Switch to GPL

[neptop]

Takeover?
Bitcoin will always be open source, because it is now licensed using the MIT license. This means you can take the code and do pretty much everything with it. Everyone can therefore grab always use and add to it. It's something SOME (I am not talking about anyone in this forum) GPL advocates forget sometimes. I did so multiple times
You can't somehow take it away.

Rearing IP. I don't know too much about it, but isn't it somehow free, because there is no patent or something similar yet? As I said I have only minimal knowledge on this topic.

[Syke]

If the only library is closed source, then there's a project to make an open source one.

If the only library is GPL, then there's a project to make a non-GPL one.

If the best library is MIT, Boost, new-BSD or public domain, then we can stop re-writing it.

I don't question that GPL is a good license for operating systems, especially since non-GPL code is allowed to interface with the OS.  For smaller projects, I think the fear of a closed-source takeover is overdone.

Satoshi,

So you support people taking your code, modifying it to skim bitcoins off the miner, and then releasing the binary without releasing the modified source code?

[mizerydearia]

Syke: While I understand some of the community may not appreciate the manners in which a seemingly friendly community member has established their proprietary product, consider businesses that offer products or services for profit.  Now consider that one of those businesses or entrepreneurial individuals has established a kind of closeness within the community.  Instead of focusing negatively on their activities, they should instead be appreciated that they are nice enough to participate amongst the community directly.  From my observations, puddinpop is providing a valuable service and is handling the negative reviews/feedback as an accepted or nonevil organization/business would.

While I particularly advocate and look forward to an open source implementation, it isn't necessary to aggress on those that conform to such expectancies.  To do so establishes bad reputation for the consumer.  Consumers can be evil too. ^_^

In regards to your comment, an MIT license allows the code to be open source and for anyone and everyone to use and implement bitcoin into anything and everything, whether it is released as open source or established into a viable for-profit product.  To aid in the acceptance of Bitcoin for already-established businesses, MIT is a convenient license that will allow a business to consider Bitcoin.  If the license were GPL, then it would hinder growth/acceptance of Bitcoin throughout well-established businesses.

The particular friendly community member that is offering a product that generates profits is not a well-established business and perhaps the product isn't a kind of product you would expect to pay for, but it is a product nonetheless, and therefore, you must consider evaluating it as such and handle responding, considering and associating with the individual as you would any other business.  Blah blah blah, I'm sure you get my point.

[Macho]

Satoshi,

So you support people taking your code, modifying it to skim bitcoins off the miner, and then releasing the binary without releasing the modified source code?

That's exactly the point! I get the impression that many people commenting do not understand the issues involved quiet well enough and get confused a little ... no offense

So I would like to hear a response to that question too ... are you going to support and condone someone taking the code, adding a little eye-candy and little nice features perhaps and distributing this proprietary closed-sourced client to the bitcoin community? As you seem to be actually encouraging that by releasing the code under MIT license.

How are you going to check if there isn't any "bonus" to that eye-candy in the form of a backdoor, if it does follow the protocol correctly, if it doesn't at some point in time just transfer all the bitcoins from everyone to some hardcoded address? Is there any reliable way or are we just supposed to *trust* the publisher?

Why all the trouble to develop a decentralized system with specific goal of eliminating the need for central authority to trust when you then allow this exact thing to be reintroduced in the form of trusting the publisher of your software?

For those who would trade their principles for wider adoption ... I would rather want smaller community of a system I can rely on than a big one that is compromised. As far as I'm concerned, the corporations and "businesses" that won't respect the principles of transparency and openness can stick it! They're going to corrupt it and destroy anyway if they are allowed any larger influence. Did it not occur to you that there is a reason some corporations won't touch GPL? They can not cheat and rip people off with that kind of software ... that translates into lower profits. I don't know about you but I certainly wouldn't miss that kind of "company" ... That's like asking Microsoft to help popularize your OS - they are going to screw you over the first chance they get, it's going to get popular allright but it's going to be no longer the OS you had in mind at first.

Your argument Satochi that it creates duplication of work is a valid one ... it makes cheating and compromising the bitcoin community a hell of a lot easier for anyone who wishes to do so. But I did not realize that was the kind of duplicity we were worried about.

I would like to hear any reasonable scenario where distribution of a proprietary bitcoin software would be acceptable, would *you* accept a proprietary implementation to run on your system? Would you recommend to anybody to use such an implementation? I'm really curious about this ... thank you.

