OSNews

Pardon me, this is going to be a little bit offtopic (feel free to mod me down): Can someone tell me the differences in compatibility with other licenses between GPLv2 and GPLv3 (if, of course, there are any)?





edit: clarification (i presume)

Edited 2007-09-22 19:27

Why would it be unfair to consider a company's larger strategic position when assessing a particular tactical move?

Microsoft's history demonstrates a certain goal and their strategies support that goal. I think it's perfectly reasonable that in the face of that strategy, the smaller individual, and sometimes seemingly unrelated moves get scrutinized more closely.

And just to ensure I get modded down, here's a political analogy: Can you really say that George W. Bush performed his national guard duty, in the face of all the over whelming evidence, just because one of the single piece of that evidence turned out to be a mistake (even though it was never denied)?

Should we ignore the bigger picture because of a smudge?

These are two shitty arguments, as far as I'm concerned.

The first argument at least has some logic behind it, and even though I don't find it a rather good argument, a point does get made. It is preferable if a license's name corresponds to its content - but is that really that big of a deal? I mean, are 'GPL' and 'MIT License' descriptive?

The second argument is where they get really overboard. Incompatible? What? As if the GPL is compatible! Compatibility is a two-way street in my book, and in that respect, the GPL is anything but compatible.

Weird arguments.

Thom,

I really do see some merit in the first argument. Where as GPL and MIT are not really descriptive, those names are not trying to be. The use of the adjective "permissive" in the name of Microsoft's license is likely *intended* to cause confusion. In the same way that OOXML (Office Open XML) was intended to be confused with Open Office's XML format.

It may be too late for the OSI to fight license proliferation without being hypocritical. But I believe it is right for them to refuse to participate in a plot to interject further confusion into the OSS license landscape intentionally.

That said, their second argument is indeed a big, flying, turd.

"Where as GPL and MIT are not really descriptive, those names are not trying to be. "

I agree with you on MIT, but GPL? The name "General Public License" would seem to suggest "Public Domain", which GPL definitely is not.

The name "General Public License" would seem to suggest "Public Domain", which GPL definitely is not.

What straw man are you attacking? Not that I claim to be an expert on the GPL, but I think it's terms do a decent job of keeping control of code and other resources in the public domain.

Just what do you think public domain means anyway?

The second argument is where they get really overboard. Incompatible? What?

Not really. License proliferation is a big concern, and the OSI should have been reigning this in years ago. The Computer Associates license that is OSI approved is just one amongst a few licenses that should simply be struck off. It hasn't just appeared out of nowhere with these Microsoft licenses.

Incompatible licenses build walls between open source communities, and doesn't break them down. Quite frankly, I don't think the OSI should be encouraging wall building.

I also believe that the OSI should just state what is on peoples' minds. They are under no obligation to simply waltz in and approve new, incompatible and pointless licenses developed by a company who doesn't have a track record with the OSI, isn't sociable and can't get along with its peers, and hasn't helped the open source movement that the OSI has tried to help build either - quite the opposite, in fact. Many people seem to believe there is some sort of conflict there or something. There isn't. The OSI owes Microsoft nothing. It is up to Microsoft to prove themselves.

As if the GPL is compatible! Compatibility is a two-way street in my book, and in that respect, the GPL is anything but compatible.

People are actually using the GPL.

Hopefully, all this will make the OSI function better as an open source organisation, and get companies, not just Microsoft, to prove their open source credentials first.

"The OSI owes Microsoft nothing. It is up to Microsoft to prove themselves.
...
People are actually using the GPL.

Hopefully, all this will make the OSI function better as an open source organisation, and get companies, not just Microsoft, to prove their open source credentials first."

---------

OSI owes Microsoft nothing, but they do owe themselves their own credibility. And they don't appear to be making the arguments that you're making. If your arguments are the real issue, let OSI have the guts to say that rather than hiding behing what appear to be BS arguments (I say "appear", because as I said earlier, the article provides virtually no details).

You say Microsoft needs to "prove themselves", well I wouldn't be surprised if Microsoft has released lots more code under MS-PL than the total code released under many of the current OSI licenses, many of which were created for a single product or single group of related products, and are still only used for those products today. Obviously the amount of GPL code dwarfes the amount of MS-PL code, but the amount of MS-PL code dwarfes that of many OSI licenses.

