversions). Again, the bandwidth and labor would be killer, unless you used BitTorrent.

I think the real winning situation would be to hit it at a higher level, by trying to convince a major producer of online video (EA perhaps) to offer their stuff in Theora alongside the standard offerings. Or by getting Adobe to include Theora in Flash.

That said, if anyone wants to help me out with the Theora conversion/BitTorrent site, let me know.

Re:I have a better idea

[evilviper]

There is no viable alternative. Theora is just too obscure to be useful.

MPEG-1 isn't obscure. The quality is much better than VP3/Theora. In low-bitrate situations like web video, MPEG-1 often looks better than MPEG-4, when encoded with a modern codec like lavc (found in mplayer/mencoder, ffmpeg, avidemux, et al.)

The audio codec (MP2) isn't quite as good as MP3 (commonly used with MPEG-4/Divx), but you only need about a 25% higher _audio_ bitrate for similar quality. Besides, audio is always a tiny fraction of the video bitrate anyhow.

So, for ensuring that everybody can watch a video, you have to go with WMV or Quicktime, or avi with mpeg4 video (divx, xvid).

MPEG-1 will play on every video player program made for many years now, as well as DVDs, and many portable hardware players. No extra codecs to be installed, no specific player program needed. No platform issues. It works everywhere. It's far ahead of any other codec in many respects, and just happens to be free due to patent expiration.

Re:I have a better idea

[evilviper]

MPEG-1 suffers because it simply isn't as good in a quality vs. bitrate sense as the modern video codecs.

MPEG-1 video is nearly as good as MPEG-4/Divx when encoded with a modern encoder, such as Libavcodec.

At worst, it's still better than VP3/Theora, which was crappy years ago when it was introduced.

MPEG-1 is also only good at resolutions near or below 352x240

That's positively ridiculous. Idiots can put up webpages as easily as experts.

No mention of methodology (which codecs, which coding method, what keyframe settings, etc.), basis for the conclusion (PSNR? Visual? Other?), etc.

Re:I have a better idea

[fossa]

Is there a guide to getting the most out of MPEG-1 video? I obviously don't know much about choosing the best encoding options, but in my experience, it has been inferior (in terms of quality vs. bitrate) to both ffmpeg's mpeg4 and Theora.

I just did a very rough test using a 720x480 source file. I encoded four files, all at overall (based on filesize including mpeg-1 or avi container) bitrate of 290 kbits/s. Two were MPEG-1 video, one at 720x480, the other at 352x234. The other two were ffmpeg's mpeg4, again at 720x480 and 352x234. They were, probably stupidly, encoded using the -qscale option alone to control file size, e.g. ffmpeg -i input -vcodec mpeg1video -qscale 5 output.mpg. The qscales for the four above files were respectively 18, 5, 15, and 4. Playback using MPlayer with the nearest-neighbor software scaler at 720x480 resolution shows a very pixelated look in the 720x480 unscaled videos, and a artifact-full but non-pixelated look in the 352x234 files that MPlayer scaled back to 720x480. The 352x234 mpeg4 file appeared the highest quality, having fewer visible artifacts than the 352x234 MPEG-1 file. One big trouble with this test is the scaling algorithm acting as a filter on the low resolution files, possibly smoothing so variations. Text from a credits scene looks unsurprisingly pixelated, but the video itself looks pretty good; I expected the nearest neighbor scaler to look worse.

I was not expecting to see lower resolution win out in both MPEG-1 and mpeg4 files. Of course, ffmpeg has dozens of encoding options, and it's likely that poor options were chosen. That said, the comment about MPEG-1 video is not inconsistent with my experience, as rough and uninformed as it is.

Re:I have a better idea

[fossa]

Oops, I meant to add that using ffmpeg's "sameq" option produces a 2.1 Mbit/s MPEG-1 and a 1.6 Mbit/s mpeg4 file. Also, I made a Theora file at a similar bitrate (a bit lower; couldn't get the same value) using the alpha5 encoder and ffmpeg2theora. It looks nicer than the others, not having the extreme pixelated look of the other 720x480 files. I have heard some mentions of VP3/Theora smoothing the video... In this case, the smoothing seems to produce a superior looking video, but in other cases it may have a negative effect compared to MPEG-1 or mpeg4.

Re:I have a better idea

[evilviper]

Is there a guide to getting the most out of MPEG-1 video?

MPlayer's guide to encoding (with MPEG-4) will give you a good overview of libavcodec's options, and how to get the best quality. Most of which apply directly to MPEG-1 equally as well.

The one big "gotcha" is that the guide recomends *cmp=2, which will make the video extremely blocky, in expecting everyone will use deblocking on playback (which often isn't the case). *cmp=3 is very good.

They were, probably stupidly, encoded using the -qscale option alone to control file size,

Yes, that is decidely... sub-optimal. Just switching to 2-pass encoding will make a world of difference.

ffmpeg -i input -vcodec mpeg1video -qscale 5 output.mpg

That's probably your problem. ffmpeg is automatically setting very frequent keyframes for MPEG-1 (for VCD-compatible hardware decoding), whereas it uses very infrequent keyframes for MPEG-4. You can override this manually (in a lavc config file) or you can use mencoder, which won't "automatically" change settings on you.

the other at 352x234.

Dimentions always need to be divisible by 16. Otherwise, you're just wasting (lots of) bits.

I was not expecting to see lower resolution win out in both MPEG-1 and mpeg4 files.

At such a low bitrate, it's not surprising. Double the bitrate for the (4X) higher resolution and you'll get much-better-looking results.

Nice try, but...

[sid77]

I think someone should point them to the ffmpeg changelog. Actually it does open lots of proprietary formats and VLC uses the same codebase for his own engine.

Re:Nice try, but...

[Nadir]

I think they already know about it.

Re:Nice try, but...

[GauteL]

Several of the codecs in mention are heavily patended and FFMPEG's legality in many countries is not clear. The codecs from Fluendo, however, have a clear legal status as Fluendo has signed agreements with the patent holders.

That is the major difference between the offerings from Fluendo and FFMPEG.

Re:Nice try, but...

[loki_tiwaz]

ah yes, also, snow will be stable fairly soon, it certainly played ok a year ago, and wipes the floor with every codec i've seen bits/pixel-wise (and my hardware can't play h264 but it plays snow smooth as). who needs a stupid proprietary codec, there's open ones for most files one wants to play and within a year or two every DRM yet invented gets pwned.

i fell for paying for a driver for linux once, cos i needed to use a dialup modem with a connexant chip. i'm quite happy to say that my modem is now a dsl modem inside a little dedicated box running embedded linux and giving me access with ethernet, and not some stupid proprietary modem which eats cpu cycles instead of providing silicon.

nvidia will eventually suffer the ignominy of being the wide open backdoor into linux systems and the reasons for the complaints of the 'purists' will become manifest in the form of a lot of irritated customers who paid real money for nvidia chips. i am already annoyed that they've dropped support for gforce4 video cards, why is it still supported in the windows version?

eventually gpu acceleration will be sold with an open interface by someone with an open driver and it will do polygon for polygon as good as someone else's. or maybe ati will decide to do a netscape on nvidia and open up their hardware. gpu's are incredible accelerators for more than just 3d vectors, there is several other uses for that type of processing and a closed spec prevents their use except by a select number of developers, the rest of the devs being too numerous to be 'manageable' (a-la microsoft's prohibitive development licenses for hardware utilising their DRM) meaning the incentive arises for developers to target the less popular but accessible platform. doom 3 anyone?

Re:Nice try, but...

[evilviper]

ah yes, also, snow will be stable fairly soon,

Don't count on it. Development is continuing, but slowly.

Still, as slow as it's going, it still might beat Theora to a final/stable release.

World Domination

[rowama]

If you are from the ESR tribe you will see this as a positive step towards world domination.

http://catb.org/~esr/writings/world-domination/wor ld-domination-201.html

ESR, et al, believes the ability to play codecs such as these is so vital to the 2008 world domination deadline, that we should put up with these binary blobs. For a while, at least.

Lindows is supposed to be working on this also.

Re:World Domination

[QuantumG]

Is anyone a member of that tribe anymore? ESR is the Trotsky of the open source revolution.

Re:World Domination

[kripkenstein]

It's "Linspire" these days, and ESR now works with them.

Re:World Domination

[Abreu]

So that means he'll run for his life and end up being killed in Mexico City? I don't think so!

Re:World Domination

[QuantumG]

You've gotta reinterpret these things to the current revolution. He's already been ostracized by a dictator for failing to uphold the idealism of the mentor, even though the dictator believes even less in the ideals of the mentor. As for being killed? What's the open source equivalent of that? Sued into bankruptcy? Nah, you can be broke and still hold sway over the community. Personally, I think he's gunna end up cracking rocks for some kind of fraud.

There is a market

[dhuv]

First of all, look at the Open Sound project. They filled a niche by selling drivers for certain sound cards that had more features compared to the OSS drivers.