(and sorry, I'm kind of perturbed by the responses ... if you sense any tension from my post it's from frustration that people would even consider running a black box software dealing with bitcoins, incredible!)

[jgarzik]

So you support people taking your code, modifying it to skim bitcoins off the miner, and then releasing the binary without releasing the modified source code?

That's one obvious consequence of MIT licensing, and has been going on for decades.  I doubt it is a surprise to satoshi, or anyone else.

Either MIT or GPL, both licenses are fine.  MIT has been working great for *BSD and X11; there's no reason why MIT would be problematic for bitcoin.  GPLv3 adds some helpful patent language, that's about it.  Bitcoin's patent problems are in the area of linked libraries (openssl's EC-DSA), not with bitcoin itself, so that does not seem like a large concern here.

Speaking only for myself, as a programmer who has created or worked on dozens of GPL'd projects, including some of the largest in the world (kernel, gcc).

[BioMike]

Satoshi,

So you support people taking your code, modifying it to skim bitcoins off the miner, and then releasing the binary without releasing the modified source code?

That's exactly the point! I get the impression that many people commenting do not understand the issues involved quiet well enough and get confused a little ... no offense

So I would like to hear a response to that question too ... are you going to support and condone someone taking the code, adding a little eye-candy and little nice features perhaps and distributing this proprietary closed-sourced client to the bitcoin community? As you seem to be actually encouraging that by releasing the code under MIT license.

How are you going to check if there isn't any "bonus" to that eye-candy in the form of a backdoor, if it does follow the protocol correctly, if it doesn't at some point in time just transfer all the bitcoins from everyone to some hardcoded address? Is there any reliable way or are we just supposed to *trust* the publisher?

Then those users get what they deserve for using a closed source version, while a reliable open source/MIT version would still be available. If the closed source version is eating their bitcoins they will abandon it soon. There is even the possibility that some person develops a closed source version from scratch and does the same, no MIT/GPL/other license of the current bitcoin client is going to change anything on that.

[Macho]

That's one obvious consequence of MIT licensing, and has been going on for decades.  I doubt it is a surprise to satoshi, or anyone else.

yeah, it's not surprising at all ... that's why MIT should not be considered in a first place, exactly because it allows this kind of abuse. The question is not whether it is surprising but whether it is desirable, wouldn't you say? If any kind of proprietary software stemming from this would get popular, it's not going to be surprising at all when a hell of a lot of people will get robbed of all their bitcoins ... that doesn't mean we should allow it, does it? I said people are just confused ... talking about whether it is "surprising" the the license allows for obvious abuse rather than talking about eliminating that abuse. Do you support that kind of "obvious consequence" then? It seem to me that you're arguing in its favor.

Either MIT or GPL, both licenses are fine.  MIT has been working great for *BSD and X11; there's no reason why MIT would be problematic for bitcoin.  GPLv3 adds some helpful patent language, that's about it.  Bitcoin's patent problems are in the area of linked libraries (openssl's EC-DSA), not with bitcoin itself, so that does not seem like a large concern here.

X11 or BSD are not small client programs that handle your freaking financial transactions, I bet you would be all happy to accept this program I made, it makes credit card transactions a lot easier, you just type in your credit card information into it ... it's real convenient, shall I send it to you? I'm sure you wouldn't have any problem with that, would ya?

Speaking only for myself, as a programmer who has created or worked on dozens of GPL'd projects, including some of the largest in the world (kernel, gcc).

It's hard to believe that such an intelligent person would make such comments you just did then ... I'm startled.

[jgarzik]

That's one obvious consequence of MIT licensing, and has been going on for decades.  I doubt it is a surprise to satoshi, or anyone else.

yeah, it's not surprising at all ... that's why MIT should not be considered in a first place, exactly because it allows this kind of abuse. The question is not whether it is surprising but whether it is desirable, wouldn't you say? If any kind of proprietary software stemming from this would get popular, it's not going to be surprising at all when a hell of a lot of people will get robbed of all their bitcoins ... that doesn't mean we should allow it, does it? I said people are just confused ... talking about whether it is "surprising" the the license allows for obvious abuse rather than talking about eliminating that abuse. Do you support that kind of "obvious consequence" then? It seem to me that you're arguing in its favor.

Binaries may contain suspect code regardless of the license.  Yet 999 out of 1000 users prefer binaries, because they are not programmers and would have no clue what to do with source code in their hands.

The source (ie. download origin, and PGP signer) of the software is always far more paramount, if you are worrying about bitcoin theft and other abuse.

