YESTERDAY we gave a sample as large as we could gather of the coverage regarding Microsoft's patch. We showed without great effort that Microsoft had turned this into a public relations event (manufactured self-congratulatory interviews, press release, correspondence accompanied by a lot of publicity). Some of the loudest supporters were the Microsoft-sympathetic bloggers and journalists, based on empirical evidence alone. In addition, one notable supporter from Linux was Greg Kroah-Hartman. Yes, he praised it, but many people forget that he is a Novell employee (who also bashes Ubuntu).
Novell played a major role in making this happen. Novell Fellow Greg Kroah-Hartman, who leads the Linux Kernel Device Driver project, approached Microsoft and helped guide them though the submission process.
The long and short of it, though, is that Microsoft has contributed roughly 20,000 lines of code to the Linux kernel related to their Hyper-V drivers.
The most vivid headline on yesterday's news that Microsoft is releasing various Linux kernel modules under the GNU Public License may not have been the most accurate. That would be InfoWorld's "Linux slips into Microsoft's warm, deadly embrace," which cast an agreeable horror-movie glow over the proceedings.
Fun stuff, but despite Randall C. Kennedy's fine and impassioned argument that this is all an embrace-and-extend plot to allow Hyper-V to feast on the blood of the open-source movement, that's probably not where things are heading.
The Linux driver project team, a part of the Linux kernel development group, is known for pursuing companies and asking them if they want their drivers included in the kernel, according to Greg Kroah-Hartman, the Linux driver project lead and a Novell fellow.
Another kernel community member noticed the [Microsoft] drivers and pointed them out to me," he said. "Through the contacts I have at Novell and through the Microsoft/Novell interoperability agreement, I contacted Microsoft and worked out the details."
In a statement, Novell said: "As a leading Linux solutions provider and an active player in the Linux community, Novell was influential in bringing this about and has worked closely with Microsoft to make this a reality. Under the direction of Novell Fellow Greg Kroah-Hartman, who leads the Linux Kernel Device Driver project, Novell proactively engaged with Microsoft to provide valuable guidance and feedback to the Open Source Technology Center, which enabled the team to contribute the code in a way that was in line with community processes."
I’m proud of Novell’s role in this. I’m proud that our partnership brought clarity on the technical optimization need. I’m proud of the personal role played by Novell Fellow Greg Kroah-Hartman and the Linux driver project.
And the vendor operates a Linux/Windows integration lab with partner Novell.
[...]
The driver code that Microsoft open sourced and submitted to the Linux kernel was first developed and certified specifically for Novell's Suse Linux and Red Hat Linux.
Microsoft’s release of 20,000 lines of Linux drivers, under the GPL, may be a shocker to headline writers, but it’s actually smart business.
This is about Novell.
Comments
Jose_X
2009-07-22 12:38:00
With that covered and working, they have methods to give their integrated stack advantages over everything else (using "bugs", bit rot, etc). [Ie, EEE].
And along the way, they can (use proxies to) hold the patent dagger above users'/customers' heads. [Ie, make FOSS/Linux expensive to use if you don't own Windows license.]
Jose_X
2009-07-22 12:41:28
*****
I think people are underestimating the patent threat from APIs. I would like the FSF to say something about that instead of just Csharp.
All APIs can be a patent trap, and yes we can't in practice avoid using some API or other. Of course, the threat level depends on the odds that the creator of the API took out patents to parallel the API and that they will use those patents against FOSS. Here, again, Microsoft is the big worry, and this means the dotnet API (whether you use csharp of Fsharp or dflat).
Any API can be a patent trap (in the US anyway) because of the way patents work and how general in scope they can be. See for example: LINK 1
After the API patent trap comes the extension patent trap.
I am working on a better explanation of the patent extension trap. [See LINK 2 and LINK 3 ] This is where you are allowed to get patent protection for the core spec (only). [Like the Microsoft Community Promise would do if it weren't such a large loophole.] The reason this is a trap is because of how patents are granted.
Many software patents have an initial claim that sets a given amount of context. If you look, you will see that claim 1 is large while several further claims follow that are each short but leverage claim 1. What this means is that if a new invention (claim 1 of patent 1) requires (eg) 10 items all to be present in order to avoid prior art, then those 10 claims plus one more will also avoid prior art (claim 2 of patent 1). And this can be extended across patents (or by renewing patents and adding these extensions) because the new short patent claim's extra requirement on top of the other patent's 10 requirements would be an invention (claim 1 of patent 2). Since the same company would own both patents, the company would not tax itself nor go for an injunction (duh).
If you wonder about "obvious" tests. Well, you might not be paying attention to what is being granted patents. Plus, any obvious test would likely not consider what a reasonable team of engineers could come up with in say 5 years (<20 years). PLUS, do you have a million bucks per patent to fight each patent to hope to show obviousness? I don't think many patents get struck down on obviousness tests.
[I'll provide LINK 1-3 when I find them, hopefully very soon. Use freepatentsonline.com to search for some software patents to get a taste of them. Eg, http://www.freepatentsonline.com/7013469.html . Despite all the noise on the page, all that really counts are each of the individual claims. ]
*****
OK, I forgot to finish the prior comment.
