5 ms·

I Love the GPL (Except When it Applies to Me)

29 points by blasdel 17y ago

blur 17y ago
Comments like the ones on this article are why younger devs are avoiding the free software movement (notice I said movement and not free software itself). Instead of honestly addressing the shortcomings of the GPL you get all defensive and huffy, like a 10 year old who's been told his skateboard sucks.
dschobel 17y ago
I can't tell whether the angry/swearing diatribe tone is a dramatic effect or just lack of creativity/intelligence on behalf of the author.
- macmac 17y ago
  And his assertion about license proliferation is certainly not correct ref http://www.blackducksoftware.com/oss/licenses#top20 http://www.blackducksoftware.com/oss/licenses#top20
  - nothingHappens 17y ago
    Except, is that chart of how many projects use each license, or how much much use software with those licenses sees? Those could be very different. But ups for commenting on the substance rather than the style.
    - callahad 17y ago
      Both Blackduck and Gartner Group are pushing a doomsday scenario of GPLv3 and the AGPL seeing surging use, respectively. The patent / IP language in the GPLv3 still scares a lot of larger companies, my employer included. It's in Blackduck's commercial interest to present information in a way that exaggerates the prominence of that family of licenses. Thus, the chart is of the number of projects released under that license, and does not take into account the maturity or market penetration of a project. For instance, a new entry on Sourceforge with nothing but a README in its repo would be included in Blackduck's count. That's not to say Blackduck itself is all bad, mind you. I've sat through a few Continuing Legal Education webinars that they've hosted, and they've all been quite objective and accurate. In each case, they brought in fairly prominent figures from the Open Source law space (Karen Copenhaver, Mark Radcliffe, etc) for the sessions.
  - mark_l_watson 17y ago
    That is interesting, thanks. I did not realize how few projects use the AGPL. One problem with the AGPL is that it seems to work for projects that are totally owned by one individual or company because there is a clear stratgey for supporting alternative commercial use licenses with no software distribution requirements. If alternative use licenses are low cost for received value and if the commercial license passes corporate lawyer scrutiny, all is good. The big problem with the AGPL is projects with multiple contributers - makes a paid for commercial license difficult.
jacoblyles 17y ago
I don't get his point. Does he think copyleft is a good ideal failed by the specific personalities of the current maintainers, or a flawed dream unable to survive in the reality of the economic marketplace? The essay feels unfocused and disorganized. Now, I'm an open-minded child of the modern age and I don't agree with the oppressive old-fashioned idea that every essay must have a point. And truthfully this piece does bring up some interesting food for thought (the SaaS loophole). But it is nothing we haven't heard before, with less juvenile tone.
- tentonova 17y ago
  My reading was that he was presenting the GPL as an inherently flawed dream unable to survive in the reality of the economic marketplace. In succeeding in market adoption, the proliferation of the GPL actually leads to its own failure due to its knocking the legs out of most of the higher-margin business models required to fund the R&D necessary to produce software. That might also be my own bias showing.
jmillikin 17y ago
Anybody who believes the GPL was invented to coerce users into sending patches ought to have their head examined.
- j_baker 17y ago
  I might not agree with the GPL, but I'm pretty sure that the intent behind it is to encourage people to contribute back to the open source community (even if not by directly sending patches).
  - Herring 17y ago
    No. Look, it's right there in the preamble. First two sentences...
  - blasdel 17y ago
    The intent is to require source contribution downstream. Any encouraging of upstream communication comes from the difficulty of keeping up with a fork and the knowledge that most code is a liability not an asset. The GPL only pushes you that way indirectly.
    - dagw 17y ago
      Not just source contribution, but also the contribution of new GPL applications. I've written a couple of apps which are licensed under the GPL, not because I necessarily thought it was the best license for me, but simply because I wanted to use a GPL library. Had the library I wanted to use been some other license, my app might have been some other license. But it wasn't and now there is one more GPL app. Of course there is the other side of the coin where I haven't used a library I really wanted to use because I couldn't GPL my app.
- tentonova 17y ago
  The GPL was invented to leverage network effects to win in the market place, ensuring an exceptionally high entry barrier to non-GPL market entrants (a natural monopoly) should the GPL be successful in this goal. It hasn't won in the market place -- and it won't -- primarily due to the fact that the coupling between funding and licensing is so indirect -- it's very difficult to earn a high margin on writing GPL licensed code, and most revenue models require an indirect and complex approach to earning money from the software without charging directly for the software itself. (Yes, you can charge for GPL software. No, that's not a sustainable business model. No, RedHat is not a counter-example. RedHat charges for support and their trademarks, not the software itself, and in doing so appeals to a very specific niche enterprise market.)
blasdel 17y ago
