Weekly Edition
Return to the Announcements page
| |||||||||||||
|
| |||||||||||||
Tiemann: time to start enforcing "open source"
Michael Tiemann, the leader of the Open Source Initiative, has come to the conclusion that it is
time to start coming down on companies which falsely use the "open source"
label. "We should never put the customer in a position where they
cannot trust the term open source to mean anything because some company and
their investors would rather make a quick buck than an honest one, or
because they believe more strongly in their own story than the story we've
been creating together for the past twenty years. We are better than
that. We have been successful over the past twenty years because we have
been better than that. We have built a well-deserved reputation, and we
shouldn't allow others to trade the reputation we earned for a few pieces
of silver."
(Log in to post comments)
Tiemann: time to start enforcing "open source" Posted Jun 21, 2007 16:37 UTC (Thu) by sfeam (subscriber, #2841) [Link] What a crock. Who are the OSI, that they think they can impose adefinition of common English language words on the rest of the world? To me the OSI definition of "open source" reads like a "me too" version of the FSF philosophical boilerplate. They are of course welcome to define their own terms and advocate for whatever they like. But it is extremely arrogant of them to think that the rest of the world must adopt a definition that patently does not match the obvious and common usage of the term it defines. And it is beyond arrogant to refer to someone who doesn't accept your definition as "flagrantly abusing" it.
This smells like a side-skirmish in the ongoing GPL2 v. GPL3 shouting
A simple Google search turns up no lack of other institutions who give
Tiemann: time to start enforcing "open source" Posted Jun 21, 2007 17:56 UTC (Thu) by rickmoen (subscriber, #6943) [Link] Who are they? OSI are of course the people who invented and promoted that consensual, carefuly definined meeting of "open source" in the software context. And a damned good thing that's been, too. (A search a few years ago by an obsessive on the license-discuss mailing list turned up, if memory serves, two very slightly earlier, arguable proto-uses by individuals on Usenet, predating OSI's, but they never went anywhere with that concept.) As to uses in other fields entirely: I'll bet you can figure out the relevance of context, if you work at it. By contrast, if you can't figure out why it's important to discourage description of blatantly proprietary codebases as "open source", then I'm afraid there's little basis for further discussion. Rick Moen
Tiemann: time to start enforcing "open source" Posted Jun 28, 2007 13:22 UTC (Thu) by lysse (guest, #3190) [Link] What case does the OSI make for the importance of redistribution, as it relates to the term "open source"?
Tiemann: time to start enforcing "open source" Posted Jun 21, 2007 18:46 UTC (Thu) by EmbeddedLinuxGuy (guest, #35019) [Link] it is extremely arrogant of them to think that the rest of the world must adopt a definition that patently does not match the obvious and common usage of the term it defines. There is no "obvious" definition for the phrase "open source", any more than for "object-oriented" or "realtime". Somebody introduced these terms and gave them their specific meanings. To give the terms other contradictory meanings after they've become accepted does not indicate a good-faith effort to communicate. There are many real-world cases where the overriding and critical property of of open source code is, in fact, simply that it is available for inspection. This is an important freedom, but it should be described with a different term such as "source code available", since "open source" already has a different well-established meaning. The NIH has its own definitions of these terms that do not, as I read it, fully agree with the OSI definition. Please provide a reference to substantiate this.
Tiemann: time to start enforcing "open source" Posted Jun 21, 2007 19:31 UTC (Thu) by sfeam (subscriber, #2841) [Link] The NIH has its own definitions of these terms that do not, as I read it, fully agree with the OSI definition.Please provide a reference to substantiate this. Here is an example: http://grants.nih.gov/grants/guide/rfa-files/RFA-GM-05-010.html The RFA refers to "the open source code policy specified in Sharing Research Resources under Section IV 6 of this document", and you can find a 4 point definition in that section. I don't know to what extent different RFA documents use identical language, but it is common to have some such clause present. Please note that I am not taking issue with the admirable goals of the OSI, nor am I saying that the NIH policy is necessarily better. I am simply trying to point out that the OSI does not have a monopoly on the term "open source", and their definition is not the only one used. I would have no problem with strict enforcement against false claims of using an "OSI-approved license", or similar wording. But trying to legislate the use of ordinary words in a language calls to mind the story of Canute and the tide.