So what is the extension trap?
Well, you can be offered real protection (the Microsoft Community Promise doesn't even get that far) for the invention in claim 1 patent 1 (the 10 requirements) but not for anything else (like claim 2 patent 1 or claim 1 patent 2 or anything else).
So if you are working under the context of those 10 requirements thinking you are safe, you will find that to start stepping on patents, you now have to acquire potentially only a single new requirement instead of the usual say 10 requirements that might be needed to hit a distinct/independent patent. [Remember: that large initial set of requirements exists since avoiding prior art means building something sufficiently unique.]
By Microsoft providing inducements to the base requirements (ie, making core core dotnet an "open standard" and giving it "patent protection"), they entice people to use their core invention. This means Microsoft will then have a field day anticipating all requirements you might want to ever create in an app. They only need to guess one requirement at a time, however [each single additional requirement can be captured in a new dependent claim].
From the point of view of probability, it's much more likely they will guess a single requirement than say 10 requirements exactly. Put another way, if they guess 3 of 10 new features of your new mono app, that would be three patent claims you would be violating. If you don't use mono, then they would have to guess say those 10 things before you start to violate even a single patent.
Further, Microsoft, by virtue of their head start and the initial secrecy of each new patent, is just about the only one that can set those traps on that core invention and cash in on them.
Did I just hear, "help strengthen monopolies?"
Did I just hear, "Microsoft has a huge investment riding on dotnet patents?"
Remember that as mono apps grow in number, Microsoft's patents grow in value, and they have a lot of patents. I'm sure some could multiply in value 100 fold or more.
I would actually not be shocked to learn that Canonical/Novell or others had arrangements to profit from some of those patents already.
So if you use mono, (a) you make it easy to fall prey to extension patents that leverage mono, and (b) the owner of the patent will almost surely be Microsoft or someone doing close business with Microsoft (eg, a contractor that likely gives Microsoft a cut of the proceeds or a patent troll).
Preserve monopolies? Make FOSS unusable without paying tolls? I don't think so.
*****
eet
2009-07-22 12:53:27
Well, I can only repeat myself that there's a very nice and quiet OS out there, called GNU/Hurd. Check it out sometime, it's got 'RMS' all over it.
DiamondWakizashi
2009-07-22 15:07:51
DiamondWakizashi
2009-07-22 15:21:14
Novell is a Mono/Moonblight company, not a Linux company.
roadelland
2009-07-22 15:48:13
DiamondWakizashi
2009-07-22 15:55:06
roadelland
2009-07-22 16:07:39
DiamondWakizashi
2009-07-22 16:45:26
"Adult Conopid flies attack bumble bee workers feeding on flowers or in flight, then lay an egg in their abdomens. Later, the egg hatches and the larva develops inside the bee, until it fills the whole of the bee's abdomen and kills it"
http://www.newscientist.com/article/mg13017754.100--science-how-a-parasitic-fly-can-fool-bumble-bee-studies-.html
Novell uses a similar parasitic strategy against Linux:
Novell minions attack Linux users trying to have a wonderful computing experience, infecting them with the Mono/Moonblight parasitic infection. Later, it develops inside the Linux distro and kills it.
Linux users should have as much enthusiasm for Novell as a bumble bee would have for a Conopid fly.
zatoichi
2009-07-22 18:51:40
There's one little point I'm a bit hazy on:
Later, it develops inside the Linux distro and kills it.
I think we should definitely assemble a list of every single Linux distro that has met its untimely demise at the hands of this horrible parasite.
1. ?
Hm. That didn't take as long as I thought it might.
Now, why are you taking us on this walk through "The Discovery Channel", exactly? As I've commented elsewhere, if you're that worried about potentially Microsoft encumbered patents, patents we know Microsoft says they hold, into our free software systems, why aren't you campaigning to have the kernel removed from the distros?
You know, it occurs to me that if you're that concerned that Microsoft is going to come and sue you, there's a very, very simple way to guarantee that will never, ever happen to you:
Buy yourself a nice copy of SLED. Seems like pretty cheap insurance.
zatoichi
2009-07-22 22:37:39
zatoichi
2009-07-22 19:02:17
DiamondWakizashi
2009-07-22 20:07:42
What Microsoft Linux patents? The Microsoft criminals refuse to show even a single mythical patent the Linux supposedly violates?
"Buy yourself a nice copy of SLED. Seems like pretty cheap insurance."
I'll pass, Microsoft patent FUD doesn't scare me.
"Hey, Mr. Wakizashi? Is there any truth to the rumor that you’re Mark Fink?"
No, I would never use a ridiculous name such as "Mark Fink".
roadelland
2009-07-22 20:22:41
DiamondWakizashi
2009-07-22 21:34:16
zatoichi
2009-07-22 22:33:02
Well, you don't show your cards in a poker game, either, until you're ready to figure out who gets the chips. But I doubt they have anything that's terribly worrisome or that they're likely to use, if there's actually an issue at all.
“Seems like pretty cheap insurance.”
I’ll pass, Microsoft patent FUD doesn’t scare me.