I'm constantly struck by just how much of a turd the AGPL is -- it's a baldfaced EULA! It takes a huge leap beyond what is enforceable via straightforward copyright law and wallows right into the deep end of clickwrap licenses (the GPLv3 bullshit at least still hinges upon distribution). With all other FOSS licenses, it's lazy bullshit to use them in the LICENSE field of your installer maker -- I absolutely do not have to accept the terms of the GPL to use your software, it's right there in the text of the license, so don't you dare make me click an 'I accept' checkbox. With the AGPL it's absolutely necessary to do this, because I don't actually have The freedom to run the program, for any purpose The RMS of twenty years ago would have never approved of such a thing, that he has lent it the imprimatur of the FSF now is testament to his intellectual bankruptcy. The AGPL will never ever succeed. It will only ever be used by egotistical idiots and dying companies that wish to sear a scarlet letter into their code -- noone but I may profit directly from this work -- as you still retain copyright by default. Is the AGPL transitive across the network the way the GPL is by linking? It seems to me like either answer would show how poorly considered it is as a license.
- jmillikin 17y ago
  The RMS of twenty years ago would have never approved of such a thing, that he has lent it the imprimatur of the FSF now is testament to his intellectual bankruptcy. I doubt he approves now. I've not been able to find a single essay or even email from him regarding the AGPL, which is interesting considering his standard evangelism regarding other licenses. And he regards SaaS as non-free, regardless of whether the source is available. Is the AGPL transitive across the network the way the GPL is by linking? There's no legal precedent (that I know of) for the GPL being transitive across dynamic linking, and if it is, that opens a number of legal problems when linking with Free implementations of proprietary APIs. Static linking is, of course, still covered under standard copyright. From my non-lawyer reading of the AGPL, it does transit between processes and systems because it covers the output of the application.
  - blasdel 17y ago
    his public silence is odd -- but given that the GPLv3 contains a clause explicitly granting an exemption for use with the AGPL, and the degree to which the FSF is essentially an extension of his person, he must approve in some sense.
mark_l_watson 17y ago
I don't agree with the trash talk on the AGPL in these comments to the original article. I would like to see fewer open source licenses: in my opinion the Apache 2, LGPL 3, and AGPL 3 would suffice since I am not that clear why Apache 2 and BSD styles licensed projects just couldn't be Apache 2. That said: creators should obviously have the only say in what license they choose to use. I know someone who has a small AI + semantic web company, they use the AGPL, and I think that they are doing OK. The trick is selling a "commercial use" exception. Also, my understanding is that if you use AGPL code in a web service, then you need to make the entire web service implementation AGPL. However, other systems that use the newly created AGPL web service do not need to be AGPL - if I am wrong about this, I would appreciate being corrected.
- tedunangst 17y ago
  "I am not that clear why Apache 2 and BSD styles licensed projects just couldn't be Apache 2." $ wc -w isc apache 192 isc 1395 apache
freetard 17y ago
> Free Software: I found a loophole in my student loan documentation that lets me defer payments for decades, so long as I stay in the Ph.D. program! I'm sure Stallman and the like who have given their whole life to free software will appreciate. Meanwhile, this guy hasn't done much and yet not only complains about free software while using them but also calls them lazy at the same time. Shameless.
tptacek 17y ago
I have no idea what his point is, but from a software business perspective, the difference between BSD-style and GPL-style is simple: GPL-style licenses make it harder for competing companies to use your code. GPL is therefore an easier sell inside of a company, because it gives the software owner special rights over the code.
- tentonova 17y ago
  In contrast, as a company looking to use and contribute to open-source software, BSD licensed code is an easier sell because you aren't locked to purchasing a commercial non-OSS license from the copyright holders (see: Oracle/Sun/Monty regarding MySQL).
  - tptacek 17y ago
    Yes, the BSD license sure is better for software users. You can lift a BSD package wholesale into a new white-label product, and you don't even have to tell the author! That sure is convenient. You use the BSD license when you don't care about the commercial value of your code. But, for the most part, I think using the BSD license means you're not going to care about the commercial value of your code, because if it's valuable, people will freeload. In my industry, the GPL has done a servicable job protecting some key software projects (Snort, Nessus, Wireshark). The counterexamples (notably Metasploit) are utterly dependent on the continued effort of the original authors --- Metasploit goes stale faster than almost any other security tool.
    - tentonova 17y ago
      Yes, the BSD license sure is better for software users. It's also better for software contributors who want to leverage the code in their own products, and use the resulting revenue to contribute back their changes (but not their entire product). But, for the most part, I think using the BSD license means you're not going to care about the commercial value of your code, because if it's valuable, people will freeload. Some people will free-load. Many won't. Look at the Apache project, where a vast majority of development is funded by corporations using the software in their closed source products while contributing improvements back to the original product. In my industry, the GPL has done a servicable job protecting some key software projects (Snort, Nessus, Wireshark). How do you define "protecting", and what evidence would you use to demonstrate that protection? What would you say those projects were protected from?