X11 or BSD are not small client programs that handle your freaking financial transactions, I bet you would be all happy to accept this program I made, it makes credit card transactions a lot easier, you just type in your credit card information into it ... it's real convenient, shall I send it to you? I'm sure you wouldn't have any problem with that, would ya?

BSD OS's handle tons of financial transactions.  Wall Street loves Linux, but it's got plenty of BSD in there too.

[Macho]

Then those users get what they deserve for using a closed source version, while a reliable open source/MIT version would still be available.

Oh gosh, I have to calm down ... it really is hard for some people to grasp this. Please listen to yourself, you're saying that using closed source client is a bad idea, do I understand that right? And you are encouraging people to use the open-source MIT one over the closed source one, right? That means people should really use only the open-source version to not get screwed ... what. is. the. point. of. the. MIT. license. then!? The MIT license is specifically designed to allow for closed source derivatives! If you're saying that people should only use open-source one without the risk of getting "what they deserve" if they don't ... you are in total agreement with me and you are making the exact same argument for using GPL as I am. Why do you feel the need to make excuses for MIT license then? Do you not like me personally that you just have to disagree with me for some reason? Or is this some ego thing about "winning an argument" or something? I honestly do not understand why people do this, it's frustrating. You may not like me, you may not like my style, you may enjoy arguing or whatever but I beg you to leave that behind and look at the facts and logic instead, please!

If the closed source version is eating their bitcoins they will abandon it soon. There is even the possibility that some person develops a closed source version from scratch and does the same, no MIT/GPL/other license of the current bitcoin client is going to change anything on that.

Ok, one by one:

1) Proprietary version doesn't mean that it will be doing only things that you'll easily notice, it can have backdoors, it can be dormant for several years and then rob a half the community at some point effectively destroying bitcoin (but still making a huge profit for whoever has done this)

2) Yes, there is a possibility that someone will develop a client from scratch, is there any reason you can think of that we should make that possibility much higher by making it easy to do? I love this logic, let's all put wallets on our front porch ... they could just taken them by stealing them in a crowded bus anyway, so what's the harm.

All you've written are excuses to do nothing ... not reasons to have MIT license, you did not say why MIT would be preferable to GPL, you just made excuses for the additional and unnecessary pitfalls of MIT. There is no reason to keep these pitfalls if there are no benefits that outweigh them. I see no benefits to MIT, only unnecessary dangers. The only supposed "benefit" is the possibility of closed sourced forks which you yourself said are dangerous and discouraged. So I really do not see why would you have any rational reason to disagree with me.

[Macho]

Binaries may contain suspect code regardless of the license.  Yet 999 out of 1000 users prefer binaries, because they are not programmers and would have no clue what to do with source code in their hands.

I can't believe this, what are you arguing here? The same question, is this something about "winning an argument"? NONE of what have you written is ANY reason WHATSOEVER to use MIT over GPL. Why are you making up excuses (ridiculous at that) for the pitfalls of MIT rather than stating why would MIT be more advantageous to GPL? There is an easy answer for that, because there isn't any rational advantage to it! Is it surprising that I'm going mad here over that kind of silly things you say?

There is really simple thought process behind this to decide this question, please answer these one by one:

1) Would you personally accept closed source software dealing with your bitcoin transactions running on your station? (in that case I'm going to send you that credit card utility I told you about)
2) Would you recommend to anybody else to use closed source implementation of the bitcoin protocol? (not yours with your backdoor to make a buck I mean third-party software)
3) If you would not use closed source yourself and did not recommend to others to use such a software, what good is an option to fork existing implementation and make one?

See? Real easy .. I do not see any reason why would you want to fight a battle against GPL, why would you be against GPL? It makes no sense.

The only other option would be to *trust* the publisher of the software, in that case I would ask why use bitcoin at all? Its very design is centered about the idea of eliminating central authority that has to be trusted. You can just as well use regular dollars and *trust* the federal reserve and the US government to take care of your currency ... that turned out well, right? Do you want the same thing to happen to bitcoin?

[mizerydearia]

Binaries may contain suspect code regardless of the license.  Yet 999 out of 1000 users prefer binaries, because they are not programmers and would have no clue what to do with source code in their hands.

The source (ie. download origin, and PGP signer) of the software is always far more paramount, if you are worrying about bitcoin theft and other abuse.

Binaries compiled from open source code can potentially be confirmed by the community that the binaries are indeed related to a particular svn snapshot of source code.  I am not sure how, but I am sure this is possible to confirm.

