Talk:Comparison of free and open-source software licenses

This article was nominated for deletion on 2006 July 19. The result of the discussion was keep.

This article is rated List-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects:

Computing: Software / Free and open-source software Low‑importance This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.ComputingWikipedia:WikiProject ComputingTemplate:WikiProject ComputingComputing articles Low This article has been rated as Low-importance on the project's importance scale. This article is supported by WikiProject Software (assessed as Mid-importance). This article is supported by Free and open-source software (assessed as High-importance). Open (inactive) This article is within the scope of WikiProject Open, a project which is currently considered to be inactive.OpenWikipedia:WikiProject OpenTemplate:WikiProject OpenOpen articles

Here are some tasks awaiting attention: Cleanup : Treat licence compatibilites outside of the Approvals section

On 28 March 2022, it was proposed that this article be moved from Comparison of free and open-source software licences to Comparison of free and open-source software licenses. The result of the discussion was moved.

The problem with this page, although well-intentioned, is that a global opinion about license compatibility is just that: global, and an opinion. Whether one license is compatible with another really isn't your, mine, or his decision to make. It is a decision that only a licensor can make. I may think that the BSD license is compatible with the GPL. If I do, then it doesn't matter if somebody else says it isn't, because I'm not going to initiate a lawsuit. Whatever the licensor says, goes, even if everybody else disagrees with him. RussNelson 01:12, 30 April 2006 (UTC)[reply]

That's a very interesting point of view

But i prefer to focus on the legal aspect only,that is to say if there are compatible or not

A text explaining your point of view can be added

We can also talk about additional permissions such as in the gpl where someone can give aditional permission on his code

This could be useful in order to cope with incompatibilities

00 tux 22:20, 4 May 2006 (UTC)[reply]

Sorry, 00tux, but you cannot wiggle out of this philosophical problem simply by claiming that you are focussing on the legal aspect only. It's not enough to say whether a license is compatible or not. You have to say who says it, and who will listen to them. So, when the FSF says that the Apache 2.1 is not GPL compatible, and yet the ASF says that it is, who is right? You can't say who is right -- not if you want a NPOV. You have to say who says it, and who will listen to them. RussNelson 02:47, 5 May 2006 (UTC)[reply]

But there are global opinions pertaining to some of the licenses. If there's any debate as to whether certain licenses are compatible, their cell can just say 'contested', right? --Snarius 05:01, 17 June 2006 (UTC)[reply]

For me it`s interesting. —Preceding unsigned comment added by 84.183.187.69 (talk) 17:22, 15 May 2008 (UTC)[reply]

Great points. But in the end, in the US and most countries, it's the courts who decide what each license means. Anything that is vague enough to be argued about, won't be clear until the wording is changed or it is challenged in court. I like Snarius' 'contested' idea, or at least adding *lots* of footnotes. --GlenPeterson (talk) 12:45, 3 September 2008 (UTC)[reply]

I'm not really sure what to do with the gigantic table; it's not readable as it is, and there appears to be only one entry filled in... Cheyinka 10:16, 26 May 2006 (UTC)[reply]

This article is not worth putting any effort into it because of the philosophical problems I explained above. This article cannot be repaired because everyone who reads it has a POV that is important, and yet which may differ with the contents of the article. An NPOV article is simply not possible. Rather than try to achieve the impossible, better to not try. RussNelson 14:45, 26 May 2006 (UTC)[reply]

I've reduced its size somewhat. It might be easier to edit now. --Snarius 05:18, 17 June 2006 (UTC)[reply]

I added two columns on "license features" which treads dangerous close to NPOV. (Specifically, the default "yes/no" templates seem to suggest "yes" is good and "no" is bad, implying a value judgement.) In this case, I made "yes" be the least restrictive option, which is arguably a value judgement. Oh well. There's a reference to the chart this is based off (no original research) but I haven't really figured out a good way to link it. So right now it's a footnoot in the standard <ref> style. I don't really like that, but - oh well.