Second, I think that it is even easier to sell these kinds of things today. They can make a deal with somebody like Novell or Xandros who want to provide their users with a fully functional fully LEGAL linux desktop. This will help them to do that.

I don't see this being so popular with non-commercial distros like Debian because its a different set of users. But with commercial distros like Xandros (who already offer things like Codeweavers Office), I think its a great fit.

Re:Sounds great. If...

[Pecisk]

If you have read Christian (Fluendo most visible VP) blog post, they plan to provide upgrades via distribution upgrade/installation system (apt-get, yum, etc.)

Breaking news.......

[AlzaF]

finally porn has been ported to *nix.

This could help acceptance of the Linux desktop

[jonnyj]

I hate DRM as much as the next person, but this is good news. Acquiring and installing proprietary codecs is a dark art that is major obstacle to wider acceptance of the Linux desktop.

Given some further development, I can see a few opportunities:

• distros like Suse and Ubuntu could integrate their package management systems with Fluendo and offer fully legal point and click codec installation for a small (compared with the Windows anti-virus tax) fee
• suppliers of Linux based PCs and laptops could offer products that play nicely with the rest of the web
• system builders might be able to start offering Linux-based media centres build around applications like MythTV

As someone who absolutely refuses to pirate software unless I have no choice, I'd be prepared to pay a few ££ extra to stay legal.

Re:This could help acceptance of the Linux desktop

[zakezuke]

As someone who absolutely refuses to pirate software unless I have no choice, I'd be prepared to pay a few ££ extra to stay legal.

I wonder if keeping an extra copy of windows would make use of the propriority .dlls needed for mplayer WMV support would be considered legit and not piracy.

I'm not so much anti-piracy but pro working shit. For example turboprint. If I ever needed to print to my canons under linux, this is something i'd consider paying for.

Re:This could help acceptance of the Linux desktop

[kosmosik]

> Acquiring and installing proprietary codecs is a dark art
> that is major obstacle to wider acceptance of the Linux desktop.

This is BS. Installing multimedia support for decent distro is as easy as:
- enabling an additional repository
- issuing a command

You can do this either clicking with your mouse or just with terminal. I'll show with terminal since it is more strict, for Fedora:

% su -
(here enter your root password)
% rpm -ivh http://rpm.livna.org/livna-release-6.rpm
% yum install mplayer-gui vlc xine

And you basically get players to play most of the content. The content you won't be able to play are extremely DRM infested files like WMV9 (which Macs also won't play) and so on. It is for FC6 but I am pretty sure it works in similar way in any other major distro - just ask their helpfull community.

Now real question - was it that hard? If you find it hard (you don't need to understand it, just copy and paste the lines) I find you retarded. Sorry. It was not harder than getting Windows to play DVDs or some obscure codecs.

> As someone who absolutely refuses to pirate software
> unless I have no choice,

WTF you are saying? It is possible to play most of media files just fine with OPEN AND LEGAL codecs. It is not like you need to go to PirateBay and download some codecs to make it work in Linux. These codecs have DISTRIBUTION restrictions in some countries (namely USA) but not USAGE restrictions. It is perfectly legal for me to play WMV, MP3 and others using libmad, ffmpeg or smth. similar.

> I'd be prepared to pay a few ££ extra to stay legal.

They probalby look for uninformed people like you with such offer. :)

Re:This could help acceptance of the Linux desktop

[Pecisk]

They are covered by patents are therefore is a subject of USAGE restrictions.

And no, vlc and xine is no use for simple user. Computer geek - maybe.

Re:This could help acceptance of the Linux desktop

[scdeimos]

I agree that this will help with the take-up of Linux, but I don't see how you can equate DRM with proprietary codecs.

Proprietary codecs often have some benefit to end-users in the form of improved quality, improved resolution (not the same thing) and/or improved compression. SVQ (Sorenson video) codecs in Apple Quicktime certainly helped to make that product popular for some of those reasons, at least until it was surpassed by MPEG-4.

DRM is a benefit to nobody, except when it comes to lining the pockets of content producers. I equate DRM with user-inhibiting and the theft of pre-existing rights.

Re:This could help acceptance of the Linux desktop

[kosmosik]

> They are covered by patents are therefore is a subject of USAGE restrictions.

What patents you are talking about? I bet you don't have any idea.

> And no, vlc and xine is no use for simple user.

Even if so... type:

% yum install totem-xine

It is Totem with xine backend. If you find it hard to use I really find you retarded.

Re:This could help acceptance of the Linux desktop

[repvik]

This is BS. Installing multimedia support for decent distro is as easy as:
- enabling an additional repository
- issuing a command

That ain't BS. How the FUCK is grandma going to know this?

% rpm -ivh http://rpm.livna.org/livna-release-6.rpm
% yum install mplayer-gui vlc xine

WTF is that? What does -ivh do? Where'd you get that URL?

Now real question - was it that hard? If you find it hard (you don't need to understand it, just copy and paste the lines) I find you retarded. Sorry. It was not harder than getting Windows to play DVDs or some obscure codecs.

Oh really? My windows plays DVD's. An app came with my PC that does that. Fucking moron. The fact that you know exactly what to do to get these codecs doesn't mean that J. Random Luser does. Infact, it's quite unlikely that he will figure it out during his 25 second attention span.

Get over it. Linux sucks when it comes to usability when compared even to windows. There should be *ABSOLUTELY* no need for a user to open a terminal window and type obscure commands to do simple stuff like that.

Re:This could help acceptance of the Linux desktop

[Kjella]

Get over it. Linux sucks when it comes to usability when compared even to windows. There should be *ABSOLUTELY* no need for a user to open a terminal window and type obscure commands to do simple stuff like that.

There is absolutely no need, the 1990s called and want their "but linux has no graphical package manager" FUD back. It's just that slashdot isn't well suited for a graphical tutorial page (inline goatse.cx anyone? *shudder*).

Re:This could help acceptance of the Linux desktop

[jonnyj]

I assumed - perhaps mistakenly - that the Fluendo plugins will allow access to DRM'd files. Like it or not, if the Linux desktop is to hit the mass market then it will have to provide a mechanism for people to be able to play back music and video that they've bought online.

Re:This could help acceptance of the Linux desktop

[p0tat03]

I find your suggestion of using the command line for a "simple user" to be somewhat retarded. Ease-of-use is end-to-end, not isolated to one section of the experience. However easy your player is to use, if the user cannot grok how to install the damned thing, it's of no use to them.

Re:This could help acceptance of the Linux desktop

[ratboy666]

Wow!

In no particular order - online systems currently use text as the primary communications medium. Not mouse gestures and clicks. Windows used to come with a "macro recorder" -- I imagine that something similar could be cooked up, and the GUI events captured, canonicalized into a text format, added to a post, and fed back. Roughly, I imagine the sequence would look like:

MOUSE MOVE ABSOLUTE 15 10
LEFT CLICK
MOUSE MOVE ABSUTE 50 100
LEFT CLICK ..

and so on for a bit. Not that useful, when compared to cut and paste of mostly readable commands. More recent Windows versions have dropped the Recorder application, in favour of scripting, by the way, which is closer to the Linux (Unix) way.

After all, you seem to understand most of it, except "-ivh" (which stands for "install,verbose, hash", if it matters to you).

Cut, paste: which is what a GUI is good for. Symbolic, dense commands: which is what a CLI is good for. This example is a happy medium, not a negative example. The correct comparision is to obscure RegEdit keys; against that, -ivh wins. How to determine "-ivh"? By lore (as you have just done), reading the RPM guide, typing "rpm", looking on the web, etc.

As to livna. Generally, found by Google for the query: "how do i play dvd in fedora 6" and following the first link.

However, the end user (who is NOT a J. Random Luser, and generally has greater than 25 second attention span) should use a distribution with local support (someone around who can help) if they are not computer savy. This is one of the "network effects" and is CERTAINLY the case with Windows adoption. If this is the case, a simple personal question would do.

Generally, most Linux distributions are simple enough to use that a network query will work as well (as it did for this case).

My "Grandma" (Grandma to my kids) doesn't know that. All Grandma knows is that a functional computer that does email, light web browsing (looking for discount air fares), and webcam is $30. She also knows that a Windows XP license for that computer is $230. Office (for some light WP and bill payments) is hundreds more.

As a result, Grandma uses Windows 98 (which license CAME with the computer), or Linux. Pretty much the only application of interest is OpenOffice, and "MSN Messenger".

Grandma was very happy with Linux, until she wanted to run MSN Messenger with Webcam. She then put the old Windows 98 license back into use. And reloads the computer every 6 or so months. She is getting to be quite the tech. Number 1 computer complaint from Grandma? When will MSN Messenger with Webcam be out for Linux... Oh, and Grandma even uses Gmail, for ease of rebuilding computer.

[tech note - yes I know amsn, etc. but required is support for logitech webcam with only v4l2, not 1 support. So it doesn't quite work yet... maybe next update to amsn (and I don't have time to fix it personally, sorry Grandma).