Seems to scare you when it's about Mono, though, doesn't it?
No, I would never use a ridiculous name such as “Mark Fink”.
Watch that: my grandmother's last name happens to have been "Fink". I Am Not Kidding Around Here. It's the German name for the bird we call a "finch", so spare me the second grade humor, please.
The question is would you post email under a name like "Mark Fink"? Or post comments places under names like "NastyMicrosoft", "BoycottMicrosoft", or "cyberphoenix"?
Is this a new astroturfer name game?
Not at all, this is the New Boycott Novell, where debates are polite, and we all have to log in. I think Elland and I are just interested in knowing who's who.
Ã¥ÂâºÃ£Â®ãŠæ¯ÂãÂâ¢Ã£ââã¯åÂâºÃ£ââãâ¬Å½Diamond Wakizashiãâ¬ÂÃ¥â¼ã°ãªãÂâ¹Ã£Â£ãŸã ãâÂãÂâ ãÂÂ？
Oh, sorry. I'd just sort of assumed you spoke Japanese, and just realized you hadn't responded to my other questions! Maybe I'm wrong. I was just saying, "I don't suppose your mother named you "Diamond Wakizashi"".
Or do I owe Mrs. Wakizashi an apology?
zatoichi
2009-07-23 01:59:32
(And any decisions about what goes into GNOME 3.0 will be made by the 3.0 team, not by Novell, so stop making nonsensical claims. Of course, if you really wanted a say about GNOME 3.0, you might consider showing up at a conference, getting involved in a project, etc. Those are the folks who get to make decisions. We do not award points for "repetitive whining".)
You really seem to be emotionally overwrought about this whole thing, Mr. Wakizashi, talking about "parasites" and such. I hope that you're not playing with sharp objects in such a state of mind.
(How many times a day do you wash your hands, if you wouldn't mind my asking? I'm just curious.)
roadelland
2009-07-23 02:52:49
Answering a question with a question, implying that the person is an astroturfer, and trying to change the subject, that usually only happens here when someone is too close to the truth.
It was just a question, nothing more. Other people will read this and they may possibly think that you are those nyms because of the reasons I gave above.
zatoichi
2009-07-22 22:49:19
Novell supports Linux like a tick supports the dog it’s attached to.
This is actually a complete inversion of demonstrable facts. Novell supports the GNOME Foundation in a big way, they sponsor major conferences like the Gran Canaria Desktop Summit, and Novell engineers, like Jeffrey Stedman and Miguel de Icaza contribute to the community in important ways.
I don't suppose you've ever gone to a community conference, have you, Mr. Wakizashi? Are you a Friend of GNOME? Do you even support the FSF?
(I still haven't heard from Roy as to whether or not he supports the FSF. He's not on the donor lists from 2004 to the present. I'm an FSF-Europe Fellow, über-fancy personal certificate in a smart card, fsfe,org address and all, and I support the GNOME Foundation, the EFF, Creative Commons, not to mention the Long Now Foundation and others.)
Hey, here's an idea for you guys! I suggest you write to the nice folks over at FSFE and tell them all about how I'm a "secret Mono advocate" and "not really a Linux guy".
zatoichi
2009-07-22 22:50:57
DiamondWakizashi
2009-07-23 00:07:54
"GNOME 3.0 may have more Mono apps"
http://www.itwire.com/content/view/25967/1231/
Novell "supports" Gnome by poisoning it with Microsoft's Mono garbage. Like the Conopid fly wants the bee to thrive until it's larva is fully developed, Novell wants Linux thrive to feed the Mono parasite.
Novell wants people to create programs for Mono not Linux, Linux is merely the Mono parasite's host.
Mono is based on Microsoft's defective technologies, there is a very real patent threat(like FAT). Linux is not based on Microsoft's defective technologies, their patent threats against Linux are ridiculous, which is why they won't show the patents.
zatoichi
2009-07-23 02:11:44
Wasps, poison, parasites... You seem to be a very unhappy person, Mr. Wakizashi.
...there is a very real patent threat...
Oh, Mr. Wakizashi, one benefit of growing up in a family of lawyers is that you learn that there's no such thing as "facts", there's only "evidence".
Now, the nice lawyers at Canonical (who are quite able lawyers, I know some of them) seem to have a decidedly different opinion than yours. I know the lawyers at SFLC have their opinion as well, but I've read them both, and based on my legal understanding (which is not insignificant), I would tend toward Canonical's opinion, personally.
( A point to note: a lawyer, when pressed will always make the case his client demands be made: the customer is always right, and the SFLC's client is arguably pre-disposed to a particular conclusion, shall we say. Canonical, on the other hand, has no axe to grind either way--I don't believe that Mark Shuttleworth cares strongly if Mono is in Ubuntu or not (and I know Mark. Do you, Mr. Wakizashi?)
zatoichi
2009-07-23 02:17:20
I mention this only in passing.
NotZed
2009-07-22 23:18:46
"here, now you only have to write your drivers once for windoze, and you can avoid all that gpl nastiness ..."
zatoichi
2009-07-23 02:14:47
What?