The table columns are a little wider than I'd like, since I tried to make the column titles as descriptive as possible. It may be better to choose simple titles for them and then add footnotes describing what is intended. — Xenoveritas 22:09, 16 October 2006 (UTC)[reply]

Obviously I meant "non-NPOV" above - blah. I should have noticed that earlier. — Xenoveritas 02:47, 27 October 2006 (UTC)[reply]

License features are interesting. I guess not all licenses are resistant to any law tricks, like foul tricks with patents? A feature is for me for example resistant against foul patent tricks, or compatible with GPL, or copyleft, or license change allowed, many more... What about a new table for this?

Microsoft has some shared source licenses now. Two or so got also approved. Shouldn we add those licenses to this overview? —Preceding unsigned comment added by 84.183.187.69 (talk) 17:26, 15 May 2008 (UTC)[reply]

The LGPL has a "Yes" in the "Link from code with a different license" column. While strictly true, it requires that you allow your code to be disassembled and reverse-engineered. I think this deserves at least a footnote... --GlenPeterson (talk) 12:53, 3 September 2008 (UTC)[reply]

No. It only require that the author of the combined work of an LGPL and non-LGPL work to not actively restrict users to reverse-engineer the LGPL part, and only the LGPL part of the work. The intent here is that the licensees should be able to modify the LGPL library part, and then re-link it with the application (like applying a security fix to a PDF parsing library being used in a browser). Any footnote would had to include the complete context, or it would confuse the reader. Belorn (talk) 07:46, 5 November 2012 (UTC)[reply]

Is it possible to relicense LGPL code as GPL code? It was possible with LGPL v2.1, wasn't it? Originalfmg (talk) 23:20, 4 November 2012 (UTC)[reply]

Yes. Any work solely licensed under the LGPL can be relicensed under the GPL (of the LGPL version). If its a combined work of LGPL and non-permissive licensed parts, then the combined work as a whole can not be relicensed. The purpose of allowing relicensing is mostly to allow copy and paste to be used from an LGPL source to a GPL project. Belorn (talk) 07:46, 5 November 2012 (UTC)[reply]

I'd like to see the details from this table [1] included in a "features" table on this article. I believe something like this would be very useful. --Hm2k (talk) 00:54, 29 October 2008 (UTC)[reply]

Doesn't CDDL have a copyleft similar to MPL? Shouldn't "Release changes under a different license" be "No" and red? I'm not 100% sure about this, thus I didn't want to change it before asking. (Puce77 (talk) 12:39, 16 June 2010 (UTC))[reply]

Ni Faiwaiza (talk) 05:10, 25 September 2013 (UTC)[reply]

Could someone comment on Common Public Attribution License? I'm no software licensing expert or lawyer so if anyone knows how to classify CPAL, it would be greatly appreciated! (Note, it claims it is a free software license, but I don't see it on here... unless it is the same thing as "Common Public License"? I don't think so though as CPAL seems more conditional... here's an example of a CPAL for a popular ESB software with a "community" open source version released under CPAL, and an "Enterprise" commercial version: http://www.mulesoft.org/licensing-mule-esb —Preceding unsigned comment added by Bcmoney (talk • contribs) 20:24, 15 September 2010 (UTC)[reply]

I think List of FSF-approved software licenses should be merged here. In addition, we should merge in the free software section of List of software licenses; there is no reason to maintain this information in 3 places. --KarlB (talk) 18:12, 20 June 2012 (UTC)[reply]

I agree entirely. In addition, after merging the useful content from List of software licenses, I would redirect it to Software license. The licenses under "commercial royalty-free" are open source by the common sense definition, and can go in this list. If that bothers anyone this list can be renamed to Comparison of software licenses, but in any case we ought to consolidate these lists into one. – Pnm (talk) 18:29, 21 June 2012 (UTC)[reply]

