"GPL3 is a (commercial) plague - anything that uses any GPL3 library MUST comply by GPL3 and any license that is not GPL3 becomes GPL3, so Apple had to abandon SAMBA - if they integrate a SAMBA interfacing gui into their OS (which they did under GPL2), they immediately are required to release the entire OS under the GPL."
Can I have some of your 'shrooms please, they must be *really* good:-).
This is so far from the truth it's easier just to point at this:
We already have this for all release streams. Only one person (Karolin) the release manager can commit to a release stream - all other changes have to be associated with an open bug report and go through at least one code review before going in.
Master branch is still more open for rapid development, but we're moving in the direction of more and more code reviews.
I don't like doing code reviews, but I like having my code reviewed as other people find my bugs:-).
It wasn't you Guy:-). Must have been versions subsequent to you working on it. I do know who it was who needed some help with the NetApp code, but it's not an appropriate subject to discuss here.
> I am curious; who was it that finally fixed samba to > respect/understand file streams(NTFS) and resource > forks(HFS) so we could switch between them?
The streams work was done by Isilon. As for mapping to HFS, if that's possible it was certainly Apple. Unfortunately they didn't push it back upstream so we were not aware that it had been done.
Samba has always been GPLvXX-or-later licensed. Everyone who contributes is aware of this fact. Currently we're v3-or-later. Who knows if there'll ever be a v4, but if there is we'd have no problems moving to it as it's already implicit in all code contributions.
> So, a couple interesting things related to CIFS came to light > not long ago. The first was playing with Macs. They suck at > it, horrible performance. In 10.4 they couldn't even talk to the > NetApp with CIFS at all, they could talk to Windows servers > but slowly. NFS worked but it was a disaster trying to get > permissions to work right. > > We figured this was in part because they use Samba which > is not necessarily the fastest thing out there, and was > originally designed for reverse engineering SMB, not a > reference CIFS implementation like the NetApp.
Oh dear. Is NetApp marketing really this good ?
Firstly - the Mac client is written by Apple and is called smbfs, it's not Samba at all. It is Open Source code, released by Apple in Darwin. I know the engineers who write it, and they're really good and have been working on it for a while, so I'm sure it's gotten better since you tried it.
Secondly, "a reference CIFS implementation like the NetApp." !!!!
Oh. My. God.:-). NetApps CIFS implementation was written well after Samba, with some judicious peeks at the Samba code in order to implement the hard stuff (this was before Microsoft released their docs). That's ok, that's one of the reasons the Samba code is out there, so people can learn from it.
As for being "a reference CIFS implementation". Just try running Samba's smbtorture4 test suite against NetApp's "a reference CIFS implementation" to discover how much of a "reference" they actually implemented.
> "Do you want to give a justification of why you are willing to > keep using the GPL even though it means companies like > Apple are not willing to use your software?"
Sure. Apple have never been a major contributor to Samba. Other companies like Google, IBM, Cisco, Symantec (and many other NAS vendors and OEM's) are happy to contribute and use Samba under GPL (both v2 and v3), so the GPL is still a vital tool to share development costs between companies who want to *contribute*, not just use.
IMHO Apple want to keep their ability to sue over software patents, which the GPL is designed to make difficult.
If you've been following the news recently I hope you see why this is becoming more and more important for Free Software code. Sort of off-topic, but software patents really are a threat to all software engineers and they don't distinguish between open source or proprietary code:-(.
Samba isn't owned by one entity, and so re-licensing under special terms isn't possible. It's one of the advantages (or disadvantages, depending on your point of view:-) of having distributed copyright ownership.
I won't say who offered, but tridge was once offered a multi-million deal to "sell" Samba to a networking company (a long time ago, before people understood what Free Software/Open Source really meant:-).
Apple has been moving away from the GPL in all it's forms for a while now. They just got around to us (I'm guessing we were pretty high on the list once they got rid of gcc:-).
Jeremy.
Re:That would be an antitrust violation
on
The Case For Oracle
·
· Score: 3, Informative
FlorianMueller wrote:
> I oppose software patents and particularly the use of patents against free and open source software. > In Google's case, we are however talking about a company that is very much pro-patent as far its own > patents (especially the search engine patents) are concerned and just despises everyone else's when used > against it. Now Google effectively calls on the community, but Google doesn't support the community in > the fight against software patents.
I know I'm a Google employee and therefore should be suspect in this (i.e. check my claims about Google, don't take them on trust), but the statement above is untrue. Google submitted an anti-software patent brief in the Bilski case. See here:
You might actually try *testing* the Solaris CIFS implementation, not just believing the Sun/Oracle press releases about it. You might find it's a bit.. lacking:-). What's their roadmap timetable for an SMB2 server for example ?
"Why is it, by the way, that your quite sudden anti-Microsoft slog has materialized only after you became a Google employee?"