Re:This could help acceptance of the Linux desktop

[Pecisk]

I have idea - most of ffmpeg codecs ar covered with nasty (mpeg2, mpeg4) patents and therefore making them not so legal.

And calling me "retarded" is a nice touch, thank you.

No, I don't use xine backend. I use gstreamer backend. Which is installed by default. Yawn.

Re:This could help acceptance of the Linux desktop

[repvik]

Far too long for my attention span :-P

I have absolutely no problems using the CLI myself. I administer several linux servers and run linux on my desktop and laptop. I don't run any rpm-based distro though. The point is that currently, it's far too advanced to get "normal" stuff working compared to Windows or OSX.

However, the end user (who is NOT a J. Random Luser, and generally has greater than 25 second attention span) should use a distribution with local support (someone around who can help) if they are not computer savy. This is one of the "network effects" and is CERTAINLY the case with Windows adoption

The J. Random Lusers I regularly support hardly have that long attention span. Saying that the End User should use a distribution with local support is fair enough. Except that an OS shouldn't really need this. Why do I hear that Linux is ready for the desktop already, when it's obviously far inferior compared to the two major commercial offerings? What "network effect" are you talking about? Do you mean that windows is so common because your local geek said you should install it? That's complete BS if I ever heard it. Windows comes preinstalled on PC's (with a very, very few honest exceptions). Windows is default. Most people doesn't care as long as they can use it.

Re:This could help acceptance of the Linux desktop

[tepples]

There is absolutely no need, the 1990s called and want their "but linux has no graphical package manager" FUD back.

What about "but linux has no patented codecs in the default configuration of its graphical package manager"? Fluendo will provide an alternate solution to what I emphasized.

Re:This could help acceptance of the Linux desktop

[kosmosik]

> I find your suggestion of using the command line

Then go back and reread entire post. Specially the part telling that you could also use your mouse pointer to the same thing.

Re:This could help acceptance of the Linux desktop

[kosmosik]

> I have idea - most of ffmpeg codecs ar covered with
> nasty (mpeg2, mpeg4) patents and therefore making
> them not so legal.

Depending on where you live.

Re:This could help acceptance of the Linux desktop

[ratboy666]

To summarize (since you don't like reading the long post): Don't insult people. Windows is good; mostly because lots of people use it. Linux is good because it has excellent driver support built-in and lots of useful applications out of the box.

Long stuff:

I get a call from a Windows user. The following must be done: read a "pst" file and extract emails. User: lawyer, *not* LUSER. Certainly has a reasonable attention span. Local IT help: useless (and this isn't forensic, thank the deity). Of course I bill for the phone conversation (she did). Indeed, I have a timer on my desk for these types of issues.

Typical kind of operational Window issue. Why does the law firm use Windows? Because their clients do. That is the network effect. The problems that ensue? Hey, I don't mind. Microsoft likes lock-in; doesn't matter to me -- I make money from it. No need to call the lawyer a LUSER -- what the fuck is a ".pst" file anyway?

I guess that with so many users, Windows is "easy to support", but there are motherboards which REQUIRE OEM XP installations. Go figure, I happen to have one. Load XP from CD, the CD disappars on first boot. XP cannot then load the "motherboard driver" which is needed to enable networking or CD. Figure that. Is your contention that an average user can slipstream the needed drivers into XP? Without access to a running XP? Just wondering. Exercise: compare and contrast to setting up a Linux MythTV box.

Windows is preset by manufactures to "just work". By itself? Its more of a crapshoot than Linux. Linux comes with a boatload more driver support "out-of-the-box". It much more probable that a new user would be able to set up a useful Linux (Ubuntu / Fedora Core) system EASIER than setting up a Windows XP system. (XP doesn't come with DVD playback either -- you actually have to BUY this feature)

Re:This could help acceptance of the Linux desktop

[ajs318]

Ah, well, it's likely that the upcoming Zune vs iPod vs PlaysForSure wars will wake people up to what DRM really means.

As an aside, a promotional catalogue that fell out of last week's Radio Times listed a combined record/CD player/radio with built-in MP3 encoder that can write to SD/MMC cards or USB mass storage devices. This is the sort of thing that is going to take digital music right into the mainstream.

Re:This could help acceptance of the Linux desktop

[repvik]

To summarize (since you don't like reading the long post): Don't insult people. Windows is good; mostly because lots of people use it. Linux is good because it has excellent driver support built-in and lots of useful applications out of the box.

I don't intend to offend people. It's a simple fact that I observe during my support calls that people can't be bothered to investigate things like this on their own. Things should "just work". I sometimes use windows because things "just work" when I can't be bothered to debug a problem on linux. It's lazyness, not incompetence.

I get a call from a Windows user. The following must be done: read a "pst" file and extract emails. User: lawyer, *not* LUSER. Certainly has a reasonable attention span. Local IT help: useless (and this isn't forensic, thank the deity). Of course I bill for the phone conversation (she did). Indeed, I have a timer on my desk for these types of issues.

Typical kind of operational Window issue. Why does the law firm use Windows? Because their clients do. That is the network effect. The problems that ensue? Hey, I don't mind. Microsoft likes lock-in; doesn't matter to me -- I make money from it. No need to call the lawyer a LUSER -- what the fuck is a ".pst" file anyway?

Quite a few *users* have a decent attention span and might even attempt to solve problems themselves. Lusers don't.
Now, did the lawfirm actually choose between windows and any other operating system? Or did they just buy a bunch of PC's from Dell? Buying a PC from (eg.) Dell gives you a Windows PC that people can expect to work with other PC's without much/any effort.

I guess that with so many users, Windows is "easy to support", but there are motherboards which REQUIRE OEM XP installations. Go figure, I happen to have one. Load XP from CD, the CD disappars on first boot. XP cannot then load the "motherboard driver" which is needed to enable networking or CD. Figure that. Is your contention that an average user can slipstream the needed drivers into XP? Without access to a running XP? Just wondering. Exercise: compare and contrast to setting up a Linux MythTV box.

I've had several problems installing Windows XP on "blank" boxes. Eg. getting S-ATA to work so that XP actually boots after being installed. Either requires slipstreaming or loading a driver off a floppy during installation. A google with the motherboard name and windows xp usually has the problem in the first few hits. Anyway, users don't usually install windows. It comes preinstalled. If they need to reinstall, they usually also have a "recovery-cd" with all the drivers necessary included.
I'll pick up on the MythTV thing. I've built myself a Linux MythTV box using a VIA Epia M10000 board. Installing Ubuntu was a no-brainer, since I've done it dozens of times before. Getting all the hardware to work proved a task that took several days. Linux does *not* have "excellent" driver support. Eg. my Hauppauge WinTV PVR USB2 requires downloading a driver, compiling and making sure it'll be loaded on boot. My Hauppauge WinTV PVR 250 requires an IVTV-driver. Same procedure. Now, getting my wireless stick to work (of course without network access to the box, since wireless obviously doesn't work yet). Get driver from zd1211.ath.cx, compile, install. Wow. Now I've gotten my two tuners working, and even have internet. To get mpeg acceleration working, I need to find a driver provided by VIA on a *forum* (viaarena), download, compile, install. Getting the LCD-display in the case working was "easier", just download lcdproc and compile/install. The remote control proved harder. A lot harder. After some investigation, I found the driver, downloaded, compiled, installed. It didn't work! After a lot of googling, I could get the remote control to work. Partly. It didn't work in MythTV. That required another obscure config-file, that I had to make myself. Phew. After spending a couple of days

Re:This could help acceptance of the Linux desktop

[rpdillon]

Not so much.

Sure, you give an example under Fedora. But the reality is:
1) The software is illegal for a good portion of the people who are interested in using it, therefore:
2) The repositories are "unofficial" and subject to disappearing at any given time (even though Livna has been around for a while), and:
3) They don't work on x86_64. The new ffmpeg stuff should help with some of that, but it is too new for me to have given it a try quite yet. As of November/December, my only way to watch most of these codecs was through a 32 bit chroot on my Kubuntu Edgy box using psuedo-illegal codecs from Debian universe.

So, despite what you say, it is kind of a dark art. Finding the repos is a pain (varies by distro), and making it all work with 64 is a pain, especially plugins for browsers (you have to use 32 bit wine to run the .dlls for win32 in 32 bit mplayer, but if you want to use that in firefox, you have to get 32 bit firefox, but then you lose Java because the java you run is 64 bit, so then you install 32 bit java as well...) It goes on and on.

I've been messing with it all in kind of a techno-haze for the past 5 years wondering when someone would fix the system. I started the (now huge) thread ("32 bit mplayer without 32 bit chroot") on Gentoo's AMD64 forums about this back in late 2004/early 2005 and there is still no good solution. Personally, I'm glad the drivers are available...I might just use them.