There has been lots of code released under MS-PL (I just recently incorporated a MS-PL'ed zipfile lib into one of my hobbyist projects), including major code such as the DLR (which is one of the reasons Miguel's team got Moonlight running so quickly). Other well-known MS-PL projects are IronPython and IronRuby.

Here's the thing: The MS-PL code that is in use today is "open source code" whether OSI certifies it as such or not. Now, if OSI doesn't want that code to have their impramature, that's up to them, but Microsoft and others are releasing MS-PL code with ever increasing frequency, and it wouldn't look good for OSI for there to be such a large and important set of open source projects out there that lacks OSI's impramature. In such a scenario, the OSI "brandname" begins to lose relevance.

So, IMO, it would behoove OSI to have legit technical (rather than political) reasons to reject MS-PL, if they end up doing so. And maybe the reasons *are* legit, but we need more info regarding that.

But if the bottom line is indeed political (as you are saying), then OSI needs to say that explicitly.
(The problem with doing that is that certain OSI members urged the submission of MS-PL, so to now say that they reject it for political reasons leaves OSI open to the charge that this whole thing was merely a game; just a charade to get the opportunity to publicly reject MS-PL as an "open source" license.)

For my own part, I'm inclined at this point to believe that this isn't political, that there is some technical issue at hand, but we don't have enough info to know for certain either way. But we can still have fun speculating on both the technicalities and the politics.

Edited 2007-09-23 00:25

OSI owes Microsoft nothing, but they do owe themselves their own credibility. And they don't appear to be making the arguments that you're making.

I think they know what's wrong, but because they want to try and appear impartial they're not saying what they really think.

Here's the thing: The MS-PL code that is in use today is "open source code" whether OSI certifies it as such or not.

Obviously Microsoft doesn't believe that to be true, otherwise they wouldn't be trying to get the OSI to approve their licenses.

So, IMO, it would behoove OSI to have legit technical (rather than political) reasons to reject MS-PL

Politics always comes into this with these organisations. The OSI isn't a black and white organisation that polices licenses in a black and white manner.

Microsoft has never had any legit technical reasons for not cooperating with standards bodies, using other peoples' standards properly or working with the wider open source community effectively and the OSI needs to take into account whether these licenses will benefit the wider open source community, or divide it. Quite frankly, it works both ways. There's no conflict as far as I'm concerned.

The GPL is by far the most common Open Source license.

As far as compatibility with other open source licenses goes, contrary to your assertion the GPL is quite middle-of-the-road.

Another question raised was:

http://gplv3.fsf.org/wiki/index.php/Compatible_licenses
http://gplv3.fsf.org/dd3-faq

The GPL is compatible with a vast range of open source software projects ... the overwhelming majority in fact.

I'm not sure, but I had heard that the same could not be said of the MS-PL. I must admit that in reading the MS-PL I can't actually determine what any incompatibility might be.

Edited 2007-09-23 00:38

It's incompatible with the GPL and the code can't be ripped out of the file and put into another file under a different license. But the MSPL is a file-based license, so you could combine it with GPL files into a module if the GPL would have allowed combination. Basically the permissions that MS-PL demands run afoul of the permissions that the GPL denies.

Microsoft doesn't want to give its code to the GPL crowd. It may be a move to weaken the GPL, but why should that matter? The key is that it seems to be an Open Source license under the OSD and the code could be quite useful in non-GPL Open Source projects running on Linux or elsewhere.

Because that's the most widely-used FOSS license. It is also what prevents open-source from being co-opted by MS. MS is trying to divide the FOSS community to better conquer it.

These are strategic issues. Unless MS makes *very real* moves to develop trust within the FOSS community, then the community (which by and large is GPL-friendly) will not be fooled by this wolf-in-sheep's-clothing.

Bear in mind that this is purely because of the intent of the "more permissively licensed project".

That is what I never understood about recent whining from BSD-licensed projects. Apparently it is perfectly fine from BSD point of view for a commercial company to take BSD-licensed code and make it closed-source, and give nothing back, but it is not OK for a GPL project to do something more open than that?

Why do BSD-license proponents feel that is OK for commercial exploitation of their work, without any giving back, but it is apparently not OK for any attempt to ensure through licensing that the code remains forever open?

There is something decidedly suspicious about this position. There is seemingly an orchestrated pseudo-astroturfing campaign here ... the thrust of which seems to be to come up with anything, anything at all no matter how bogus, to try to paint in a bad light any attempt to preserve open-ness of code in an on-going sense.