Tiemann: time to start enforcing "open source" Posted Jun 21, 2007 23:01 UTC (Thu) by EmbeddedLinuxGuy (guest, #35019) [Link] http://grants.nih.gov/grants/guide/rfa-files/RFA-GM-05-010.html Thanks. The three points under "Sharing Research Resources" are termed "goals for software dissemination", not a definition of open source. They comprise a significant subset of the OSI definition; for example the license "should permit the commercialization of enhanced or customized versions of the software", which is where many pseudo-open licenses fail (e.g. Scilab, Cobia). If I wrote a license which (for example) prohibited use of my software by companies who compete with me, the NIH might or might not approve my grant -- but it would be inappropriate for them to call my software "open source". Why? The term "open source" is deliberately intended to put my project in the same category as Linux, Apache, and other open source success stories. A big part of their success derives from the freedoms granted by their licenses; I don't think IBM would want to invest its resources in Linux if it were under the Scilab license. To misapply the term "open source" to licenses which don't provide these freedoms is false advertising.
Available for inspection? Posted Jun 22, 2007 5:13 UTC (Fri) by BrucePerens (subscriber, #2510) [Link] Code that is available for inspection is called "Disclosed Source Code".
Tiemann: time to start enforcing "open source" Posted Jun 21, 2007 18:31 UTC (Thu) by sab39 (guest, #2185) [Link] I agree with just about all of Tiemann's spiel except for one point: his suggestion that we should push for the definition of "open source" to be "licensed under an *OSI approved* license".
The OSI are the stewards of the Open Source Definition which they actually use to determine whether they approve a license or not. So they *have* a definition, which works and has been honed and become well understood over the years.
There are licenses that meet the OSD but are not OSI approved, either because the OSI have not yet finished the process of making a determination to approve them or not, or because the OSI rejected them on grounds that do not actually compromise their Open Sourceness (license proliferation, for example) or because they have not actually been formally submitted to OSI for consideration at all. (In this last category, I'm thinking particularly of Microsoft's Permissive License, which is generally considered to meet the definition of Free Software / Open Source, but which OSI has so far declined to evaluate).
I completely agree that we should reject claims to co-opt the phrase Open Source to refer to licenses that do not meet the OSD/DFSG. But I reject the redefinition of Open Source to exclude licenses that meet the OSD, but haven't been formally approved.
Tiemann: time to start enforcing "open source" Posted Jun 21, 2007 21:53 UTC (Thu) by Los__D (subscriber, #15263) [Link] (In this last category, I'm thinking particularly of Microsoft's Permissive License, which is generally considered to meet the definition of Free Software / Open Source, but which OSI has so far declined to evaluate)." Where did you read this? I searched a bit, and the only place that I found anything saying they refused, was another post by you on opensource.org, and a Slashdot article about someone other than MS submitting it, OSI asking MS if they should go ahead, and MS asking them not to. The only other mentions I found with MS-PL and OSI, was articles of OSI members (Including Tiemann), encouraging Microsoft to submit it for evaluation.
Microsoft Permissive License, evaluation of Posted Jun 21, 2007 22:12 UTC (Thu) by kmself (subscriber, #11565) [Link] Your best source of information is likely the license-discuss mailing list and its public archives. OSI policy IIRC is that licenses are only reviewed on submission by the license author, on general principles of preservation of bandwidth and not evaluating drafts of not-yet-complete licenses. It may be that the license has not been submitted by Microsoft. Guidelines: http://www.opensource.org/approval Formal acceptance decided at OSI board meetings (see above). Google shows no hits on the license title at the l-d archive.