Re:This could help acceptance of the Linux desktop

[kosmosik]

> Sure, you give an example under Fedora. But the reality is:
> 1) The software is illegal for a good portion of the people
> who are interested in using it, therefore:

Please be specific. I live in Poland/Europe and software that I use is perfectly legal in fact I've control in my offices and they didn't mind that every workstation (Windows to be specific) was equipped with VLC.

> 2) The repositories are "unofficial" and subject to disappearing
> at any given time (even though Livna has been around for a while),

Of course they are unnoficial. No official support from RedHat for example. Do you get official support for WinAmp from Microsoft? WinAmp is also unofficial for MS I think - so your point is?

> 3) They don't work on x86_64.

But this is not due legal and practical issues we are discussing? I don't use x86_64 on my workstation since I don't see any need for it. So I reall don't know. Does Windows XP support 64bit? No? Ach - that one version with no drivers? :) As for Livna. I've found this listing:
http://rpm.livna.org/fedora/6/x86_64/repodata/repo view/sound-and-video.group.html

But I guess it is beta or something and I really cannot confirm if it works. Looks like normal packages for me.

> The new ffmpeg stuff should help with some of that, but it is too new
> for me to have given it a try quite yet.

Why can't you just upgrade your distro?

> As of November/December, my only way to watch most of these codecs was
> through a 32 bit chroot on my Kubuntu Edgy box using psuedo-illegal
> codecs from Debian universe.

Ah! ;) You are running 64b kernel? It is possible to run 32b software in such setup but it requires hassle.

> So, despite what you say, it is kind of a dark art. Finding the repos
> is a pain (varies by distro), and making it all work with 64 is a pain,
> especially plugins for browsers (you have to use 32 bit wine to run
> the .dlls for win32 in 32 bit mplayer, but if you want to use that in
> firefox, you have to get 32 bit firefox, but then you lose Java because
> the java you run is 64 bit, so then you install 32 bit java as well...)
> It goes on and on.

Yeah. Well - you think this is pain? Try running 64b version of Windows XP. :P

> I've been messing with it all in kind of a techno-haze for the past 5 years
> wondering when someone would fix the system. I started the (now huge) thread
> ("32 bit mplayer without 32 bit chroot") on Gentoo's AMD64 forums about this
> back in late 2004/early 2005 and there is still no good solution. Personally,
> I'm glad the drivers are available...I might just use them.

Well 64b kernel mode is right now not mainstream on workstations - face it, MS does not do that, Apple does not and Linux in fact tries and it is usable (but with few glitches). Everybody is on 32b now and will be since main productivity apps are ported (no so soon).

But 64b/32b it is NO BARRIER for codecs - the law, the patents are. Not the implementation.

This is good for everyone

[gravesb]

A lot of people will complain about this, but I think it is good. The more software, of any kind, that comes out on Linux, the better. The more software that eliminates some of the work to get a system up and working, the better. The more shrinkwrapped software for Linux on store shelves, the better. As Linux gets more visibility, and day to day tasks become easier, more people will adopt. This will drive development for Linux. Hopefully some of those adopters, especially the younger ones, will get involved in Linux, see what it has to offer, and become contributers. They may or may not move to a fully open system, but some of them will, and the more people who do, the better. The more weight behind Linux, the more power the community has, even if that weight was gained, at least initially, through propriety software.

Re:This is good for everyone

[Ash-Fox]

A lot of people will complain about this, but I think it is good. The more software, of any kind, that comes out on Linux, the better.

Except we already have opensource implementations of MPEG, WMV, VC1 etc. in ffmpeg. Which is used in mplayer, vlc etc.

Re:This is good for everyone

[AusIV]

Except that this takes away some of the legitimacy of using the same codecs for free, and gives fewer reasons for media developers/distributors to use open codecs. If you had somewhere you could have paid for it, circumventing the law to get it seems more immoral than if you had to choose between ripping off the codecs and using Windows. Then, once Linux gains a little bit more popularity, media distributors might have started to take into consideration that they can't play wmv files, and if they want those customers, they should use mpeg containers or find a different container. I still think the only people who have anything to worry about as far as copyright infringement are the distributors of w32codecs (and they're generally overseas), but I don't think this is a win for everyone either. It's something like a political party gaining popularity by sacrificing their platform: they may start winning elections, but at this point what does it matter?

Re:This is good for everyone

[gravesb]

It gets people in the door, and once people are in the door, its easier to convert them. Also, some people do have issues with breaking the law, whether moral or immoral, and if this was the last hurdle for them (codecs are a big deal for some people), it will let them adopt. Sometimes you have to sacrifice to move forward. Also, there is no restriction on continuing to develop open libraries, and once those are better than the closed versions, people will move to them.

Re:This is good for everyone

[tepples]

Except we already have opensource implementations of MPEG, WMV, VC1 etc. in ffmpeg. Which is used in mplayer, vlc etc.

Using ffmpeg's patent-infringing codecs in several countries with a large IT market, notably including the United States of America, is just as illegal as using a pirated copy of Microsoft Windows OS. Fluendo's offering is under $40 per user; compare to the cost of emigration to escape software patents.

FFMPEG

[BigBuckHunter]

I haven't scrolled to the bottom yet to see if this is redundant, but FFMPEG has recently added WMV9 and VC1 decoders. http://ffmpeg.mplayerhq.hu/ It works fine for all of the Legacy content that I have. They also have MPEG/2/4 and a whole bunch of others. The only codec that I seem to be missing is Indeo 5

BBH

Re:FFMPEG

[joeljkp]

You people seem to be missing the point that FFmpeg and ilk are completely illegal to a large portion of the Linux-using public, many of whom don't harbor deep-rooted feelings about the legitimacy of software/algorithm patents, and would prefer enjoying their media without legal ambivalence.

Also, businesses, commercial distributions...

Re:FFMPEG

[BigBuckHunter]

You people seem to be missing the point that FFmpeg and ilk are completely illegal to a large portion of the Linux-using public
How is FFMPEG illegal, and in which jurisdictions?

BBH

Re:FFMPEG

[joeljkp]

In the United States, due to these links:
http://www.mpegla.com/m2/m2-patentlist.cfm
http://www.mpegla.com/m2s/m2s-patentlist.cfm
http://www.mpegla.com/m4v/m4v-patentlist.cfm
http://www.mpegla.com/m4s/m4s-patentlist.cfm
http://www.mpegla.com/avc/avc-patentlist.cfm

combined with USC 35 Sec. 271:

(a) Except as otherwise provided in this title, whoever without
        authority makes, uses, offers to sell, or sells any patented
        invention, within the United States or imports into the United
        States any patented invention during the term of the patent
        therefor, infringes the patent.

I'm not familiar with other jurisdictions.

Re:FFMPEG

[BigBuckHunter]

Ah, I see. I didn't realize that this affected source code yet.

BBH

Re:Sounds great. If...

[i_should_be_working]

I'll just wait here for the Free Software fire-breathing demons of zealotry. It's quite cold right now and my furnace needs a break.

The path of the righteous man is beset on all sides by the coding of the closed and the tyranny of proprietary applications. Blessed is he who in the name of 4 Freedoms and good will frees the source through the valley of darkness, for he is truly his brother's bitkeeper and the user of obsolete hardware. And I will strike down upon thee with great vengeance and furious anger those who attempt to poison and destroy my installation with binary blobs. And you will know my name is Saint iGNUcius when I lay my vengeance upon thee!11.

Re:Sounds great. If...

[kosmosik]

> This will only work if they are johnny-on-the-spot
> when it comes to updates.

I belive that is why they try to have stabilised API version - gstreamer is close to 1.0 and probably when it is finished the API for plugins will froze so that any new API will not conflict with old (like propertiary codecs) plugins.

> I would hate my purchased codecs to keep me from
> updating gstreamer,

As I've said - if they implement stable API in gstreamer that should not be an issue.

> the kernel,

WTF has kernel to do with codecs anyway? Linux kernel has stable API for applications since ages.

What is funny that I am keep using MPlayer with some bunch of binary Windows DLLs and it still works despite I have the DLLs since ages and update MPlayer to each new version. How is that different?

I don't know

[Necron69]

Their bundled price right now is 28 Euros, so about $36.19.

Would I pay that? I don't know. To be perfectly honest, I haven't actually _bought_ any software for a PC in a very long time. Keep in mind that this exact same functionality is _free_ under Windows.

- Necron69

Re:I don't know

[Anonymous Coward]

Keep in mind that this exact same functionality is _free_ under Windows.

No, it's not free at all. It's part of the $100+ buying price of Windows. Tell me one way that you can legally get these codecs, in any form, without buying something.

Re:I don't know

[petermgreen]

so the real questions are

1: are all the propietry bits you want on your desktop linux system (crossover, cadega, theese new drivers etc) plus the overhead of a distro who can put it in one conviant package for OEMs more or less than an OEM windows licence
2: assuming they are cheaper is it possible to put sufficant shitware on the default install that the price advantage doesn't dissapear (iirc big brands make a fair bit out of shipping windows shitware, possiblly more than they pay out for the windows license)

Re:I don't know

[eerok]

Keep in mind that this exact same functionality is _free_ under Windows.