29 points by blasdel 17y ago

blur 17y ago

Comments like the ones on this article are why younger devs are avoiding the free software movement (notice I said movement and not free software itself). Instead of honestly addressing the shortcomings of the GPL you get all defensive and huffy, like a 10 year old who's been told his skateboard sucks.

dschobel 17y ago

I can't tell whether the angry/swearing diatribe tone is a dramatic effect or just lack of creativity/intelligence on behalf of the author.

macmac 17y ago
And his assertion about license proliferation is certainly not correct ref http://www.blackducksoftware.com/oss/licenses#top20 http://www.blackducksoftware.com/oss/licenses#top20
- nothingHappens 17y ago
  Except, is that chart of how many projects use each license, or how much much use software with those licenses sees? Those could be very different. But ups for commenting on the substance rather than the style.
  - callahad 17y ago
    Both Blackduck and Gartner Group are pushing a doomsday scenario of GPLv3 and the AGPL seeing surging use, respectively. The patent / IP language in the GPLv3 still scares a lot of larger companies, my employer included. It's in Blackduck's commercial interest to present information in a way that exaggerates the prominence of that family of licenses. Thus, the chart is of the number of projects released under that license, and does not take into account the maturity or market penetration of a project. For instance, a new entry on Sourceforge with nothing but a README in its repo would be included in Blackduck's count. That's not to say Blackduck itself is all bad, mind you. I've sat through a few Continuing Legal Education webinars that they've hosted, and they've all been quite objective and accurate. In each case, they brought in fairly prominent figures from the Open Source law space (Karen Copenhaver, Mark Radcliffe, etc) for the sessions.
- mark_l_watson 17y ago
  That is interesting, thanks. I did not realize how few projects use the AGPL. One problem with the AGPL is that it seems to work for projects that are totally owned by one individual or company because there is a clear stratgey for supporting alternative commercial use licenses with no software distribution requirements. If alternative use licenses are low cost for received value and if the commercial license passes corporate lawyer scrutiny, all is good. The big problem with the AGPL is projects with multiple contributers - makes a paid for commercial license difficult.

jacoblyles 17y ago

I don't get his point. Does he think copyleft is a good ideal failed by the specific personalities of the current maintainers, or a flawed dream unable to survive in the reality of the economic marketplace? The essay feels unfocused and disorganized. Now, I'm an open-minded child of the modern age and I don't agree with the oppressive old-fashioned idea that every essay must have a point. And truthfully this piece does bring up some interesting food for thought (the SaaS loophole). But it is nothing we haven't heard before, with less juvenile tone.

tentonova 17y ago
My reading was that he was presenting the GPL as an inherently flawed dream unable to survive in the reality of the economic marketplace. In succeeding in market adoption, the proliferation of the GPL actually leads to its own failure due to its knocking the legs out of most of the higher-margin business models required to fund the R&D necessary to produce software. That might also be my own bias showing.

jmillikin 17y ago

Anybody who believes the GPL was invented to coerce users into sending patches ought to have their head examined.

j_baker 17y ago
I might not agree with the GPL, but I'm pretty sure that the intent behind it is to encourage people to contribute back to the open source community (even if not by directly sending patches).
- Herring 17y ago
  No. Look, it's right there in the preamble. First two sentences...
- blasdel 17y ago
  The intent is to require source contribution downstream. Any encouraging of upstream communication comes from the difficulty of keeping up with a fork and the knowledge that most code is a liability not an asset. The GPL only pushes you that way indirectly.
  - dagw 17y ago
    Not just source contribution, but also the contribution of new GPL applications. I've written a couple of apps which are licensed under the GPL, not because I necessarily thought it was the best license for me, but simply because I wanted to use a GPL library. Had the library I wanted to use been some other license, my app might have been some other license. But it wasn't and now there is one more GPL app. Of course there is the other side of the coin where I haven't used a library I really wanted to use because I couldn't GPL my app.
tentonova 17y ago
The GPL was invented to leverage network effects to win in the market place, ensuring an exceptionally high entry barrier to non-GPL market entrants (a natural monopoly) should the GPL be successful in this goal. It hasn't won in the market place -- and it won't -- primarily due to the fact that the coupling between funding and licensing is so indirect -- it's very difficult to earn a high margin on writing GPL licensed code, and most revenue models require an indirect and complex approach to earning money from the software without charging directly for the software itself. (Yes, you can charge for GPL software. No, that's not a sustainable business model. No, RedHat is not a counter-example. RedHat charges for support and their trademarks, not the software itself, and in doing so appeals to a very specific niche enterprise market.)