Agreed - maintain in one place only. However "commercial royalty-free" isn't necessarily the same as open source as the source may not be available. Leornian (talk) 22:58, 27 December 2012 (UTC)[reply]

Support No brainer, the list is a content fork. Pokajanje|Talk 17:27, 4 January 2013 (UTC)[reply]

I just went ahead & did it. Now somebody should unify the tables. Pokajanje|Talk 17:42, 4 January 2013 (UTC)[reply]

The tables and discussions only include the latest version of a given license - for instance GPLv3. But developers can (and do) use older versions. The GPL is a good example of this - many projects (e.g. Linux) continue to use the older version GPLv2 because they do not like the provisions of the newer one.

I suggest that the article be modified to include older versions that are still in wide use - at a minimum GPLv2.

DEJ — Preceding unsigned comment added by 128.29.43.3 (talk) 17:36, 16 April 2013 (UTC)[reply]

Should we synchronize the list of licenses in the article with the SPDX License List? --Matthew Raymond (talk) 21:59, 11 March 2014 (UTC)[reply]

Hi, I plan to remove the Copyfree column as not notable. Nothing is wrong with starting your own server and offering some free service, quite the contrary, it's brilliant. But an article with no external reference at all, and "legal experts" offering their opinion not recognized anywhere outside of this Wikipedia comparison, IANAL, this can't be a good idea, or can it? –Be..anyone (talk) 09:10, 7 March 2015 (UTC)[reply]

Update, after a 2nd AFD in April Copyfree is now a redirect to Free_content#Copyfree; still without any legal experts in sight. –Be..anyone (talk) 03:31, 4 May 2015 (UTC)[reply]

I tend to agree. There's no evidence to suggest it's a particularly noteworthy initiative. Andyops (talk) 07:05, 20 March 2018 (UTC)[reply]

Just noticed it was removed from Free content. I'm going to go ahead and remove it from this article too. Andyops (talk) 07:08, 20 March 2018 (UTC)[reply]

+1 to this. The initiative is not noteworthy. — Preceding unsigned comment added by 87.198.40.10 (talk) 11:58, 10 September 2018 (UTC)[reply]

The section Article comparison begins with a hidden inline comment This matrix/chart needs an introduction to explain clearly what each of the column entries means. In fact the complete table is bogus, I've removed three columns with no info at all (= all cells filled with yes). The rest of the table (12×11 cells) is still bogus, mostly saying that the five GNU licenses differ from the six other licenses in four columns, some unsourced details about trademarks, and 26 empty dunno cells. The best way to fix these issues could be to remove the section, or to transform the table in some referenced statements highlighting the differences. –Be..anyone (talk) 02:04, 25 March 2015 (UTC)[reply]

After a complete reorganization of the page (two instead of three tables) this comment is now obsolete. –Be..anyone (talk) 11:52, 27 March 2015 (UTC)[reply]

The row on the Eclipse Public License described linking, distribution and modification as "Limited", with footnote 7 pointing to the "eclipse public license version 1" without an explanation. What does "Limited" really limit? Daviding (talk) 15:47, 26 April 2015 (UTC)[reply]

Lots of licenses say noting about trademarks, e.g. BSD and MIT. [Or anything about patents, "Patent grant" would be "No", except no free or open-scource license would explicitly NOT grant patents so I go with "Manually".] Should the column be "Grants TM"? And is this unrelated to:

XCore Open Source License: "Neither the name of XMOS, nor the names of its contributors may be used to endorse or promote products derived from this Software or the Documentation without specific prior written permission of the copyright holder."