It's free if you didn't pay for your copy of windows, I suppose.

Otherwise the price of the codecs could well represent your entire cash outlay for a linux distro.

Re:I don't know

[Mad Merlin]

1: are all the propietry bits you want on your desktop linux system (crossover, cadega, theese new drivers etc)

Crossover is not proprietary, unless you're talking about their configuration tool. Crossover is open source and you can find their changes in the vanilla Winehq tree as well.

You are correct when you say that Cedega is proprietary though. They have a public CVS repository, but their license is LGPL incompatible and their public CVS does not update very often (and probably doesn't reflect what's in their proprietary product). They are starting to use components from vanilla Winehq, and thus those components are LGPL licenses, but Cedega in general is not.

Re:I don't know

[physicsnick]

Keep in mind that this exact same functionality is _free_ under Windows. No it's not you fool. You pay $200 for a copy of Windows; a significant portion of that goes to licensing codecs. You just don't notice it when you buy a computer with Windows bundled in, so you think it's free. Sure, if you pirate Windows, it's free, and if you use ffmpeg on Linux it's free. This solution is illegal in the US, and entirely useless to OEM vendors.

why we are releasing these codecs

[Ur@eus]

I see a lot of people confused about why we are releasing these codecs when there are things like the open source ffmpeg codecs etc.
Our goal is not to provide the community with codecs which there is absolutly no support for already as
that would be foolish. Our goal is to provide a 100% legal option which I know a lot of companies who have or
want to deploy linux desktops have been looking for. These companies like open source, but they also have policies in place
which hinders them from deploying solutions which have clear patent issues hanging over them in their country of operations. This is unfortunatly
the case with most multimedia codecs and even though we have spent a lot on resources on Xiph codecs here at Fluendo and are now working with BBC
on Dirac there is still some way to go before the need for non-free codecs are gone.

So for those in a situation where they can freely use gst-ffmpeg and similar options, more power to you! For those who the lack of licensed codecs
has been a hinderance or problem for adopting Linux (or Solaris) desktops at your company or institution or even private use, then we hope our plugins will be a good solution.

Christian Schaller
Fluendo

Re:why we are releasing these codecs

[svu]

Are you going to release the list of "supported" TV-broadcasting services? I would like to know that the service I want to use (webtelek.com) is supported - BEFORE paying any money (and would like to know exactly the codec I have to buy).

Re:why we are releasing these codecs

[schwaang]

I tried to understand the Conditions of Use but I have questions that should be in a FAQ on your site:

1) The license seems to say "per computer". Does that mean it's OK to install the same software in two versions of Linux on the same (dual-boot) computer?

2) Is there a process for me to legally move the software to a different computer (deleting it on the old one, of course).

3) Does the software "phone home" in any way?

4) The Indemnity clause demands that the buyer (licensee) indemnify you against anything. Where's the part where you promise that you have legally licensed the patents that you are implementing, thus indemnifying *us*?

5) Is the software distro-aware? If I install it on Fedora, will I know if some livna RPM tries to blow away a library?

Thanks in advance for any information.

Re:why we are releasing these codecs

[XanC]

Wow, that indemnification thing seems pretty crazy. If that's in place, what's the point of the whole exercise? And why has there been no reply?

Re:why we are releasing these codecs

[Ur@eus]

1) The license seems to say "per computer". Does that mean it's OK to install the same software in two versions of Linux on the same (dual-boot) computer?
Yes, this would be fine.

2) Is there a process for me to legally move the software to a different computer (deleting it on the old one, of course).
Yes, doing this would be fine. Be aware however that the plugins are sold per architecture, so if you do switch architecture
as part of this move you would still need to buy the plugin again for that architecture.

3) Does the software "phone home" in any way?
No, not at all. And to reply to one of the questions implied by this, we are not worried about piracy at all.
Because using a pirated version of our plugins would give the user nothing that for instance just grabbing
the open source codecs available for linux gives them. The way we see it people will either buy these plugins
as a way to support our general work on GStreamer or they buy it cause they need something legally licensed in
their jurisdiction.

4) The Indemnity clause demands that the buyer (licensee) indemnify you against anything. Where's the part where you promise that you have legally licensed the patents that you are implementing, thus indemnifying *us*?

The clause is meant as a general clause relating to issues such as 'if you install the codecs on the computer running your local nuclear plant and it blows up its not our fault' kinda events. That said I see when re-reading it that its not to well worded atm. I will talk with our lawyer to get it rewritten to make this more clear as it could be read to even cover claims of patent infringement the way it is currently worded.

5) Is the software distro-aware? If I install it on Fedora, will I know if some livna RPM tries to blow away a library?
no, its not

Re:why we are releasing these codecs

[schwaang]

The way we see it people will either buy these plugins as a way to support our general work on GStreamer or they buy it cause they need something legally licensed in their jurisdiction.

Thanks for making this very clear. This has been a missing piece for linux-on-the-desktop, so I hope the marketplace rewards you!

I'm going to buy the "everything" bundle myself for my windows-to-linux converts.

Worse than that

[Anonymous Coward]

it stops distributions that distribute where ffmpeg is legal if the distributor also wants in to the US (etc) market.

e.g. SuSE (even when a german company) didn't do MP3.

Fluendo = "Streaming Penguin"?

[Kadin2048]

I guess a native binary blob is slightly better than a MS coded binary blob

It's significantly better, actually. Not because it's technically superior (although it may be), but because it can legally be rolled into a commercial version of Linux. Right now, you can't legally distribute a Linux distro with multimedia support (at least not in the U.S.), because they depend either on MS DLLs (obvious copyright problems) or patent-encumbered free implementations (which can't be distributed with the distro for legal reasons).

This makes Linux into a second-rate desktop OS, even if you're willing to pay for it, because it means key features don't work out of the box. There have been exceptions to this from time to time (Xandros, Lindows), but they weren't well accepted by the community, possibly because they tried to leverage their use of proprietary codecs as an advantage over other Linux distros, rather than against Windows -- not a good way to make friends.

A company which wasn't involved in the actual production of a distro, might be in a good position (assuming it dealt with everyone on the same terms) to produce codecs that could be incorporated into (a non-free, pay-per-copy) version of any distro. E.g., someone could take Ubuntu, add the codecs (paying Fluendo, obviously), and sell the result as a package, suitable for pre-installation. I don't think this would violate GPL either, if the codecs were built in a way that didn't require linking or otherwise producing a "derived work."

In short, Fluendo could be in a position to be ESR's "Streaming Penguin." In that paper, he discusses some of the major problems facing Linux as a marketable desktop OS, and the lack of modern multimedia capabilities are a real deal-breaker. In fact, the lack of multimedia capabilities are more of a weakness, than simply being free-as-in-beer is a strength; people are obviously willing to pay for an OS that works, but one that doesn't work out of the box (or works only after fiddling around with some shady instructions involving PLF mirrors) won't fly, even if it's free.

While people here on Slashdot may not regard having to manually install LAME, Xvid, Flash, and the Win32 codecs as a significant problem, it's one of the many reasons why you can't go out and buy a Dell pre-configured with Linux as a home computer. Even if there wasn't Microsoft trying to torpedo it before it gets going, I'm not sure customers would accept anything that didn't work right, right out of the box. Fluendo could, if they play their cards right, be a big benefit to the adoption of Linux.

Re:Fluendo = "Streaming Penguin"?

[physicsnick]

While people here on Slashdot may not regard having to manually install LAME, Xvid, Flash, and the Win32 codecs as a significant problem, it's one of the many reasons why you can't go out and buy a Dell pre-configured with Linux as a home computer.

Exactly. I think one of the biggest problems with w32codecs, libdvdcss, etc. is that they are too easy to install. People don't realize that in many countries (most notably the United States), installing these codecs is NOT LEGAL; they just install the packages on their Linux box and complain about distros not including them by default. The critical application of Fluendo is for OEM installs of Linux; instead of charging the utterly ridiculous $60 or whatever price for Windows XP, they can now markup $4 for a codec license and finally offer the layman a completely functional operating system.

Re:Fluendo = "Streaming Penguin"?

[BiggerIsBetter]

If Novell was smart they'd leverage their MS agreemend, buy this Fluendo shop, and use their codecs in their SuSE Linux Enterprise Desktop (SLED) distribution.

Re:Fluendo = "Streaming Penguin"?

[evilviper]

[...] or patent-encumbered free implementations (which can't be distributed with the distro for legal reasons).

WTF? Why is everyone talking about patents like they're magic?

Do you know the "legal reasons" that "patent-encumbered" code isn't distributed with Linux distros? BECAUSE THEY DON'T WANT TO PAY for the patent licenses.

If they did, they could include slightly modified programs like MPlayer/Xine/etc., legally. Why pay for Fluendo's codecs, which include the cost of patent license fees, when you can just pay the patent license fees directly, and distribute popular, open source software?

Re:Fluendo = "Streaming Penguin"?

