samba is a biggy and pretty vital to Novel and their deal since most interoperatiblity between windows machines
and *Nix machines is provided through this service, so iirc Novel will have to fork samba

Thank you sir! Hopefully many more projects will follow your lead.

So doesn't this mean that smbfs is now dead? Or stuck at 3.0.x? Since the Linux kernel will not be going GPLv3, from my understanding of what Linus has said.

We could soon see NAS, etc. vendors forking Samba 3.0.x so they don't have to deal with GPLv3 ... then the whole fun mess of was this patch copied from 3.2+ into the fork or does it just happen to look the same since the fix is only sensibly done in a limited number of ways.

When will 3.2 be released, and when will Novell include it in SUSE?

So what? Does switching to GPLv3 change anything about [LOTS of vendors re-packaging Samba and selling it as NAS's and such]? There's nothing those vendors do with Samba that will have to change because of the new restrictions under GPLv3. So who cares?

There is one thing: if a NAS manufacturer has put DRM on the device blocking its owner from changing the underlying software, that manufacturer either won't use the GLPv3 Samba, or to be allowed to use it he will have to at least disclose in full the instructions needed for one to change the underlying software (or more probably, remove the DRM "feature" entirely).

So, yes, this is major news for everyone developing/manufacturing/deploying/using/etc. anything Samba-related.

Well, this message's (actual) parent is gonna get modded Redundant because I was a bit slow in composing my reply. (My excuse: if I keep my session open too long, every clicked link pegs my CPU for what feels like half a minute, new tabs longer.)

It would be nice if slashcode's Preview would inform a poster about other replies that were made to the parent posting since the new posting was started or last previewed. That might cut down on the number of redundant follow-ups where some posters compose slower than others and don't think to click the parent's message number to reopen it in a new tab to check for other replies first.

The limitation is not because Linus is some asshole, but a practical realisation that GPL3 cannot be retro-fitted to existing kernel code.

Only the owner of the code is allowed to assign the license and people made submissions to Linux under the GPL2-flavored license. Linus has no authority to release all the Linux code under a new license since he only owns a small percentage of the code. There have been thousands of people submitting to Linus under the GPL2-flavored license and it is impractical, if not impossible, to track those submittors down and secure a GPL3 agreement from them.

Sure, Linux could adopt the SMB strategy of committing to make future release of Linux GPL3 (eg, say Linux 3.0). Then all submissions into that new version would have to be GPL3. Practically though, many of the big players in Linux might prefer GPL2 over GPL3 and that could force a fork.

Technically speaking, wouldn't this manufacturer only have to make it possible to run modified Sambas and other modified GPLv3 bits they might of used? Say the rest of the userland was some modified BSD code, that could still stay shut?

In practice, I imagine such DRM would be done by signing an entire firmware image. Future practitioners of such DRM would just have to isolate the bits that really need to be sooper-sekrit if they want to use GPLv3 code.

G'luck with that, is all I have to say. Having once waded into the Samba codebase trying to ferret out a bug, I can't see them getting very far unless they manage to snipe one of the core developers. Samba is giant and the amount of resources needed to backport every bugfix (to say nothing of feature additions) and be at all subtle about it has got to exceed just accommodating the new license. And don't forget Samba 4 is on the way, so you lose ADS too if you want to fork 3. No, I think they'll either put up or shut up.

From the GPLv2:

From the GPLv3:

I highly doubt it. I don't know the legalese well enough (or law) to say for sure, but one of the main purposes of the GPL3 is to prevent exactly that.

Now, they could modify the kernel to implement the DRM, and release an unmodified Samba >=3.2. Since you could implement pretty much any DRM system in the kernel (and it's probably the best way to do it, short of hardware measures), Samba doing this stops very little. But it is cool, I think. Even though we may not be able to circumvent the DRM, we're free to do something I've wanted to do to half the electronics I own, and make modifications to it that have nothing to do with circumventing DRM, and just lets me use the product the way I want.

As such, I believe the BSD style licenses are more idealistic than copyleft licenses, but less effective.

Actually, it still think it's the other way around because in YOUR ideal world the BSD license would be enough and there would be no need for all the bits in the GPL that are there to somehow ensure everlasting freedom for the code. This makes YOU idealistic, which is good, but the BSD license itself seems to me to be far from idealistic, it seems more for people who don't want to be bothered to ever think about the consequences of their choice of license. In that light it is still the GPL that is the license for the idealists because at least a percentage of them have the hope that somehow it will make a difference, exactly because it forces people to think. Every time a company has to consider if they can use a particular piece of GPL code they need to do a bit of introspection and who knows what can happen after years of that...

Here's how at GPLv3 is playing at my Fortune 50 company, who makes contributions to lots of OSS projects.

1. We make tivoized device.
2. OSS project which we use for the device switches to GPLv3.
3. We start looking at other alternatives, decide project is no longer useful to us.
4. We stop contributing to the project.

I anticipate there will be a lot of corporate contributors quietly exiting their Samba involvement in the near future. A few of these exits will see some pub when a major developer switches employers as a result, but most corporate OSS contributions will disappear with a whimper. GPLv3 will return OSS to the original egalitarian ideals, but it's probably going to reverse all the corporate uptake that has happened in the last few years. If you're about to say that this only applies to tivoized devices, you should take a look at the market and see that the majority of corporate uptake of OSS has been in internet-connected appliances.