Hahahahaha ! You mean all the work I put into the EU antitrust Lawsuit against Microsoft *before* I joined Novell (while I was at HP) was an example of pro-Microsoft activity ?
I don't think you're very familiar with my history:-).
Jeff was *definitely* one of the architects Novell/Microsoft deal, and had been part of the leadership for at least a year when it was finalized. I know. I was there.
Dateline: Wed Jan 27th 2010, San Francisco: 10:00am
Not since their release of the iWaterboard "enhanced interrogation" playset (tm) has so much excitement been generated over a new Apple product. The new iShackle (tm), demonstrated by Steve Jobs on an anti-DRM protester on stage at a packed San Francisco event this morning, takes customer lock-in to a new level. "With the new iShackle (tm), content companies will literally OWN their customers", said Jobs in front of an ecstatic audience. "No more pesky choice, no more confusing options, just pure, simple, buy this or ELSE corporate power". As expected, fans were completely delighted with the new Apple product. "I can't wait to buy the new iShackle (tm) and take it home and put it on !" said one excited member of the crowd. "Steve says I never have to take it off or think about other products EVER AGAIN !"
Not everyone was pleased with the announcement however. A Microsoft spokesperson said, "We've had the Microsoft zOBEY software for 2 years now, this isn't a new concept. Apple once again copies the market leader." Sales of the Microsoft alternative have been poor however, except for Microsoft employees, who have had use of the software mandated in their employment contracts since Steve Ballmer announced the product at a marketing event in Guantanamo Bay, Cuba in 2007.
Your argument that the Mono case is different than the Tom Tom case doesn't stand up I'm afraid. Microsoft only wants developers to develop for Windows..NET is an attempt to provide an environment so compelling that people continue to use Windows to get access to it. Mono weakens that case for using Windows, so attacking Mono makes perfect sense. Remember, Microsoft loves Open Source code that runs on Windows. Mono isn't designed to run on Windows - Windows already has a perfectly good.NET environment, therefore Mono serves no purpose in Microsoft's eyes. Promoting threats around Mono forces people who want to use.NET to do it on the "One True Platform" (Windows). Attacking Mono forces.NET developers to move to Windows, it doesn't mean an exodus from.NET. Just like in Microsoft's mindset the Linux kernel has no business implementing a Microsoft proprietary technology (FAT32) when people should be buying the Windows implementation instead. The parallels are obvious.
> In the past 3-4 years they've shown a much different attitude towards OSS developers..
This is true, they have. But they've also aggressively sued a company using Free Software over claimed patent violations in Free Software (Tom Tom). This negates all the attitude changes in the world. I actually feel sorry for the pro-Open Source people at Microsoft (and there are many). They're not in control and they are left trying to defend the indefensible from their masters. Hopefully this might change eventually, but I'm adopting a wait and see attitude. Actual lawsuits trump noble promises IMHO.
> When Microsoft owned the market they could do that where ever and when ever they wished. Now that they are losing ground everyday on > almost all fronts trying to do something like that would only give another reason for people who were thinking about using.NET to avoid it. > In short, it would be suicide.
It didn't stop them in the Tom Tom lawsuit, it wouldn't stop them here. As I've said before, Tom Tom changed *everything*. Microsoft lost the moral high ground (and yes they did at least have that position over patents) of not being a patent aggressor but only using patents for defense, stopped the rhetoric and became a direct patent threat to Free Software/Open Source.
Since Tom Tom, nothing they say about patents in Mono/.NET can be trusted any more. Only a legally binding non-assert offer can change this now.
I don't think I'm wrong on the GPL compliance issue. Even if there are few people with standing to sue (which I'm not convinced of), such a license from Microsoft would still violate the GPLv2 and put whoever agreed to that license into copyright violation. Now they might not get sued over it, but reputable companies don't knowingly do things like this.
Slashdot Mirror
We've handled that case since the mid 1990's...
Jeremy.
Creepy wrote:
"GPL3 is a (commercial) plague - anything that uses any GPL3 library MUST comply by GPL3 and any license that is not GPL3 becomes GPL3, so Apple had to abandon SAMBA - if they integrate a SAMBA interfacing gui into their OS (which they did under GPL2), they immediately are required to release the entire OS under the GPL."
Can I have some of your 'shrooms please, they must be *really* good :-).
This is so far from the truth it's easier just to point at this:
ftp://www.samba.org/pub/samba/slides/linuxcollab-why-samba-went-gplv3.pdf
and hope people read it than to try and rebut your ravings.
Jeremy.
We already have this for all release streams. Only one person (Karolin) the release manager can commit to a release stream - all other changes have to be associated with an open bug report and go through at least one code review before going in.
Master branch is still more open for rapid development, but we're moving in the direction of more and more code reviews.