[Anonymous Coward]

it's not just the codecs that are missing. you also need the activeX that runs the codecs. I have the windows codecs on my Mac but still lots of content does not play.

Re:Fluendo = "Streaming Penguin"?

[iburrell]

It is quite likely that the patent holders won't license open source codecs even if people are paying money for them. Or they may restrict how the licensed versions can be distributed even if the source is available. The Fluendo MP3 plugin is only patent licensed for the binary version. The source is available under an MIT license but people building it from the source themselves aren't licensed under the patent. They must get the blessed version from Fluendo. These new plugins are binary-only which suggests that the patent holders would not allow the source to be available. It would be much nicer if we could buy patent licenses for the codecs and then use them for any open-source implementation. But the patent holders don't seem to like that.

Re:Fluendo = "Streaming Penguin"?

[evilviper]

It is quite likely that the patent holders won't license open source codecs even if people are paying money for them.

You're just completely wrong.

The relevent patent holders have no say in the matter.

MPEG-2, MPEG-4, h.2646, AAC, mp4, etc., are handled by MPEG-LA specifically to ensure open, documented, non-discrminatory license terms and fees.

VC-1 is an SMPTE standard, preventing Microsoft from exerting any control over licensing of that (major) codec, as well.

That takes care of the vast majority of codecs in popular use today.

These new plugins are binary-only which suggests that the patent holders would not allow the source to be available.

No, it suggest Fluendo doesn't want them to be open source. That's all.

Great, but not what I hoped for

[damienl451]

I guess plenty of people (or companies) will be glad to have a 100-percent legal alternative to using Windows DLLs, but what the community badly needs now is a legal DVD player. Right now, the only way to watch DVDs on Linux in most Western countries is breaking the law (i.e. the DMCA or its local equivalent), which is clearly a no-no for most users, including myself. The saddest things is that the software is available (PowerDVD Linux), but Cyberlink will only sell it to OEMs. Couldn't someone strike a deal with them and release it ?

And yes, I'm aware that Mandriva includes it, but it's not my distro of choice (and I'm not sure the RPM would work as-is on other distros).

Re:Great, but not what I hoped for

[evilviper]

Right now, the only way to watch DVDs on Linux in most Western countries is breaking the law (i.e. the DMCA or its local equivalent), which is clearly a no-no for most users, including myself.

There's nothing illegal about libdvdcss. The DMCA specifically allows reverse engineering for compatibility.

Oh yeah, and whatever you do, don't look here: http://www.linspire.com/lindows_products_details.p hp?package_name=los-xine-ui

Bundle it with a even nicer player than xine

[Anonymous Coward]

Bundle it with an even nicer player than xine/mplayer (something that doesn't crash .1% of the time), add DVD support, and sell it separated or bundle it with DVD players, as powerdvd is distributed. That would give an even better value and fill a really need spot ("legal" dvd playback in linux/solaris, not that I think that if I buy a DVD it is not legal to play it in whatever way I want).

Ah, somebody's been paying attention!

[Idaho]

Looks to me like somebody has been paying attention to ESR's World Domination 201 article.

Putting aside whether you like it or not, sometimes you have got to admit the man has a point :)

Re:Ah, somebody's been paying attention!

[Anonymous Coward]

Looks to me like somebody has been paying attention to ESR's World Domination 201 article. Putting aside whether you like it or not, sometimes you have got to admit the man has a point :) Aww knock it off, Eric. We'll still read your stuff without the posts to Slashdot. Sheesh!

This is a good idea

[Anonymous Coward]

I like to have source code as much as the next guy however...

does my girlfriend, my family, the average Joe/Jane user care about the details. NO they don't care about the source code. They don't mind the idea they they have to pay a little money for software that WORKS. If wide spread mass adoption of Linux is going to be a reality then things like common MPEG-2/4 video just has to work. Codecs, 3D graphics drivers, audio drivers, Web browsers, email clients, Office applications, printers, USB mass storage, connecting to cable/DSL modems via Ethernet or USB these are all things that the computer just has to do. The end user will I would venture 90%+ NOT CARE IF IT IS COMMERCIAL CLOSED SOURCE OR OPEN SOURCE OR FREE OPEN SOURCE THEY JUST WANT IT TO WORK. Proprietary codecs, drivers, and so on are just fine and should not be shunned. Open source can have just as many "bugs" as closed source. For now most home computer users are more then happy to shell out money to Microsoft because they have a product that comes on computers and just plainly works. I am an advocate of Linux and Open Source Software however I am not about to bash Microsoft because simply for the vast majority of home users it does exactly what the need/want it to do. A goal for the Linux community should be to come up with something that is just as simple to use with no harsh learning curve or "oh Linux does not do that because WMA / MPEG-2 is just M$ being stupid and or evil" the response will be "I just wanted to watch the video email of my grandkids I want my Windows back thank you" Software needs to be People friendly and just work with little to no customization frustration.

Re:This is a good idea

[Ash-Fox]

If wide spread mass adoption of Linux is going to be a reality then things like common MPEG-2/4 video just has to work.

And what doesn't just work in ffmpeg? Which is used in mplayer, vlc, gstreamer (available as a plugin gstreamer-ffmpeg)?

Obligatory Ogg Theora Post

[Cygnus v1]

Use Ogg Theora! It's a free and open spec.

Re:Obligatory Ogg Theora Post

[evilviper]

Use Ogg Theora! It's a free and open spec.

It's ridiculously CPU-intensive, and has terrible quality as well.

Use MPEG-1! It's free, open spec, and works on nearly every video player ever made.

Re:Obligatory Ogg Theora Post

[draethus]

Ogg codecs are used, just look at what's happening in the gaming industry. Heroes of might and magic 3 used MP3s for music, recent versions of Unreal tournament and Heroes of might and magic 5 use ogg vorbis and ogg theora.

Anyone surprised that

[Anonymous Coward]

Ffmpeg just sprouted support for WMV3 (aka Windows Media Video 9), and then these people appear from virtually out of thin air? I smell something not entirely unlike fish: a handful of people can't possibly reimplement a set of highly developed video decoders, even from specification, in a couple of months.

Let's see if these codecs turn out to have bugs that correspond 1:1 to bugs in ffmpeg...

Re:Anyone surprised that

[Wesley Felter]

If you read Fluendo employees' blogs, they've been working on these codecs for a year or so. Also, when you license Windows Media from MS they give you the source code AFAIK. All Fluendo did was port MS's code to GStreamer; they didn't try to reimplement it.

How Much is That?

[zifferent]

Could they please provide their prices in American dollars?

Not everyone lives on the other side of the pond, you know.;)

Re:How Much is That?

[Anonymous Coward]

Try google:

http://www.google.com/search?q=28+EUR+in+USD

I think you're on the right track...

[StressGuy]

Speaking for myself, I've been using Linux for a while now and, these days, it is the primary OS I use on my home desktop computer. Windows is still there, but for legacy programs that I still need and I switch off the internet whenever I start Windows (it's WinME).

If you can give me a Linux distribution that is supported and comes bundled with legal implementations of all the codecs at a reasonable price point, I'll buy it. Throw in something like Cedega for games or Crossover Office for other apps (if I even need it at this point) so I can get rid of my Windows installation entirely and that would be an ideal solution for me.

Proprietary, non-free applications are going to be essential to getting Linux into the mainstream. Software sellers aren't interested in an OS they can't sell software to run on.

Re:I think you're on the right track...

[evilviper]

If you can give me a Linux distribution that is supported and comes bundled with legal implementations of all the codecs at a reasonable price point, I'll buy it.

It's called Linspire. They've been doing it for a while.

Re:I think you're on the right track...

[Anonymous Coward]

wine 1.0 will be essential to getting linux into the mainstream, and not linux binaries without source code provided.

How would you sell that?

[Kadin2048]

If you have read Christian (Fluendo most visible VP) blog post, they plan to provide upgrades via distribution upgrade/installation system (apt-get, yum, etc.)

How would this work? Does APT or YUM work via authenticated HTTP connections? They'd obviously need some way to keep just anyone from sticking "apt http://fluendo.com/updates nonfree" into their sources.list and grabbing their software -- and I don't think their business model is going to fly if they attempt to do shareware.

Maybe they could sell people digital certificates which were used in the authentication process, allowing them to access the repositories; that seems like it might be viable. Not perfect (because you'd need to keep people from sharing the certificates, but at the same time you wouldn't want to tie them to IP addresses, because that would impact traveling people or those on dynamic IPs), but I could at least see it being possible.

Re:How would you sell that?

[ADRA]

Or, they'll have an on-machine license that'll be checked when a flumotion-nonfree gstreamer stream gets triggered. How they'd handle leaked internet keys is a more interesting problem for them to solve. Maybe its a combination of online 'checkup' once and a while. Who know, maybe they don't have any of that and its just in the good-will of the populace to do the right thing.

Re:How would you sell that?

[jZnat]

Maybe they'll do what RAR does and just check for your rarreg.key file (~/.rarreg.key also works as a location) to see if you have a legit licence for the program.

