Re: Is inherited class a derivative work?
Michael Beck wrote:
Some people believe that when you subclass a new class,
you are creating a derivative work in the copyright sense,
especially when you override existing methods.
The scary scenario is that somebody will inherit a
class, make some modifications to it, and then claim
that since it's not a derivative work,
s/he doesn't have to publish it,
and the new class is his own proprietary code.
IANAL, TINLA, IMHO, YADA YADA YADA:
interesting fear you've got there.
first of all, you're mixing USE (inherit)
with MODIFY. A derived class is not a modification
of the original.
say Alice writes a class, AlicesRestaurant, which
does whatever it does, and it contains
a method called pretty_print
sub pretty_print
{
print $line_number $text \n;
}
Bob comes along, inherits from the
class AlicesRestaurant, and does this
solely for the purpose of overriding the
pretty_print method.
sub pretty_print
{
# 5,000 lines of code
}
And your fear is that Bob will claim
that his pretty_print code is proprietary.
that's an interesting fear you've got there.
Just what, exactly, did Alice do that allows
her to claim Bob's 5000 lines of code as
derived work? I'll emphasize the word WORK,
in this case, since Alice didn't do a damn
thing to create those 5000 lines of code.
well, she wrote the original method,
and Bob overrode the method by writing
a method of the same name.
Is using the same method name
pretty_printer the same as using
the name Luke Skywalker in a novel?
if I write a sci-fi epic about
Luke Skywalker, Han-Solo, and Princess Leia,
without first getting Lucas's approval,
I'd be in hot water.
this is basically how you're looking at class
inheritance. And I don't think it applies.
Good God, I hope not.
If you accept this premise, then it
is impossible to use anyone's code
without it being a derived work.
USE and DERIVED WORKS become collapsed.
Any USE becomes a DERIVED WORK.
Alice's pretty_print method is simply
an interesting plot twist in an entire novel.
The butler did it.
you can still have an endless supply
of mystery novels where the butler did it,
and not infringe on each other.
They just do it in different ways.
Bob's pretty_print method is the same
plot twist, but he implements it in
a completely independent way.
Alice should probably be reminded that
she saw code in textbooks at school that
contained pretty_print methods of their
own. So, by her own standards, her
code is actually a derived work of
someone else's.
By her premise, all fiction is derived
from ancient greek mythology,
and all source code is derived from
Ada Lovelace.
Alice needs to chill out a bit.
Not all code flows from her.
IANAL TINLA IMHO YADA YADA YADA
Greg There is nothing new under the sun London
--
license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: binary restrictions?
Steve Lhomme wrote: - Original Message - From: Greg London [EMAIL PROTECTED] To: Steve Lhomme [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sent: Sunday, October 07, 2001 5:30 PM Subject: Re: binary restrictions? | Steve Lhomme wrote: | | A binary is a derived work. | | Are you sure of that ? When you compile | you USE the code not MODIFY it. | There's no derivation. Otherwise using | a software and changing the default | settings would be a derived work... | | | converting source code to binary is in | effect a *translation* from one language | to another. That makes good sense. But in this case, why is their different rules for source code and binary versions of a work in most open-source licenses ? I mean if it's a derived work, the rules applied are the same one of a derived work. whoah, wait, there are not different LAWS for open source software versus novels. They both follow copyright law. The original question was whether or not compiling source code was USE or DERIVED WORK in the eyes of the LAW. Which is a separate issue from why open source LICENSES treat source code and binaries differently. The LAW says the copyright holder retains all rights to derived works. Compiling is a derived work. Therefore LICENSES do not have to grant rights to binaries. why open source licenses treat them differently is a separate issue. But the original question basically was whether or not a software author has the right to restrict binaries in the first place. I think copyright law says authors are given the right to control binary versions of their source code. Greg -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: The Invisible Hand
Free is the right word for some people, because free software is right for their situation. Open is the right word for some people, because open is right for their situation. The conflict arises when people confuse what is right for them and mistakenly think they have figured out what is right. It is a common human condition that everyone (including myself) suffers from occassionally. The best I've come up with is to try and remind the person that whatever 'truth' they've come up with is not necessarily the best for everyone. RedHat may be right for me. Suse may be you're best choice. KDE versus Gnome, Xemacs versus vi, etc. And if I can't the person to acknowledge that 'vi' is not the best editor for everyone, at least I can walk away in peace, rather than get wrapped up in a flame war. I've gone through the decision making process, and I've figured out what's best for my situation. If they want to go off on some rant of righteousness, I don't have to waste my time to listen. I just have to remind myself in the midst of the emails... ;) Greg London The Anti-Terrorism Act goes too far: http://www.theregister.co.uk/content/6/21854.html Stop the madness before it's too late: http://www.indefenseoffreedom.org eom -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: The Invisible Hand