Binaries compiled from closed source code can not be confirmed by the community that the binaries are indeed related to a particular svn snapshot of source code without modification, because the code has obviously been modified and there is no way to know what the modified code looks like.

[mizerydearia]

Linux has various including GNU General Public License, BSD License, Apache License, MIT License, and other licenses
Apache has an Apache license
MySQL has GNU General Public License (version 2, with linking exception) or proprietary EULA
PHP has a PHP license

Asterisk has GNU General Public License
Avidemux has GNU General Public License
Blender has GNU General Public License v2 or later
Cinelerra has GNU General Public License
ffmpeg has GNU LGPL 2.1+, GNU GPL 2+
Gimp has GNU General Public License
GNU Compiler Collection has GNU General Public License (version 3 or later)
LiVES has GNU General Public License version 3 or higher.
MPlayer has GNU General Public License
OpenOffice has GNU Lesser General Public License v3
OpenVPN has GNU GPL
OpenX has GNU General Public License
VirtualDub has GNU General Public License
VLC has GNU General Public License v2 or later
xine has GNU GPL

There are very few open source applications that I've noticed that have MIT license.  Perhaps in deciding upon MIT license or another license a kind of role model should be considered and to perhaps discuss with other communities about their perspective on licensing.

http://www.cio.com/article/32146/Open_Source_The_Myths_of_Open_Source

Alberg is careful to make clear that his commitment to open source isn’t the blind buying behavior of a zealot. He wouldn’t, for example, go open source if it were more expensive than proprietary code.  "Solaris is a strong commercial operating system. We’d choose it over open source if we found it to be less expensive," he says. "[While] cost is a huge driver for our decision-making process, we cannot risk choosing an inferior solution to save money. We couldn’t even consider open source if it weren’t at par with?or in some cases better than?commercial alternatives."

As Hein points out, most open-source projects have a large corps of developers, Internet mailing lists, archives and support databases?all available at no cost.

Proprietary versions of Bitcoin client will profit AND escalate all support to be handled by the open source, willing-to-help community at no cost to proprietary designer.

[jgarzik]

Binaries compiled from open source code can potentially be confirmed by the community that the binaries are indeed related to a particular svn snapshot of source code.  I am not sure how, but I am sure this is possible to confirm.

Theoretically yes.  Practically?  No.

Given all the variables -- compiler, compiler platform, compiler version, compiler options, linked library versions of all our dependent libraries -- this is quite difficult.

[jgarzik]

1) Would you personally accept closed source software dealing with your bitcoin transactions running on your station? (in that case I'm going to send you that credit card utility I told you about)

It depends on the source.

2) Would you recommend to anybody else to use closed source implementation of the bitcoin protocol? (not yours with your backdoor to make a buck I mean third-party software)

It depends on the source.

3) If you would not use closed source yourself and did not recommend to others to use such a software, what good is an option to fork existing implementation and make one?

Question invalid due to previous answers.

See? Real easy .. I do not see any reason why would you want to fight a battle against GPL, why would you be against GPL? It makes no sense.

Anyone who says "why would you be against GPL? it makes no sense" is a zealot.  Each licensing option has its own time and place.

[ribuck]

A closed source client is a bad idea.

But this is why users should prefer an open source client. It's not the software developer's job to decide this for the end user.

[LZ]

I believe that users themselves may decide which client is better to use.
If a proprietary client steals money, there is no fault of the community...

[Macho]

It depends on the source.

You mean like whether you can *trust* it? THAT'S MY FREAKING POINT and it's spelled out in my previous post! Based on what are you going to trust the source? Whether it is large enough? Microsoft sure can be trusted ... or how nice of a logo they have? Maybe your friend is CEO? Is he expected to be personal friend with everyone who uses the software? How does it depend on the source? Please... share this ultimate wisdom with me.

Anyone who says "why would you be against GPL? it makes no sense" is a zealot.  Each licensing option has its own time and place.

No, anyone who posts several messages without giving ANY reason why MIT would be preferred to GPL is an irrational person. You can not give a reason because there is no reason. The only thing left except admitting that, is calling people names, there is no rational reason against using GPL ... if there was, you would stated it already. But you must be an Internet hero and argue your case to the death no matter how wrong you are. That kind of mentality just makes me mad, as you may have noticed and you call me a zealot, lol

[BioMike]

Then those users get what they deserve for using a closed source version, while a reliable open source/MIT version would still be available.