What makes this better than ffmpeg etc?

[jonwil]

What codecs are these guys going to release that arent:
A.Supported by ffmpeg/libavcodec/whatever else through open source implementations
B.Supported already through binaries on linux (such as the RealPlayer binary codec libraries)
or C.Not used enough for people to actually care about support

his?

[Anonymous Coward]

VLC uses the same codebase for his own engine.

Don't anthropomorphise software.

It hates that.

Because OSS development IS better, honest.

[Dr. Manhattan]

Nvidia have paid people on the job, with the relevant experience. What makes people think that the oss community can do a better job than nvidia's own people, when they can't even keep their own codebases bugfree?

You were asking the right question up until the last bit. Nobody can keep their codebases "bug free". Humans make mistakes. I assume you're human, ergo you make mistakes too, right? There's probably no program on Earth bigger than twenty lines that's bug-free. Not even LaTeX, though it's been quite a while since anyone's found one.

But as to why "people think that the oss community can do a better job than nvidia's own people", it's because OSS development has, when comparisons have been possible, proven to be better-written than the commercial alternatives. There are objective tests that illustrate this, over and over.

(Note, I still think this Fluendo stuff or something like it is a good idea. But I still want the OSS work to go on, too.)

Re:Because OSS development IS better, honest.

[x2A]

There are many cases where oss does seem to be providing us with better software, many many cases, no argument there. But when you start getting more specialised, what matters is experience with the specialised target, not brute force manpower.

"Nobody can keep their codebases "bug free""

Too right, so wouldn't you want the people with the most relevent experience working on something, to reduce the likelyhood of bugs, and to recognise the things that cause bugs as quickly as possible when they turn up?

Re:Because OSS development IS better, honest.

[Dr. Manhattan]

There are many cases where oss does seem to be providing us with better software, many many cases, no argument there. But when you start getting more specialised, what matters is experience with the specialised target, not brute force manpower.

You'd think that, wouldn't you? Designing and implementing the software - you're right, in general. But in terms of finding bugs, frequently an outsider does better. It's practically a cliche in programmer circles that the bug you've been fruitlessly hunting for two days will be solved when a coworker looks over your shoulder for five seconds and says, "Oh, hey, there it is."

Many bugs are syntactic, not semantic. And someone who's familiar with the syntax, but not necessarily all the semantics, can spot things that others miss. Put in less technical terms, they see what the code actually does whereas someone who works with it constantly may just see what it's supposed to do.

And other bugs are related not to the problem domain per se but the way it relates to other domains. The Nvidia/Apple bug discussed here recently was due not to any deep graphics issue, but by improperly using memory - probably a signed value where an unsigned is required, so that hitting memory over 2GB does Bad Things. Yeah, I'd believe a relatively untrained programmer might spot that.

Redundant

[spoonboy42]

MPEG2 and MPEG4 have LONG been supported by ffmpeg, mplayer, xine, and vlc. As for Windows media, libavcodec in mplayer has supported decoding Windows Media 7 and 8 natively for some time now, and just recently support was added for WMV9 (aka VC-1, the latest Windows Media codec used on HD-DVD and Blu-Ray) using, again, a fully open-source decoder.

Beat G.729 for Glory

[Doc Ruby]

How about an alternative codec to G.729, that isn't constrained by the G.729 patent? That means a different algorithm, but which still gives the high quality and low network/CPU bandwidth to "telephony" audio that makes it the favorite for VoIP? With the patent, G.729 codecs are all proprietary, licensing fees costing at least $10 per simultaneous call leg capacity. Which is the greatest ecomomic bottleneck to the growing revolution in packetized voice: an $800 server with an $80:mo ($1000:y) FTTP connection can use otherwise free SW to carry up to 400 call legs, which cost $4000 for the codecs.

A $free codec competitive with G.729 were available, it could unleash "phone servers" the way that multi-IP httpd unleashed webservers, powering the entire Bubble that got you to read Slashdot.

Re:Beat G.729 for Glory

[Wesley Felter]

What about Speex or iLBC? Good luck finding other people to interoperate with weird codecs, though.

Re:Beat G.729 for Glory

[Doc Ruby]

Speex and iLBC aren't as hifi as G.729, not for all telephony (which is somewhat wider than just voice). If the new codec were $free, and its TCO as low as G.729 for quality due to low network/CPU bandwidth, everyone, especially carriers and an Internet full of small operating peers, would use it.

Thank you!

[XanC]

I deal with libraries in my line of work, and I think Linux and libraries are very compatible ideas. Librarians are also sensitive about respecting copyright, and I think codecs are possibly the biggest item slowing down adoption of Linux by librarians. Bravo!

No lindows

[Anonymous Coward]

It hasn't been Lindows for a long time now. Linspire and Freespire can be had with multimedia playback out of the box legally in the US. And they are the ones slowlly chipping away at OEM linux installs, which is eventually the only way you'll see "linux on the desktop" for a lot of people outside of geekdom, when it comes pre installed.

A little late?

[nukem996]

I am able to play every type of media on my system using open source codecs including WMV3, MPEG4, and everything else via ffmpeg. They might of had a shot a few years ago but who would want this now?

Good for non-i386 Linux

[ChrisA90278]

This is very good news for people with non-i386 Linux systems. For example Linux on SPARC, Alpha, ARM or PPC. Windows DLLs only run on i386.

I wonder if these Codecs could also be made to run under Solaris or BSD or other open source OSes?

Even for the people who use i386, this means there would be a legal codec so the big distos could include it with video players out of the box

Re:Good for non-i386 Linux

[comay]

According to Fluendo's site, support was announced for *both* Linux and Solaris (x86/x64 and SPARC).

Re:Good for non-i386 Linux

[evilviper]

Windows DLLs only run on i386.

What Windows DLLs do you need?

Looking at the MPlayer (essential) binary codecs, there are only a few dozen Win32 DLLs included now, and almost all of them are:

A) Already reverse engineered and natively working, just not on 100% of videos yet, so the Win32 version is still included.
B) Oddball codecs you're quite unlikely to see in videos in the wild to begin with.

Even for the people who use i386, this means there would be a legal codec so the big distos could include it with video players out of the box

There's nothing about a codec which makes it legal or illegal. The only issue in question is if license fees have been paid by the distributor.

A commercial Linux distro could just as easily license their own binary of MPlayer, as they could buy these codecs (the former would likely be cheaper, faster, etc).

That they don't, suggests they simply don't want to be responsible for paying the necessary fees.

Re:Sounds great. If...

[jitterysquid]

> I belive that is why they try to have stabilised API version - gstreamer is close to 1.0 and
> probably when it is finished the API for plugins will froze so that any new API will not conflict
> with old (like propertiary codecs) plugins.

Sorry, I don't buy it. "stable" is the last adjective I would use to describe most APIs. gstreamer may be massively different, I don't know how fast the winds shift in that project. Even if the API for gstreamer 1.0 is stable, how long is it before distros start shipping the next beta instead of the stable version? That will require Fluendo to be a bit more agile.

This isn't an indictment of the projects in question, or Fluendo, but rather me being *extremely* cheap and worrying about things that probably aren't really a problem. It's what I do.

> WTF has kernel to do with codecs anyway? Linux kernel has stable API for applications since ages.

Beats me. I don't know what the hell the kernel is up to these days. I keep seeing things osmotically moving back and forth between kernel-land and user-land I don't know what to think. I can certainly imagine a codec that needs tight integration with a particular hardware driver, once we all bend over and get our DRM innoculations.

Again, I worry about these things so that other people don't have to. It's like a service. Allow me to repeat my first phrase: Rah rah. I am happy for Fluendo.

Because it's easier than emigrating

[tepples]

Why pay for something that can be played natively by lavc, and by extension mplayer et al?

Because it's easier than paying damages when you are sued for patent infringement and lose. Because it's easier than moving to a country that does not recognize patents on communication methods that are defined by a novel data processing algorithm and relocating all your customers as well.

List of patents

[tepples]

They are covered by patents are therefore is a subject of USAGE restrictions. What patents you are talking about? It's not hard for anybody experienced with Google to find a list of MP3 patents, the ASF patent, a list of MPEG-2 patents (PDF), and a list of MPEG-4 patents (PDF).

And no, vlc and xine is no use for simple user. Even if so... type:

% yum install totem-xine You forgot the step of obtaining a patent license.

Re:List of patents

[ajs318]

You forgot the step of obtaining a patent license.

You don't need a licence if the patent is bogus in the first place.

Re:List of patents

[tepples]

You forgot the step of obtaining a patent license.

You don't need a licence if the patent is bogus in the first place. You forgot the step of hiring legal counsel and getting a judge to declare that the patent is bogus in the first place.

It's a patent licensing issue

[tepples]

What codecs are these guys going to release that arent:
A.Supported by ffmpeg/libavcodec/whatever else through open source implementations The only way for United States residents to lawfully use open source implementations is to leave the country.