I don't like doing code reviews, but I like having my code reviewed as other people find my bugs :-).
Jeremy.
Samba has stable SMB2 support which will be shipped in 3.6.0, which last time I checked is before 4.0..
It's already being used in production on some large sites (who helped us debug problems with it).
3.6.0 is going to be officially shipped before the next release of OSX.
Try again, it's not SMB2 support that's the problem :-).
Jeremy.
It wasn't you Guy :-). Must have been versions subsequent to you working on it. I do know who it was who needed some help with the NetApp code, but it's not an appropriate subject to discuss here.
Jeremy.
> I am curious; who was it that finally fixed samba to
> respect/understand file streams(NTFS) and resource
> forks(HFS) so we could switch between them?
The streams work was done by Isilon. As for mapping to HFS, if that's possible it was certainly Apple. Unfortunately they didn't push it back upstream so we were not aware that it had been done.
Jeremy.
Samba has always been GPLvXX-or-later licensed. Everyone who contributes is aware of this fact. Currently we're v3-or-later. Who knows if there'll ever be a v4, but if there is we'd have no problems moving to it as it's already implicit in all code contributions.
Jeremy.
Where in Sheffield ? (Just curious, my home town you see :-).
Jeremy.
> So, a couple interesting things related to CIFS came to light
> not long ago. The first was playing with Macs. They suck at
> it, horrible performance. In 10.4 they couldn't even talk to the
> NetApp with CIFS at all, they could talk to Windows servers
> but slowly. NFS worked but it was a disaster trying to get
> permissions to work right.
>
> We figured this was in part because they use Samba which
> is not necessarily the fastest thing out there, and was
> originally designed for reverse engineering SMB, not a
> reference CIFS implementation like the NetApp.
Oh dear. Is NetApp marketing really this good ?
Firstly - the Mac client is written by Apple and is called smbfs, it's not Samba at all. It is Open Source code, released by Apple in Darwin. I know the engineers who write it, and they're really good and have been working on it for a while, so I'm sure it's gotten better since you tried it.
Secondly, "a reference CIFS implementation like the NetApp." !!!!
Oh. My. God. :-). NetApps CIFS implementation was written well after Samba, with some judicious peeks at the Samba code in order to implement the hard stuff (this was before Microsoft released their docs). That's ok, that's one of the reasons the Samba code is out there, so people can learn from it.
As for being "a reference CIFS implementation". Just try running Samba's smbtorture4 test suite against NetApp's "a reference CIFS implementation" to discover how much of a "reference" they actually implemented.
Jeremy.
> "Do you want to give a justification of why you are willing to
> keep using the GPL even though it means companies like
> Apple are not willing to use your software?"
Sure. Apple have never been a major contributor to Samba. Other companies like Google, IBM, Cisco, Symantec (and many other NAS vendors and OEM's) are happy to contribute and use Samba under GPL (both v2 and v3), so the GPL is still a vital tool to share development costs between companies who want to *contribute*, not just use.
IMHO Apple want to keep their ability to sue over software patents, which the GPL is designed to make difficult.
If you've been following the news recently I hope you see why this is becoming more and more important for Free Software code. Sort of off-topic, but software patents really are a threat to all software engineers and they don't distinguish between open source or proprietary code :-(.
Jeremy.
Samba isn't owned by one entity, and so re-licensing under special terms isn't possible. It's one of the advantages (or disadvantages, depending on your point of view :-) of having distributed copyright ownership.
I won't say who offered, but tridge was once offered a multi-million deal to "sell" Samba to a networking company (a long time ago, before people understood what Free Software/Open Source really meant :-).
Jeremy.
Apple has been moving away from the GPL in all it's forms for a while now. They just got around to us (I'm guessing we were pretty high on the list once they got rid of gcc :-).
Jeremy.
FlorianMueller wrote:
> I oppose software patents and particularly the use of patents against free and open source software.
> In Google's case, we are however talking about a company that is very much pro-patent as far its own
> patents (especially the search engine patents) are concerned and just despises everyone else's when used
> against it. Now Google effectively calls on the community, but Google doesn't support the community in
> the fight against software patents.
I know I'm a Google employee and therefore should be suspect in this (i.e. check my claims about Google, don't take them on trust), but the statement above is untrue. Google submitted an anti-software patent brief in the Bilski case. See here:
http://en.swpat.org/wiki/Bilski_v._Kappos_amicus_briefs
for details.
Jeremy.
You might actually try *testing* the Solaris CIFS implementation, not just believing the Sun/Oracle press releases about it. You might find it's a bit .. lacking :-). What's their roadmap timetable for an SMB2 server for example ?
Jeremy.
Anonymous wrote :
"Why is it, by the way, that your quite sudden anti-Microsoft slog has materialized only after you became a Google employee?"