Microsoft Permissive License, evaluation of Posted Jun 22, 2007 3:10 UTC (Fri) by rickmoen (subscriber, #6943) [Link] kmself wrote:It may be that the license has not been submitted by Microsoft. This surmise is exactly correct. Rick Moen
Microsoft Permissive License, evaluation of Posted Jun 22, 2007 7:34 UTC (Fri) by Los__D (subscriber, #15263) [Link] - Then isn't this is more a case of Microsoft refusing to submit it?
Microsoft Permissive License, evaluation of Posted Jun 22, 2007 12:47 UTC (Fri) by sab39 (guest, #2185) [Link] Microsoft has no particular desire to be OSI-certified Open Source. It's not in their interests, because they'd much rather be able to spread FUD around Open Source in general without advertising the fact that they're doing it themselves.
On the other hand it is in our interests as a community to be able to take advantage of their code when it IS open source - for example, recognizing that we have the option to include their "ASP.NET Ajax toolkit" or "Dynamic Language Runtime" (running on top of Mono) in our distros without compromising our principles.
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 12:57 UTC (Fri) by sab39 (guest, #2185) [Link] I said "declined" to evaluate rather than "refused" for a reason, although admittedly I might have been a little over-subtle in the distinction I was trying to draw.
If Microsoft had submitted it for evaluation and they'd said "no we won't, because you're Microsoft", then I'd have said "refused". What I meant, instead, was that they haven't chosen to take the pro-active step of evaluating it regardless. They have no particular obligation to do so, but the result is that there's code out there that (probably, based on the informal analyses I've read) meets the OSD, but isn't OSI certified. Taking what Tiemann said literally, he's asking that we not refer to that code as Open Source, and that's the bit I'm objecting to.
There's at least one example of the OSI evaluating a license without the request of the license author: I'm *very* sure that the FSF never submitted the GPL for OSI approval ;)
(I was disappointed to not find the MS-PL on the FSF's license-list page last time I checked, too. I'd feel much more secure about the fact that Mono is planning to depend on that code if I knew that the FSF, OSI or Debian had evaluated it and found it acceptable...)
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 14:39 UTC (Fri) by Los__D (subscriber, #15263) [Link] "There's at least one example of the OSI evaluating a license without the request of the license author: I'm *very* sure that the FSF never submitted the GPL for OSI approval ;)"
OSI can display the GPL on their site, as the GPL has a license to make verbatim copies of the GPL itself. MS-PL has no such permissions, so OSI has to get permission from the author (MS) to include it in their collection of licenses.
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 3:19 UTC (Fri) by rickmoen (subscriber, #6943) [Link] sab39 wrote:I reject the redefinition of Open Source to exclude licenses that meet the OSD, but haven't been formally approved. I'd speculate that, if you asked Tiemann about this point, he'd say "Oh, sorry, I didn't mean to imply that licences that are plainly OSD-compliant on the basis of their plain language must nonetheless be considered proprietary until formally approved, and will be subject to tsk-tsking in the interval." Nobody at OSI has in my experience held such a view before; I doubt they do now. For obvious reasons, they do however encourage authors to submit candidate licences for certification (if not duplicative of existing options, etc.). Rick Moen
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 13:04 UTC (Fri) by sab39 (guest, #2185) [Link] That's exactly what I'm hoping he meant, but it does go contrary to the language of what he wrote in his article. I did post a comment on the article too, I'm hoping he'll respond. But what he wrote was specifically:
"So here's what I propose: let's all agree--vendors, press, analysts, and others who identify themselves as community members--to use the term 'open source' to refer to software licensed under an OSI-approved license."
I'm saying, replace "OSI-approved" with "OSD-compliant", and I'm all in.