Oh gosh, I have to calm down ... it really is hard for some people to grasp this. Please listen to yourself, you're saying that using closed source client is a bad idea, do I understand that right? And you are encouraging people to use the open-source MIT one over the closed source one, right? That means people should really use only the open-source version to not get screwed ... what. is. the. point. of. the. MIT. license. then!? The MIT license is specifically designed to allow for closed source derivatives! If you're saying that people should only use open-source one without the risk of getting "what they deserve" if they don't ... you are in total agreement with me and you are making the exact same argument for using GPL as I am. Why do you feel the need to make excuses for MIT license then? Do you not like me personally that you just have to disagree with me for some reason? Or is this some ego thing about "winning an argument" or something? I honestly do not understand why people do this, it's frustrating. You may not like me, you may not like my style, you may enjoy arguing or whatever but I beg you to leave that behind and look at the facts and logic instead, please!

Whooh there!!... I'm not trying to be against you. Just trying to have a normal discussion, nothing personal. It is normal that people might have different opinions about something. As for my reasons in this discussion, see below.

If the closed source version is eating their bitcoins they will abandon it soon. There is even the possibility that some person develops a closed source version from scratch and does the same, no MIT/GPL/other license of the current bitcoin client is going to change anything on that.

Ok, one by one:

1) Proprietary version doesn't mean that it will be doing only things that you'll easily notice, it can have backdoors, it can be dormant for several years and then rob a half the community at some point effectively destroying bitcoin (but still making a huge profit for whoever has done this)

2) Yes, there is a possibility that someone will develop a client from scratch, is there any reason you can think of that we should make that possibility much higher by making it easy to do? I love this logic, let's all put wallets on our front porch ... they could just taken them by stealing them in a crowded bus anyway, so what's the harm.

All you've written are excuses to do nothing ... not reasons to have MIT license, you did not say why MIT would be preferable to GPL, you just made excuses for the additional and unnecessary pitfalls of MIT. There is no reason to keep these pitfalls if there are no benefits that outweigh them. I see no benefits to MIT, only unnecessary dangers. The only supposed "benefit" is the possibility of closed sourced forks which you yourself said are dangerous and discouraged. So I really do not see why would you have any rational reason to disagree with me.

Ok, to be clear about my point. I've been using FOSS for over 15 years now and I don't care if it is GPL or MIT. They are both open source and free. That is the pro against closed source licenses for me. If a developer chooses one or the other, it is his choice, I can't force him to use an other license if I haven't written any code myself. If I don't agree with this, I write my own code and place it under my license of choice, or I shut up. I would love to see the/a client under a GPL license and I think I would prefer to use that one over the MIT one, but there is currently no GPL one, so I have to do with the MIT one and that is fine with me (as long as I have the basic rights that define it as FOSS).

Now, back to the GPL vs MIT discussion. Both licenses were once made to give the user the freedom to use the software how he/she wants to use it. But the licenses have a different view about the freedom. GPL restricts the user in taking these freedoms away (copyleft), making it in BSD eyes less free. BSD/MIT on the other hand give the freedom to take freedoms away from the user. Two different point of views causing a dilemma about which one is more free. But, IMHO both are valid to use. Personally I don't mind the copyleft restriction of the GPL, but you also have to respect other peoples opinion that this "hypocritical" restriction should not be needed in the first place.

I have enough thrust in FOSS that the open source MIT client will keep its leading edge over future closed source ones as long as it keeps actively developed and keeps support of its community.

[lfm]

Now, back to the GPL vs MIT discussion. Both licenses were once made to give the user the freedom to use the software how he/she wants to use it. But the licenses have a different view about the freedom. GPL restricts the user in taking these freedoms away (copyleft), making it in BSD eyes less free. BSD/MIT on the other hand give the freedom to take freedoms away from the user. Two different point of views causing a dilemma about which one is more free. But, IMHO both are valid to use. Personally I don't mind the copyleft restriction of the GPL, but you also have to respect other peoples opinion that this "hypocritical" restriction should not be needed in the first place.

Note who gets and loses freedom in each case. The GPL restricts the freedom of the developer and maintains the freedom of the user to modify the system. BSD/MIT gives the developer the freedom to restrict the freedom of the user to do such modifications. Which license you like can depend on who you are.

Of course most of us are really both here. It is really interesting to note how the gcc (GNU compiler) remains under the GPL, where the programmers are the users, and is even used on many BSD operating systems. (Yes I know there are efforts to change that.)

[aceat64]

Macho, the GPL license doesn't stop anyone from making a closed source client. It just requires them to write it from scratch (or be dishonest and use Bitcoin code).