Hahahahaha ! You mean all the work I put into the EU antitrust Lawsuit against Microsoft *before* I joined Novell (while I was at HP) was an example of pro-Microsoft activity ?
I don't think you're very familiar with my history :-).
Jeremy.
Jeff was *definitely* one of the architects Novell/Microsoft deal, and had been part of the leadership for at least a year when it was finalized. I know. I was there.
But don't let facts get in the way of your post.
Jeremy.
Yet more proof, if it were needed, that once you reach the CXX level there are never any consequences for any of your actions :-(.
Jeremy.
Dateline: Wed Jan 27th 2010, San Francisco: 10:00am
Not since their release of the iWaterboard "enhanced interrogation"
playset (tm) has so much excitement been generated over a new Apple
product. The new iShackle (tm), demonstrated by Steve Jobs on an
anti-DRM protester on stage at a packed San Francisco event this
morning, takes customer lock-in to a new level. "With the new iShackle
(tm), content companies will literally OWN their customers", said Jobs
in front of an ecstatic audience. "No more pesky choice, no more
confusing options, just pure, simple, buy this or ELSE corporate power". As
expected, fans were completely delighted with the new Apple product.
"I can't wait to buy the new iShackle (tm) and take it home and put it
on !" said one excited member of the crowd. "Steve says I never have
to take it off or think about other products EVER AGAIN !"
Not everyone was pleased with the announcement however. A Microsoft
spokesperson said, "We've had the Microsoft zOBEY software for 2 years
now, this isn't a new concept. Apple once again copies the market
leader." Sales of the Microsoft alternative have been poor however,
except for Microsoft employees, who have had use of the software
mandated in their employment contracts since Steve Ballmer announced
the product at a marketing event in Guantanamo Bay, Cuba in 2007.
Your argument that the Mono case is different than the Tom Tom case doesn't stand up I'm afraid. Microsoft only wants developers to develop for Windows. .NET is an attempt to provide an environment so compelling that people continue to use Windows to get access to it. Mono weakens that case for using Windows, so attacking Mono makes perfect sense. Remember, Microsoft loves Open Source code that runs on Windows. Mono isn't designed to run on Windows - Windows already has a perfectly good .NET environment, therefore Mono serves no purpose in Microsoft's eyes. Promoting threats around Mono forces people who want to use .NET to do it on the "One True Platform" (Windows). Attacking Mono forces .NET developers to move to Windows, it doesn't mean an exodus from .NET. Just like in Microsoft's mindset the Linux kernel has no business implementing a Microsoft proprietary technology (FAT32) when people should be buying the Windows implementation instead. The parallels are obvious.
Jeremy.
The Microsoft Community Promise is not good enough. See this legal analysis for details:
http://www.softwarefreedom.org/resources/2008/osp-gpl.html
Microsoft lawyers are good enough to produce a better document than this, they just chose not to. See this document:
http://www.samba.org/samba/PFIF/PFIF_agreement.html
for a better agreement and an analysis on why all the terms in it are needed (especially the "Patents" section).
Jeremy.
aztracker1 wrote:
> In the past 3-4 years they've shown a much different attitude towards OSS developers..
This is true, they have. But they've also aggressively sued a company using Free Software over claimed patent violations in Free Software (Tom Tom). This negates all the attitude changes in the world. I actually feel sorry for the pro-Open Source people at Microsoft (and there are many). They're not in control and they are left trying to defend the indefensible from their masters. Hopefully this might change eventually, but I'm adopting a wait and see attitude. Actual lawsuits trump noble promises IMHO.
Jeremy.
AmaDaden wrote:
> When Microsoft owned the market they could do that where ever and when ever they wished. Now that they are losing ground everyday on .NET to avoid it.
> almost all fronts trying to do something like that would only give another reason for people who were thinking about using
> In short, it would be suicide.
It didn't stop them in the Tom Tom lawsuit, it wouldn't stop them here. As I've said before, Tom Tom changed *everything*. Microsoft lost the moral high ground (and yes they did at least have that position over patents) of not being a patent aggressor but only using patents for defense, stopped the rhetoric and became a direct patent threat to Free Software/Open Source.
Since Tom Tom, nothing they say about patents in Mono/.NET can be trusted any more. Only a legally binding non-assert offer can change this now.
Jeremy.
Yeah, lots of the original Samba code was written on Sun 386i's. Ah, memories :-).
Jeremy.
Unless of course the person in question is a *known* paid anti-odf shill from Microsoft. As in this case.
Jeremy.
I don't think I'm wrong on the GPL compliance issue. Even if there are few people with standing to sue (which I'm not convinced of), such a license from Microsoft would still violate the GPLv2 and put whoever agreed to that license into copyright violation. Now they might not get sued over it, but reputable companies don't knowingly do things like this.
Jeremy.