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 17:27 UTC (Fri) by rickmoen (subscriber, #6943) [Link] sab39 wrote:I did post a comment on the article too, I'm hoping he'll respond. I fear you're getting bogged down in trivia, here, and risk getting sidetracked from Tiemann's actual point, which concerns the problem of firms misappropriating the term "open source" for licenses and deployment scenarios that are clearly proprietary. That is a serious problem, especially in the ASP field. Please consider focussing on that instead of on fine points of rhetoric -- unless and until you see OSI harassing users of obviously OSD-compliant but as-yet-uncertified licences, which scenario in fact never happens. Rick Moen
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 18:10 UTC (Fri) by sab39 (guest, #2185) [Link] I didn't mean to give the impression that this is any more than a nitpick. I said in my initial comment here that I agreed with his entire spiel except that one point, and in my comment on his article I was even more explicit that I'm grateful to him for posting it, I think it's tremendously important and I'm wholeheartedly in favor of cracking down, as much as possible, on people abusing the term "open source". In fact I've been doing that (within my own tiny sphere of influence) myself for years.
Having said that, I don't think that the fact that it's a nitpick means it can just be summarily dismissed. He could have said "OSD-compliant", but he didn't, he said "OSI-approved"; he's replied to various comments on the article (but not to mine) and in those replies still seems to be sticking to the same line, which suggests to me he didn't mis-speak but meant what he wrote.
I know that my tendency to overemphasize nitpicks is a character flaw which often leads to not adequately communicating the fact that I *really do* agree and feel very strongly about my agreement with his main point. I'm not sure how to avoid it, though, without just ignoring the nitpick entirely - which I do still think is a valid and significant one.
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 8:29 UTC (Fri) by copsewood (subscriber, #199) [Link] Given the fact that they didn't get a trademark on the term "Open Source" but they did get an "OSI Certified" mark, there is no point in Tiemann and others at OSI complaining about missattribution of "open source" given that they clearly can't enforce it, and if they want to talk about enforcement as opposed to encouragement they will have to start trying to get people to respect the term "OSI certified". It seems that OSI need to sort their own act out and decide whether the term they want to promote is the one more frequently but loosely used or the term they have registered and so can feasibly protect. Software suppliers who use the "Open Source" term misleadingly are very unlikely to be worried about what the free software community thinks about their actions in any case. The whole thing smacks of lack of direction at OSI.Here is an article about the OSI Certified mark registration.
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 10:14 UTC (Fri) by jfj (guest, #37917) [Link] Ok, there are two things in the article:
1) Using the term Open Source for projects which are anything but Open Source is something we don't like.
2) What can we do about it.
Now, do we agree on (1)? Is Tienmann convincing that (1) sucks? To me, yes, at least.
Now once we agree on (1) we can talk about (2).
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 11:31 UTC (Fri) by copsewood (subscriber, #199) [Link] OK I agree that community pressure can be effective and should be used in cases where we are dealing with people who care about what we think. But that is just a proportion of the market or user community. The problem is that the other part of the market can only be brought into line based on a different term, the term that was or could in future be trademarked. The problem arises when a term acceptable and understood within the community is different from the trademark that was obtainable. This should have all been sorted out far, far earlier as it makes the whole idea of an "open source movement" look ridiculous to outsiders if we can't defend out basic definitions.
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 12:41 UTC (Fri) by maxxedout (guest, #45888) [Link] > didn't get a trademark on the term "Open Source"
That statement is contrary to this Press Release:
Issued by and for the Board of Directors of OSI
http://www.opensource.org/pressreleases/osi-launch.php
BD
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 12:57 UTC (Fri) by copsewood (subscriber, #199) [Link] Yep. Read it carefully. The press release is dated 1998. They applied to trademark "Open Source" but this application failed. So they registered "OSI Certified" instead. See also:
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 14:23 UTC (Fri) by maxxedout (guest, #45888) [Link] It's a shame that there are no links to any Articles, Emails or Press Releases regarding this issue from the People that were actualy involed with the trademark application procces. Unless of course I missed it.