Everyone breath for a second. My understanding of David's original post was to assert that open and free were meaningless distinctions. because of Adam Smith's notion of Invisible Hand, it didn't matter where you start, you end up at effectively the same end point. Therfore there is no need to divide the community into factions. What then flamed up was a debate that basically highlighted the very rift that David spoke of. The question is whether open and free are effectively meaningless distinctions. From a legal standpoint, I think they are meaningless. But not because of any invisible hand coming in and mucking up the works. You cannot create a license that says this work shall be open or this work shall be free and have it have any concrete meaning in court. At least not in the sense of free or open we're talking about. The only distinction that has any real world effect is the wording of the license, the rights granted to the world by the copyright holder. anything else has little real meaning. Greg -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: click, click, boom
Rick Moen wrote: begin Greg London quotation: Look, nobody's going to force-feed common sense to people who don't want to read the OSD in the spirit intended. One has to find one's own. If someone puts out a bunch of source code under the MIT license, and the distro is OSI certifiable, there is nothing to prevent someone else from redistributing it in binary form only. Their only penalty is that they lose OSI certification. There is no legal enforcements in the MIT license to require the source be included with the dist or be made publicly available. And OSD certification is not enforcable, it can only be revoked. so, all I'm saying is that if someone looks at the OSD and likes it, they can't just go and pick any OSI approved license and have it give legal enforcability of all the OSD bullets. If I pick the MIT license, then OSD #2 is not enforcable. The OSI certification can only be revoked. Other than revoking the certification, I see no teeth to OSI certification. any other disputes then drop down to the level of the license, which may allow actions that are against the OSD. I don't care what the spirit of the OSD is. You can force-feed all the common sense you want about the spirit of the OSD. But the OSD is not a license. And it is the license that controls how a distribution may be re-distributed. Other than the threat of revoking teh OSI certification mark from a distribution, OSI has no recourse to enforce the spirit of the OSD if the license used in the distro allows said spirit to be broken. Unless there is some other implication of enforcemnet to OSI certification that I am unaware of. Greg -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: click, click, boom
Rick Moen wrote: begin Greg London quotation: If someone puts out a bunch of source code under the MIT license, and the distro is OSI certifiable, there is nothing to prevent someone else from redistributing it in binary form only. Their only penalty is that they lose OSI certification. _Licences_ are OSD-certified. Software is open-source or not, in accordance with its nature (including but not limited to licensing). http://www.opensource.org/docs/certification_mark.html The OSI Certified mark applies to software, not to licenses. Licenses can be approved by OSI, but that does not guarantee certification on a piece of software. Do you have a point, or are you simply ruminating on the vagaries of power and influence? point (1) it is not clear by the OSI website that there is a distinction between approved licenses and certified software. You confused the two above yourself. someone who has a program they wish to license could come to teh OSI website, do a quick readthrough, and come away with the understanding that an approved license will guarantee the OSD is legally enforced in the license. With 26 licenses, some of them extremely long, most people will not read all of them,nor understand the implications of them. I skipped over to the OSD, read that, and assumed that I could pick any approved license, and the OSD would be enforced. point (2) This is at the root of the whole yet another public license discussion. OSI has little incentive to approve YAPL, since OSI's only contribution to open source is through it's certification of software. Approval of another license is independent of software certification. programmers have little incentive to get OSI certification, because it does little measurable for them. The only thing that is concrete for the developer is the wording of the license. Therefore you get all these developers trying to get a slightly modified version of a license approved. A developer submits a license to OSI and says none of teh currently approved licenses do exactly what I want. OSI (or at least a number of people on this list) respond you should just use an already existing license and get certified Developers want certification, but they also know enough that they want a legal license that gives them what they want. OSI will certify software under the MIT license, which effectively means that OSI will certify software licensed in such a way to give away all rights. So it is no surprise that OSI isn't too concerned about YAPL that splits some fine hairs between this right and that right. OSI will certify something that licenses away all rights. Why worry about whether or not you can distribute changes in the form of patches? The MIT license says you can do ANYTHING. Greg -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: books about free software open source movement
[EMAIL PROTECTED] wrote: Who can tell me which books about free software movement and open source movement are popular in America, or worth reading for those who are interested in these movements? Could you list about five books? Thanks. I can tell you 1 offhand. Open Sources by O'Reilly. Greg -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: GPL vs APSL (was: YAPL is bad)