B.Supported already through binaries on linux (such as the RealPlayer binary codec libraries) Are the w32codecs lawful to redistribute under copyright and patent law?

Cool

[bberens]

Without going off on some religious tangent, I just wanted to say that I am appreciative of this support for the linux community. I'm pretty happy with my windows binary codecs, but if this works well and has a fair price I may pick it up to show my financial support for companies which support linux.

Re:Sounds great. If...

[kosmosik]

> Sorry, I don't buy it. "stable" is the last adjective I would use to describe most APIs.
> gstreamer may be massively different, I don't know how fast the winds shift in that

Gstreamer is (should be) designed to be modular and backwards compatible so older plugins should work with new versions. Gstreamer is partly developed by Fluendo so they do have some inpact on its design and decisions.

Also you've ommited part of my comment regarding mplayer - mplayer is something like the described setup (but illegal). It uses some binary plugins (think Windows DLLs), changes hell lot and is quite different on diferent distros - but still it works with these binary codecs. Is'nt it similar situation to Fluendo releasing some libraries and other software to hook into them?

(...)

> I can certainly imagine a codec that needs tight integration with a particular
> hardware driver, once we all bend over and get our DRM innoculations.

But you are talking like 2 years from now - please focus on current stuff. :) Maybe in 2 years everybody drops DRM and some new whizz-bang-open codecs will be developed. Same speculation as yours.

Public good

[ichthus]

That may be true if you don't live in a capitalist society. There are many people (I suspect, as I am one) who enjoy using OSS, but don't believe that absolutely everything must be done in accordance with RMS's communist (and I use that word in the sense of the theory of communism, not the Soviet Russia sense) ideals. I'm a firmware engineer by trade and, while I enjoy using Linux, I do believe that there is a place for closed-source, money making, IP-protecting code -- especially in a capitalist setting, where company A makes money and provides jobs because they do something company B can't.

And, I believe that most people are at least somewhat ok with the idea of closed-source whether they realize it or not. I don't hear many people complain that their BIOS, or HDD firmware or [insert any other peripheral]'s firmware is closed-source. (Yes, I'm aware of the Linux BIOS project.) Really, many of these pieces of software are lower-level than even the kernel and are located in potentially dangerous areas of the system architecure.

So, while I may agree with your statement in a near-sighted way, I believe that ultimately closed-source is necessary for companies, and the people who work for those companies and the companies that sell products to those people, and (deep breath) the people who work for the companies that sell products to the people who work for the companies that have closed-source software. And, the people who live on the welfare provided by the taxes paid by the people of said companies. In short, capitalism. Yes, Linksys opened it's firmware and has arguably sold more routers as a result. But, not all of their IP (the HDL that went into making their (or Broadcom's) ASICs, their PCB designs, etc.) was opened. If that were made available, how could that possibly benefit them? It wouldn't. It would hurt them as a company, and then their employees, and ultimately the "public good."

Better idea; get tough

[Anonymous Coward]

Instead of cursing the darkness, let's light a flamethrower.

Reverse-engineer, de-compile and publish. One "blob" at a time. We have a right to the Source Code -- let's take it by force, if we have to. No harm to life or tangible property, please; just this false concept of "intellectual property".

As one of us falls, let another step up to take their place.

Still no legally licensed CSS playback

[jopet]

I like to see that businesses are popping up around Linux and that licensed, legal software is becoming available for media playback. However, one of the major issues is still not resolved: there is no legal way to play back CSS protected DVDs. It seems the same company is planning to provide a player, but the original plan to publish it in the second half of 2006 did not seem to hold :)
There are similar issues with mp3-encoding and many video encoders.

Re:Still no legally licensed CSS playback

[evilviper]

there is no legal way to play back CSS protected DVDs.

There's nothing illegal about libdvdcss. The DMCA specifically allows reverse engineering for compatibility.

Oh yeah, and whatever you do, don't look here: http://www.linspire.com/lindows_products_details.p hp?package_name=los-xine-ui

Maybe they wouldn't have to?

[Kadin2048]

It seems like Linux would be a pretty rough platform to try and implement DRM (of any sort, media or software) on. Unless they have mandatory BLOBs in the kernel, it seems like their license-checking and enforcement codes would be pretty trivial to bypass, by intercepting the software's communications with other parts of the OS and the outside world. And if they depend on kernel-space stuff, I think they'll hurt adoption tremendously (I wouldn't compile anything into my kernel that was there just to provide copy-protection or DRM to some piece of software -- if I wanted that, I'd run Windows!).

I think at the end of the day, they'll basically have to be relying on the good-will of the community, or on their relationships with OEM installers. The more I think about it, the more I think this latter group will really have to be their bread and butter. It's where Linux fails hard, right now: if you want to sell Linux machines in the U.S., you basically have no choice but to sell them crippled, or face patent lawsuits. These guys could sell the rights to the codecs to Dell, etc., who would build that cost into the price of a Linux-preinstalled system, and in return users would get a machine that was ready to go out of the carton.

That doesn't require any draconian license enforcement, because would-be pirates of their software, are probably the people who are just going to download ffmpeg and Xvid from the Penguin Liberation Front (who I am no way slandering -- they rock) anyway. If the people at Fluendo have any brains at all, they'll realize they can't compete with FOSS for the Linux users who are out there right now. But what they can do, is go after the users that FOSS hasn't managed to pick up so far, and are still using Windows.

Re:DRM, codecs... same thing

[Kazoo the Clown]

Yes, it's not as if we don't have enough codecs out there. If a given piece of media can't be viewed with a legal and free codec, I just don't need to see it. There's plenty more things to view out there that don't require restricted codecs. Rather than accomodate proprietary restrictions on media play, we should avoid them and help restricted codecs to die out. Making them available on Linux suggests that it is Linux' problem that a codec won't run on it, rather than the codec's problem. In either case it's the users problem, but covering up the fact the codec is restrictive is doing noone any favors except those who wish to exert proprietary controls over media-- and while they may have that right, we also have the right to ignore them. As soon as some Madison Avenue exec realizes that their ads are missing a segment of the market by using a restrictive codec, they'll toss it and use something else, and ad men can be pretty darn influential in media production...

Vote with your feet-- against restricted codecs.

found answer to question 5: no

[schwaang]

5) Is the software distro-aware? If I install it on Fedora, will I know if some livna RPM tries to blow away a library?

I found the answer to this part on Christian's blog.

The codecs are distributed inside tarballs together with instructions. We realize this is not as painless an install as one could wish for. But doing packages for a million and one distro's was not a plausible solution either. That said we are working on a codec installer/updater which will automatically download and install any codec bought in the shop.

I hope Fluendo will consider providing packages for the top three distros. But I guess distro communities could provide a source package a la jpackage.org for Sun Java.

Another thing to keep in mind

[goaty_the_flying_sho]

This exact same functionality is _free_ under Linux, right now.

we could sandbox it

[r00t]

Write a number to /proc/self/seccomp to sandbox an app.

It's even good for fully open software, just in case there might be a buffer overflow.

it's only 64 kbit/sec, sheesh!

[r00t]

Phone voice quality is crappy enough as it is. Please don't make it even more difficult to understand what the other party is saying. Just send the uncompressed data.

Now, if you were planning to do 24 kHz 16-bit end-to-end, maybe it would be a different matter.

Re:it's only 64 kbit/sec, sheesh!

[Doc Ruby]

You really don't know much about Internet telephony, do you? G.729 is quite good for telephony, as is GSM. They use around 10Kbps. That's 6x as many calls on a single CPU, which runs maybe 100-200 simul calls. To serve 10,000 simul calls, that's 50-100 servers. Which is a lot more costs in HW, operators, network bandwidth, facilities square footage, power, etc. Those savings are what makes VoIP cost-effective enough for little players to challenge the big telcos. In which change lies the revolution. Don't fight it, go with it.

Re:it's only 64 kbit/sec, sheesh!

[r00t]

Isn't GSM is the sucky compression used on digital cell phones? No thanks.

Q: "can you hear me now?"
A: "no, I don't fear your cow"

Telephone quality is shit already. Make it better, not worse.

Look, you will PAY MONEY for the shit quality. People have to repeat things when the quality is bad. That takes time, making calls longer. People are more likely to get pissed off then they struggle to communicate, and pissing off your customers is not good for the finances.

Re:it's only 64 kbit/sec, sheesh!

[Doc Ruby]

Again, you really don't know much about telephony. The GSM codec is pretty good: the robocalls I've run with them got rave reviews for their audio quality, because I tweaked everything (including the EQ) to use the GSM quality. The problem you're talking about has nothing to do with the codec, and everything to do with mobile telcos cheaping out on their transponder tech, locations and configs.

The economics of codecs are quite well established, and the audio quality of GSM and G.729 is fine. I know all about the obvious customer service details you're laboriously pointing out, too.

So stop trying to act like you know what you're talking about. You're just spreading FUD. Learn something accurate about this subject before you try teaching it to anyone else.

Click...