BD
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 17:19 UTC (Fri) by rickmoen (subscriber, #6943) [Link] copsewood wrote:Given the fact that they didn't get a trademark on the term "Open Source" but they did get an "OSI Certified" mark, there is no point in Tiemann and others at OSI complaining about missattribution of "open source.... This conclusion simply does not follow, and is in fact severely mistaken. The bad publicity that attaches to firms when they're hounded by members of the source community for deceptive use of the term "open source" has proven to be a serious concern for such firms, especially when they find out to their distress that we don't give up and don't go away. This is a significant tool; we use it because it works. Your assistance in that area, instead of just posting cheap nihilism, might be useful. Rick Moen
Tiemann: time to start enforcing "open source" Posted Jun 22, 2007 21:04 UTC (Fri) by copsewood (subscriber, #199) [Link] The point I've been making is that the term "enforcing" (please check the use of this term in the lead article) is inappropriate when legal moves to make enforcing possible through a trademark have been misdirected. I'm all in favour of gentler forms of persuasion as has also been suggested by others in this thread. But the problem with OSI's management is 1. they have chosen to persuade well-disposed people including myself to use a term "open source" which they can't enforce while they could enforce the mark "OSI Certified" and 2. OSI have not used their standing to persuade those they can influence to use "OSI Certified" in place of the term "open source", once it became clear to them that "open source" could not be legally enforceable. It must have become clear to OSI that "open source" could not be trademarked relatively soon after they stated their intention to register "open source" as a trademark in 1998. Yet they continue promoting the term "open source" and the consequences of this still hasn't dawned on them by 2007 ?This is a pretty fundamental bug in the advocacy and mark promotion and protection process, and trying to talk their way out of it or pretending it isn't there isn't going to fix this. Would Linus have achieved what he has if he had accused those pointing out bugs in his program of "cheap nihilism" rather than doing his best to accept bugs as such and fix the bug however fundamental to previous product designs ? This bug can unfortunately only be fixed either through some significant re-engineering or by giving up on the idea of "enforcement" as opposed to "persuasion". Whether OSI can do this job remains to be seen, and until we find out, the Free Software Definition is likely to be the best measure we have, given that those abusing the term "open source" are much less likely to refer to the Free Software Definition as applying to a license not on the approved list, though the term "free software" is just as capable of abuse, e.g. by being applied to free as in beer unless they decide to sue you. I do hope OSI can sort their problem out, as having a non-profit organisation capable of promoting free software based on business case and engineering quality considerations, and able to apply a widely understood and respected mark to conforming products and defend it can only be a good thing. But I don't see this happening until OSI:
Tiemann: time to start enforcing "open source" Posted Jun 25, 2007 18:44 UTC (Mon) by rickmoen (subscriber, #6943) [Link] copsewood wrote:The point I've been making is that the term "enforcing" (please check the use of this term in the lead article) is inappropriate when legal moves to make enforcing possible through a trademark have been misdirected. The term "enforcing" appears nowhere in Tiemann's piece. If you didn't bother to read what Tiemann wrote, and instead are sounding off about one word in Jon Corbet's article headline, then your error and digression from the actual matter at hand are explained but not excused. And your assumption that OSI cannot and should not assert its entirely just claim to definining "open source" in the context of software is simply mistaken. With our help, it can and does. Rick Moen
U.S. government defines "open source software" as basically Free Software Definition Posted Jun 28, 2007 15:11 UTC (Thu) by dwheeler (guest, #1216) [Link] The U.S. government has an official definition of what the term "open source software" means, and it's basically the same as the Free Software Definition. The U.S. government's OMB M-04-16, "Software Acquisition", signed July 1, 2004, says that "Open Source Softwares source code is widely available so it may be used, copied, modified, and redistributed. It is licensed with certain common restrictions, which generally differ from proprietary software."At the least, the notion that you can say software that can't be redistributed is "open source software" is nonsense.
|
|
||||||||||||