David Johnson wrote: On Monday 24 September 2001 11:08 am, Greg London wrote: You err slightly in (B). It does not mean that the source code must be made equally available to those without the binary. you missed my following paragraph that said (paraphrasing) :(B) somewhat implies public distribution :but it is possible to distribute a binary :and give a well publicized URL that requires :a password to get the source. One thing Bob can't do, according to OSD, is fix a bug in Alice's code, send her a binary that works, and taunt her, saying I'll send you the source for a million bucks. Once Bob sends Alice a binary, he must make the source available to her. In the case of the MIT license, Bob certainly *can* charge Alice a million bucks for the source, but the license would still be an Open Source license. It seems to me that the MIT does not meet item #2 of the OSD, then. The APSL goes above and beyond #2 requirements. But the MIT license seems to fall short. OSD #2 seems to be setting a clear minimum requirement that source code must be included with any distribution, or be made publicly available. looking at the MIT license, it seems to simply waive all rights and throw in a No Warranty clause. IANAL TINLA GREG -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: OSD #2 (was Re: GPL vs APSL (was: YAPL is bad))
Bruce Perens wrote: Both the MIT license and Public Domain fit under both the OSD and RMS's definition of Free Software, is it possible to take GPL'ed code, modify it, relicense it under a proprietary license, and distribute it only in binary form? my understanding is it is not possible. but MIT'ed code would allow this. The 'bar' to meet GPL is pretty high. The bar to satisfy the MIT license is on the ground. The OSD is somewhere in between. one can argue about RMS's definition of free software, but his implementation (the GPL license) set the bar a LOT higher than MIT. One could also argue that RMS's definition is moot to OSI, since OSI has it's own definition to follow. and to change the OSD to exclude them would be a travesty. travesty smavesty. I'm not saying exclude GPL. Never did. I said GPL exceeds the minimum requirements given by the OSD. I have no qualms with GPL being OSI approved. You're not talking to that which I am saying. I am saying the MIT license does not meet OSD #2. Since OSD #2 says the program MUST include source code There is nothing in the MIT license to guarantee OSD#2, so it fails to meet the definition. And OSD#2 requires this, because it uses the word MUST. I didn't put it there, OSI did. Therefore OSI should not have approved the MIT license, since MIT does not satisfy this requirement. OSI put the bar at a certain height, and the MIT license limbo'ed right under it. It isn't about it being a travesty. It's about whether or not OSI followed its own definition. If OSI certification gives absolutely no guarantees about code licensed with an OSI approved license, then OSI is moot. IMHO IANAL TINLA YRMV Greg YRMV= Your Results May Vary -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
click, click, boom
Ah, several items just fell into place. 1) The OSD and the OSI approved licenses (AL) are totally independent. 2) Some of the OSI AL's also happen to meet the OSD definition, and some do not. But OSI does not determine this. 3) OSI approved it's licenses not because of how they measured against the OSD, but because they were simply the most common, or someone requested it. 4) OSI certification requires that you A) license your software with an OSI approved license B) distribute your software in a manner that comlies with the OSD. 5) OSI certification does not mean your license meets the OSD, but that your software distribution meets the OSD. 6) You can use an OSI approved license and not be OSI certified. See, this clears up a -whole- lot of confusion. I thought the OSD was somehow related to the approved OSI licenses. Instead, OSI approved its first four licenses simply because they were the most common. The OSD is independent of the licenses. Getting software off the net that uses an OSI aproved license does not mean that software is open source, in the OSD sense of the term. Rather, if you want to know if some software off the net meets the definition of open source, it must be OSI certified. So, to put it in a nice little blurb that could go somewhere on the OSI web page: (probably the approved licenses page) Just because a license is approved does not mean the license enforces the OSD. However, to get OSI certified, a program must be licensed under an OSI approved license. If an OSI certified program is re-distributed in manner that does not meet the OSD, OSI's only recourse is to revoke certification. Not all OSI approved licences require that re-distributions meet the OSD. Choose your license carefully. Greg -- Greg London x7541 -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
The Cathedral and the Bazaar