blasdel 17y ago

I'm constantly struck by just how much of a turd the AGPL is -- it's a baldfaced EULA! It takes a huge leap beyond what is enforceable via straightforward copyright law and wallows right into the deep end of clickwrap licenses (the GPLv3 bullshit at least still hinges upon distribution). With all other FOSS licenses, it's lazy bullshit to use them in the LICENSE field of your installer maker -- I absolutely do not have to accept the terms of the GPL to use your software, it's right there in the text of the license, so don't you dare make me click an 'I accept' checkbox. With the AGPL it's absolutely necessary to do this, because I don't actually have The freedom to run the program, for any purpose The RMS of twenty years ago would have never approved of such a thing, that he has lent it the imprimatur of the FSF now is testament to his intellectual bankruptcy. The AGPL will never ever succeed. It will only ever be used by egotistical idiots and dying companies that wish to sear a scarlet letter into their code -- noone but I may profit directly from this work -- as you still retain copyright by default. Is the AGPL transitive across the network the way the GPL is by linking? It seems to me like either answer would show how poorly considered it is as a license.

jmillikin 17y ago
The RMS of twenty years ago would have never approved of such a thing, that he has lent it the imprimatur of the FSF now is testament to his intellectual bankruptcy. I doubt he approves now. I've not been able to find a single essay or even email from him regarding the AGPL, which is interesting considering his standard evangelism regarding other licenses. And he regards SaaS as non-free, regardless of whether the source is available. Is the AGPL transitive across the network the way the GPL is by linking? There's no legal precedent (that I know of) for the GPL being transitive across dynamic linking, and if it is, that opens a number of legal problems when linking with Free implementations of proprietary APIs. Static linking is, of course, still covered under standard copyright. From my non-lawyer reading of the AGPL, it does transit between processes and systems because it covers the output of the application.
- blasdel 17y ago
  his public silence is odd -- but given that the GPLv3 contains a clause explicitly granting an exemption for use with the AGPL, and the degree to which the FSF is essentially an extension of his person, he must approve in some sense.

mark_l_watson 17y ago

I don't agree with the trash talk on the AGPL in these comments to the original article. I would like to see fewer open source licenses: in my opinion the Apache 2, LGPL 3, and AGPL 3 would suffice since I am not that clear why Apache 2 and BSD styles licensed projects just couldn't be Apache 2. That said: creators should obviously have the only say in what license they choose to use. I know someone who has a small AI + semantic web company, they use the AGPL, and I think that they are doing OK. The trick is selling a "commercial use" exception. Also, my understanding is that if you use AGPL code in a web service, then you need to make the entire web service implementation AGPL. However, other systems that use the newly created AGPL web service do not need to be AGPL - if I am wrong about this, I would appreciate being corrected.

tedunangst 17y ago
"I am not that clear why Apache 2 and BSD styles licensed projects just couldn't be Apache 2." $ wc -w isc apache 192 isc 1395 apache

freetard 17y ago

> Free Software: I found a loophole in my student loan documentation that lets me defer payments for decades, so long as I stay in the Ph.D. program! I'm sure Stallman and the like who have given their whole life to free software will appreciate. Meanwhile, this guy hasn't done much and yet not only complains about free software while using them but also calls them lazy at the same time. Shameless.

tptacek 17y ago

I have no idea what his point is, but from a software business perspective, the difference between BSD-style and GPL-style is simple: GPL-style licenses make it harder for competing companies to use your code. GPL is therefore an easier sell inside of a company, because it gives the software owner special rights over the code.

tentonova 17y ago
In contrast, as a company looking to use and contribute to open-source software, BSD licensed code is an easier sell because you aren't locked to purchasing a commercial non-OSS license from the copyright holders (see: Oracle/Sun/Monty regarding MySQL).
- tptacek 17y ago
  Yes, the BSD license sure is better for software users. You can lift a BSD package wholesale into a new white-label product, and you don't even have to tell the author! That sure is convenient. You use the BSD license when you don't care about the commercial value of your code. But, for the most part, I think using the BSD license means you're not going to care about the commercial value of your code, because if it's valuable, people will freeload. In my industry, the GPL has done a servicable job protecting some key software projects (Snort, Nessus, Wireshark). The counterexamples (notably Metasploit) are utterly dependent on the continued effort of the original authors --- Metasploit goes stale faster than almost any other security tool.
  - tentonova 17y ago
    Yes, the BSD license sure is better for software users. It's also better for software contributors who want to leverage the code in their own products, and use the resulting revenue to contribute back their changes (but not their entire product). But, for the most part, I think using the BSD license means you're not going to care about the commercial value of your code, because if it's valuable, people will freeload. Some people will free-load. Many won't. Look at the Apache project, where a vast majority of development is funded by corporations using the software in their closed source products while contributing improvements back to the original product. In my industry, the GPL has done a servicable job protecting some key software projects (Snort, Nessus, Wireshark). How do you define "protecting", and what evidence would you use to demonstrate that protection? What would you say those projects were protected from?