It is almost as if there is a huge push to undermine anything that does not allow open source code projects from being taken up as a closed source program at some time in the future.

I wonder who could possibly be behind such a campaign?

Are you implying that TdR is somehow acting in favor of Microsoft? That's an amusing supposition, but doesn't pass the sniff test. Theo also complains when companies use OpenBSD and don't contribute back, so it's not like he's one-sided here. The reason it's worse for the GPL is twofold: first, companies don't like bad publicity, so they probably quietly deal with the issue if he ever raises it; second, as open-source licenses, GPL and BSD are "brothers." They are both part of the broader Open Source community, so it seems like a greater betrayal of the spirit of the whole Enterprise for one side to take from the other without giving back. You almost expect betrayal from a stranger, but it hurts when it comes from family.

If TdR is so keen on getting contributions back, why doesn't he just simply put that condition in his license? Why doesn't he simply make a license which says "if you take this code & modify it, then you must give your modifications back"?

If he put that condition in his license, then he could complain without hypocrisy about lack of reciprocity with the GPL ...

...

... oh wait.

Clue to TdR ... if you just asked everyone equally exactly the same as that which you ask of GPL projects, then your license would in fact BE the GPL, and you wouldn't have a problem.



Are you sure it doesn't pass the sniff test?

I am still smelling something very suspicious about TdR's position from where I sit.

Edited 2007-09-23 07:28

Ahem...this is about an *OSI* decision, not a FSF one. If the OSI rejects Microsoft's license on the ground that they are not compatible enough, that has nothing to do with the FSF. The two organization are separate and - if I recall correctly - far from agreeing on philosophical matter.

"""
this is about an *OSI* decision,
"""

Of course. I referenced the FSF since it is the organization which trumpets the GPL's wide ranging "compatibility". I didn't really want to imply that OSI did that, because I'm not sure that they do. At any rate, no matter who does or does not say it, the GPL's "compatibility" is very one sided. In fact, I think it's probably more accurate to consider the GPL *incompatible* with most every other license. To me, "compatibility" implies a two way flow of code.

That may be your definition, but it doesn't appear to be the OSI's. In any case, they seem to claim that the MS-PL is *less* compatible than the GPL. Of all the criticism I've read so far on this page, I haven't really seen anyone contradict this (with clear arguments to support their position).

"""

"""

If the OSI is making that claim, then I would say that the burden of proof is on them. Molly is right, though. More detail is needed.

I like the OSI. But in a very real sense, it is their credibility, more than any license, which is really on trial.

"That may be your definition, but it doesn't appear to be the OSI's. In any case, they seem to claim that the MS-PL is *less* compatible than the GPL. Of all the criticism I've read so far on this page, I haven't really seen anyone contradict this (with clear arguments to support their position)."

And I've yet to see anyone *support* this with clear arguments to support their position, including the article itself.

Nobody can contradict the claim because there's been no evidence presented to support it. You can't cross-examine evidence that has not been presented.

You appear to be ready to accept OSI's decrees at face value, with no evidence required. But most of the other posters here want details. Once the details are disclosed, then people will at least be *able* to contradict them. I assume the details will be disclosed; if they aren't, and OSI just declares MS-PL to be too imcompatible without releasing any supporting evidence, then they leave themselves open to charges of using "secret evidence" ala Kafka.

I had expected the MS-Pl license to be accepted without a hitch, since it is quite permissive. Not as far as MIT/BSD but none-the-less still quite permissive.

I agree that we can't analyse the situation yet, since the article contains virtually no technical details.

Taking a look at MS-PL http://www.microsoft.com/resources/sharedsource/licensingbasics/per... does not show anything which gives me second thoughts on accepting the license. It is more permissive than GPL and less than MIT/BSD (but much closer to MIT/BSD than GPL).

I'm surprised.

EDIT: MS Community License is more tricky but still quite easy to accept according to OSI-rules http://www.microsoft.com/resources/sharedsource/licensingbasics/com...


The only reason for not accepting them is the duplicitous behavior of Microsoft in regard to FLOSS.

Edited 2007-09-22 20:14 UTC

Which I personally also find a lousy reason, but at least it's a reason I could understand.

I hate it when people beat around the bush. I wish the OSI as a whole was just honest and up front about it: we don't want Microsoft licenses OSI approved, because we hate Microsoft. At least then you could accuse them of having balls.