(an allegory) And the Bazaar was hustling and bustling with comings and goings, a hustle and bustle, and a great many things all happening at once. A woman came through the great morass with a push cart. She found an open space on the walk and started to unload her wares. She struggled to set up her table, and then covered it with a multitude of pretty, and interesting things. She then put a large, brightly, colored umbrella through a hole in the center of the table, which put her table in the shade. A man at the table next to him began shouting at her and ran over. You cannot put an umbrella in your table. Why not? it is a sunny day, and I will surely get sunburn. We at the bazaar do not take lightly of such structures. These things lead to Cathedral-like ways of doing things. He flashed an angry side glance at the Cathedral across the corner. The man returned his gaze to the table and suddenly became angrier than before. A ROUND TABLE! What are you doing bringing a round table to the bazaar!?!? The woman explained. It is easier for me to roll a round table where I want it, than it is to carry a square table on my back. That is not the way of the bazaar! You must have a square table if you are to trade your wares here! Where does it say I must do such a thing? The man reached in his pocket and angrily waved the paper in front of her. Right here! He pointed to the top item. Square tables are approved for bazaar operators. But that doesn't mean round tables aren't approved or cannot be approved. Don't be absurd. There are 11 approved items on the bazaar's list. Round tables are not one of them. see for yourself. He threw the list at her. The woman was busy reading it when a loud racket nearby distracted them both. The man took the list from the woman and ran over to see what the ruckus was. Another woman was setting up a tent at her site, and a pole had fallen to the ground in a kabang. The man began screaming. What are you doing!? Tent's are not allowed in the bazaar! But on days it is windy, a tent is much more comfortable. And my merchandise is so small and light, that even a small breeze would blow it all away. But tents have not been approved! he pointed angrily at the paper in his hand. Don't you understand? Tents have not been approved for the bazaar. Who died and made you high priest? open-air is not the only way of trading at the bazaar. I need a tent. The umbrella woman came over and interjected. And I want an umbrella and a round table. Well, you cannot have them nor should you want them. They have not been approved, and for good reason. We can't just have a bazaar of vendors, selling and trading their wares any way they feel like it. There are specific established ways of doing things here. The two women looked at each other. Come on, sis. We're obviously not welcome here. She motioned her to follow. Maybe we can do it our way at the Cathedral. And they _ALL_ lived miserably ever after. The End. Author's Prologue: We realized it was time to dump the confrontational attitude that has been associated with free software in the past and sell the idea strictly on the same pragmatic, business-case grounds that motivated Netscape. quoted from: http://www.opensource.org/docs/history.html Finis Greg London with apologies in advance to Eric Raymond. -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: GPL vs APSL (was: YAPL is bad)
% OSD 2: % The program must include source code, ... % % When some [program] is not distributed with % source code, there must be a well-publicized % means of obtaining the source code % % preferably, downloading via the Internet without charge. so, according to OSD, you have two options EITHER (A) you distribute source and binary in one kit. OR (B) you distribute a binary in one kit. and you make the source code freely available. (preferably downloading for free on the net) those are the only two options of the OSD. (B) somewhat implies public distribution by mentioning internet and well-publicized. but it is possible to distribute a binary and give a URL that requires a password for those you allow to download the source code. and such a way could be well-publicized even though the password is not public. Option (A) does not, in any way, require publicity. Alice could download some software off teh net. Alice could modify it and send the Binary and Source to Bob. Bob and Alice could be employees of Charlie who owns a Corporation. Charlie has Alice adn Bob sign a NDA, saying they will not distribute any of their software outside the company. As long as all distributions within said company include source code from employee to employee, the source does not have to be made public. And satisfies the Open Source Definition. One thing Bob can't do, according to OSD, is fix a bug in Alice's code, send her a binary that works, and taunt her, saying I'll send you the source for a million bucks. Once Bob sends Alice a binary, he must make the source available to her. I don't have a problem with this. And I think allowing this makes corporations warmer to using Open Source within their company. The OSD does not require that all modifications be sent to the author. Nor should it require all modifications be public. The APSL raises the bar. It takes the OSD and makes it even more open. Which is fine by me. See my allegory sequel to the cathedral and the bazaar. Some dealers at the bazaar want square tables and some want round tables. IANAL Greg -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: The Cathedral and the Bazaar
Subtitle: Attack of the Bazaar-Nazi's http://crynwr.com/cgi-bin/ezmlm-cgi?3:mss:3919:200109:eocppecokloobmpbdmgf Also, change teh number 11 to 26 in the text. I did not mean to imply the OSD bullets, but the OSD approved licenses themselves. Greg -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: GPL vs APSL (was: YAPL is bad)
Whether or not you think Apple's behavior is legitimate, do you now agree that there is a real behavioral difference between the GPL and the APSL on this score? civil action is not the differntiating factor. Apple's behaviour is no different than if it were GPL, and John failed to produce source code at John's request. FSF would pursue him similarly. The OSD requires the source to be given to *at a minimum* anyone who recieves a distribution. which means universal distribution is not required. APSL requires that regardless of distribution, you must publicly distribute your code. The difference is APSL does not give you the option of limiting source code to people to whom you give your distributions. OSD allows source code to be contained within a circle of friends. IANAL Greg London -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: Backlog assistance?