I agree that the OSI should be honest.
They should say something like, we think Microsoft are using misleading words, and by the way, we think that Microsoft hates everything we stand for and would like to destroy us so we find them hard to trust.
That would be refreshing.

I don't know, I think you're assuming many things here. After all, maybe they're rejecting the licenses because they don't thing are open enough to receive their approval?

In that case, it wouldn't be because they don't like Microsoft, but simply because they see Microsoft's effort for what it is, i.e. another attempt at weakening the F/OSS movement. Because make no mistake - you and others may not care about politics, but MS does. A great deal, at that.

"We're not the Judean People's Front, we're the People's Front of Judea!"

RMS (free) and ESR (open) both seem to criticize the other. Outside and to many inside, both seem to be essentially the same thing.

"We're not the Judean People's Front, we're the People's Front of Judea!"

emmmm yes, but, by the same token....

"He is NOT the messiah, he is a very naughty boy !"

The patentlefting is not a problem at all. It is explicit in GPL3 and MS-PL, whereas it is implicit in GPL2, MIT, BSD and some other licenses.

IPL and CPL also contains explicit "patentlefting".

"""

"""

You were doing really well until you got to the delusional part which I have quoted. Torvalds, Raymond, Stallman? They are not a clique of conspirators. That list reads more like guest stars on Celebrity Deathmatch.

If the license meets OSI criteria, it should be approved. If the OSS definition has problems, as OSNews user Butters has observered in recent weeks, that definition should be revisited and existing OSI certified licenses reevaluated.

That said, considering the source of this license, I don't mind the OSI being especially vigilant to ensure that Microsoft dots all their i's and crosses all their t's regarding its approval.

>I didn't say they're conspirators. They just got their pockets filled by u-know-who.

Don't be shy, say it!

Linus Torvalds get paid by Transmeta for a long time and now from the Linux Foundation

Richard Stallman lives from the money he as made by selling Free Software and give some courses, the MacArthur award and he gets paid to held speeches.

For Eric Raymond i wasn't sure but on his homepage he wrotes "I do consulting under the Thyrsus Enterprises service mark."

Where is the problem?

By the way, RMS is definitely not a "OSIs puppet".

>As I wrote, OSI not only trashes itself by this decision but also take the bigger risk to split open-source community

Nothing new. The OSI has already split the Free Software community 1998.

You just have to look at all this flame wars on all the news sites etc. We really don't need a discussion about MS licenses to split the community. This already happens many years ago. But it's not necessarily a problem.

Edited 2007-09-22 21:07

Don't be shy, say it!
I already said that earlier. A few key figures get their pockets filled in order to be pastors for thousands of sheeps.

By the way, RMS is definitely not a "OSIs puppet".
Never said that. OSI is a puppet itself, as we are witnessing. RMS is just a puppet of people paying his bills. That's not OSI.

Nothing new. The OSI has already split the Free Software community.

You just have to look at all this flame wars on all the news sites etc. We really don't need a discussion about MS licenses to split the community. This already happens many years ago. But it's not necessarily a problem.
When you have rules, you always got splitted. But I didn't mean that. If OSI proves that rules they set are just fog in order to kick out the ones they don't like (regardless actual merit or compliance), it would just position itself as defendant of Linux-open-source.

Be aware that there are thousands of Windows open-source developers. I don't think it would be sane to make them feel they're class-B citizens in OSI because that would of course led the creation of a "W-OSI", where they could still be class-A citizens.

While I have my own reserves to the whole OS community, there's no reason to do that. Unless you want to defend non-W big guys.

"""

"""

No. I don't know who. Spell out your accusations explicitly, please.

"""

"""

Just to be clear, when I say "vigilant", I'm referring to the OSI's desire for the license name not to be intentionally misleading and confusing. We have *enough* license confusion as it is. The OSS definition may not have been crafted to deal with that particular abuse. But I would support the OSI in blocking that attempt at creating unnecessary confusion; Changing the license's name to something more accurate should be trivial if Microsoft is sincere in their desire to have the license approved.

As to objecting to the license on the grounds of its incompatibility with other OSS licenses... I've already covered that in this thread. That I would not consider vigilance. I would consider it a major change of venue.

I agree with you that we must not custom-tailor the OSS definition to exclude individual players. We must stick to our basic principles.

Or, to state it in a more whimsical and moralistic way, "we cannot triumph over evil by becoming evil".