Alex Stewart wrote: If the point is to provide a few good, clear-cut licenses for people to choose from, that's one thing, and suggests the OSI should be very picky. If the goal is to encourage open-source licensing terms amongst the software community, that's very different, and suggests that the OSI should (try to) encourage (and thus certify) anything that meats the open-source requirements. IANAL, but simply from a development point of view, OSI does not appear to be taking advantage of some of open-source's best feature: patches and evolution currently, all OSI certified licenses are One-Off applications, written from scratch, no reuse, no inheritance, no nothing. And their speed at which the approve licenses seems to be in line with that of an organization who is testing multiple, and completely independent, applications. And verification is basically what OSI does, except it's in a legal manner versus software manner. The OSD is the spec, everyone sends in their personal interpretation of that spec, and then OSI has to verify the license completely meets the spec. in order to make it easier to create your own license, perhaps someone could come up with a minimum code base and people could add to it. The OSD has 11(?) requirements. how hard would it be to come up with a minimal license that defines these requirements. THen if you want to create your own license, you inherit the minimal license and add to it. Copyright law only grants so many rights to the author, (something like copy, modify, distribute, display, public perform, but this is from vague memory) And there are only so many variations of things you can do with these rights and still be considered open source. And some are tied directly into the OSD itself copy and distribute are guaranteed. And either Modify or distribute with patches must be included as well. The biggest one seems to be that if yuo modify and redistribute, that new distribution must be under the same license as the original, to prevent code hijacking. could it be possible (and more importantly, of value) for OSI (or someone) to develop a license developers kit which includes a base-line license, and possibly, additional items from which people can select. throw in a good No Warranty clause. give a selection of what country's laws any disputes will be settled in. People could use that as a base license, select the required alternations, and then add an addendum area where they put their specific licensing requirements. (GPL could put it's political rant in it's ammendment section, for example) And the base-line license could say that whenever the ammendments conflict with the base license, the base license wins (or whatever the legal term is). Which should then prevent people from hijacking the base-license. And license incompatibilities with one another should be minimized. It would also create some unity among different variations of licenses. Currently, if the GPL were attacked legally, the people who use the Artistic License would have little at stake. but if a number of different groups use the same base-license, and the base license is legally attacked, there would advantage for a combined legal defense from all users of the base-license. and more importantly (benificially), OSI wouldn't have to manually approve every variation under the sun. This would allow OSI to gain the same benifits that open-source software has, and apply them to licensing development and approval. just a thought. IANAL Greg -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: Contract or License?
Lawrence E. Rosen wrote: Copyright law does not prohibit use. It prohibits reverse engineering (and similar activities) under certain circumstances. I didn't intend to be subtle about the meaning of the word use. but reverse engineering was part of 'fair use' before DMCA, wasn't it? Greg London -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: copyright discussion
[EMAIL PROTECTED] wrote: I'm a newcomer.Could tell me what the DMCA is? good grief. your search engine must have flooded its carbeurator. cause mine came up with a bazillion hits with just 'dmca'. (yes, exactly 1 bazillion hits, no more, no less.) ;) but to give you a jump start, I think the most relevant site is the one that shows the real implications of DMCA, namely that people can go to jail for knowing too much: http://www.eff.org/alerts/20010808_eff_sklyarov_alert.html you might want to go to the www.eff.org website and read for a while. They have a number of articles that may be of interest. Greg -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