Edited 2007-09-22 21:02

No. I don't know who. Spell out your accusations explicitly, please.
I have no "secret news" to tell: we all know who's funding most OS projects (and why, and how, and who). I already named two of them in my post, perhaps most important ones. When Big Blue stated they were going to "invest" 100millions in Linux, did you think it was for marketing maybe? lol

Changing the license's name to something more accurate should be trivial if Microsoft is sincere in their desire to have the license approved.
I don't get the point. "Permissive" term is just a simple term. Plus, that license is *actually* very permissive. This is just a pretext and you know that. You should demand that OSI used no pretexts: today they do against Microsoft, tomorrow it could be against you.

As to objecting to the license on the grounds of its incompatibility with other OSS licenses... I've already covered that in this thread. That I would not consider vigilance. I would consider it a major change of venue.
Yes, we mostly agree about almost anything in this story. But I'm just upset that OSI tries to fool me, while you don't look too upset about it.

I agree with Tom: if OSI wants to play politics, they should have balls to do that.

You've got it backwards. Open Source doesn't mean anti-Microsoft, rather Microsoft means anti-Open Source. This much has been clear for years (thanks in part to Microsoft upper management's own declarations).

MS isn't any friendlier to F/OSS nowadays - it has simply decided to wage a PR war to make it appear as if it was. On the core issues, they haven't made any significant movement towards open source. They have simply paid lip service to it.



Rubbish. The day MS Office runs on Linux, and supports ODF, then we'll be able to talk seriously about MS "playing nice with OS" (by OS, I assume you mean Open-Source...you should use FOSS, or at least OSS to avoid confusion). Until then, it is only natural that FOSS advocates continue to view MS with suspicion - to do otherwise would be incredibly naive.

How do you know the intentions of Microsoft? You seem to be so perspicacious. Microsoft is a huge organization with over 80,000 people (maybe about 10,000 devs). They don't necessarily have one viewpoint. And it's a very team-based culture. What works for one team (CLR and other dev tools or web-based technologies) would not work for other teams. Office is a highly successful commercial product that is produced by a relatively small group of people (you'd be surprised at the size of the WinWord team). Why should you reject the whole company from Open Source when it's obvious that the community can benefit from the parts that are interested in the source. This isn't the way to grow the community and strikes me as unreasonable.

Because I've followed the computing scene for decades, and I understand basic economics.

I also don't call myself "PlatformAgnostic" when virtually *all* my posts here are about defending Microsoft.



I am, thanks for noticing.



It doesn't really matter what dev #8947 thinks about upper source. As in any corporation worth its grain of salt (and MS certainly is), strategic decisions are made by upper management - and Microsoft's stance towards open-source is a profoundly strategic issue for the software giant.



Well, that's because you're parroting MS' PR message, i.e. that *they're* the victim. The reality is that MS, as a company, has been *very* hostile towards F/OSS, and I for one am not convinced that it has suddenly become F/OSS-friendly because of a few timid gestures. Given its past (and present, with regards to OOXML), MS is going to have to do a lot more to gain any kind of trust with the community.

>I always considered myself more an "open source fan" than a "free software fan". Now what should I do? It's clear that I can't really trust the OSI, because they're basing an important decision on facts that have nothing to do with their own principles. Sad...

I would recommend to go back to the roots and to support the Free Software movement.

>Incidently, does anyone know if the FSF has commented on this license?

FSFE has commented it back in 2005: http://mailman.fsfeurope.org/pipermail/press-release/2005q4/000120....

(It looks like a quite positive statement)

>If Microsofts license preserves the 4 freedoms, then I doubt you would see the FSF falling over itself to NOT approve it as a free software license.

I agree. But FSF doesn't care that much about "approving" licenses. They look at license only if it is necessary for themself or if someone especially ask them. So if you want to know more just write a mail at licensing(at)fsf(dot)org and i'm sure you will get an answer and they will list it at their list[1] as Free Software license or non-Free Software license after they have made a decision.

The FSF always said that it is not about "anti-MS"[2]. Also the FSFE stated[3] that it is not against MS and that they would even help them if they make a step toward Free Software!

[1] http://www.fsf.org/licensing/licenses/
[2] http://www.gnu.org/philosophy/microsoft.html
[3] http://www.wsa-conference.org/video/greve.mov

Edited 2007-09-23 09:41