And it's Hardware License Agreement:

"Neither the names of XMOS, nor the names of its contributors may be used to endorse or promote products derived from this Open Design or the Documentation without specific prior written permission of the copyright holder." comp.arch (talk) 18:51, 3 May 2015 (UTC)[reply]

Seems MPL does not grant trademarks (similar to Apache). See sections: 2.1, 2.3 "This License does not grant any rights in the trademarks, service marks, or logos of any Contributor (except as may be necessary to comply with the notice requirements in Section 3.4)." and 3.4 comp.arch (talk) 19:04, 3 May 2015 (UTC)[reply]

The Approvals table says that OSI approved the "Apple Public Source License 1.x", but I can’t find it on OSI’s site (they seem to have only approved the version 2 of this license). — Preceding unsigned comment added by 84.162.34.74 (talk) 16:52, 1 July 2015 (UTC)[reply]

...since the whole point of copyleft is to remove all restrictions... can we change copyleft to blue in the table? I think that would differentiate it nicely from both the non-reciprocal permissive and the restrictive licenses. — Preceding unsigned comment added by Thgntlmnfrmtrlfmdr (talk • contribs) 03:15, 17 March 2016 (UTC)[reply]

Cyan would be better I think after thinking about it a little. — Preceding unsigned comment added by Thgntlmnfrmtrlfmdr (talk • contribs) 03:20, 17 March 2016 (UTC)[reply]

To be more neutral, it is better to use "Free/Libre Open Source Software". See Stallman's reasoning here: https://www.gnu.org/philosophy/floss-and-foss.html It is a reasonable change I think. Free is not gratis, or free as in free beer but as in freedom ;-) Filiprino (talk) 11:44, 17 May 2017 (UTC)[reply]

See my longer answer at Talk:Free and open-source software § Move to FLOSS. There are many reasons to admire, listen to, and support Richard Stallman; but he (and FSF) is not a neutral source on these things, which he openly states in that article. I suggest that further discussion of this point be on the other article's talk page, to avoid fragmented discussion. Murph9000 (talk) 20:47, 1 June 2017 (UTC)[reply]

I'm planning to fill in missing blanks in the table, but I would like to propose some clarifications:

"?" should be used for incomplete fields (meaning they haven't been researched by a human yet)

"unknown" should be used for unknown values (for example, if the creator of the license is not attributed in any known source, it is unknown, but still complete and researched)

"disputed" should be used when a field's value is disputed. Disputed fields should always include sources.

Some additional proposed changes: Modification column should be more specific rather than yes/no, since many licenses allow modification, but only if modifications are labelled/documented.

In addition to filling in "?", I will add some licenses that are not included (for example, some Microsoft licenses) and breakdown some licenses further (for example, BSD license is commonly separated into 2-Clause, 3-Clause, and 2-Clause-Patent). I will also add clarifying notes when helpful. Unknowngnome (talk) 03:44, 13 June 2018 (UTC)[reply]

• Those are good clarifications, I think. Thank you. -- econterms (talk) 17:44, 29 June 2018 (UTC)[reply]

the Lesser Perceptive Licence implies proprietary way to the available result, indicating that all knowledge about the software is perceptible and accessible to the use of other tools to achieve the presented result;

perhaps thus indicating the tools used, "the work was evidenced by the developer in order to complete the available project"; reserving the perception in the interests of the end user in perpetuating actions to infer the productivity triggering efforts in losses for social participation;

The title under reverse engineering implies the deduction in which the work was already personified by the developer, again indicating the tools of third parties used for production is a result of the intellectual project already personified through the available work;

By reserving the perception of continuity of the project absent from interruptions, focusing on the embellishment, improvement in updates and continuity by the author of the work presented - indicating the fame being reserved in access to other tools to satisfy curiosities of the end user to his own learning for production ; — Preceding unsigned comment added by WDS Account (talk • contribs) 08:33, 28 August 2018 (UTC)[reply]

While a helpful summary, the values used in the table are confusing and easy to misinterpret. For instance, under Private use we have "Yes" and "Permissive" -- do they both mean the same thing? While the first use of "Permissive" and "Copylefted" in the table are linked to their respective articles, this is easy to miss and still leaves other values poorly defined. Perhaps a brief "legend" for the table would be helpful, defining what each value means.

Also, the "Private use" column may be a bit misleading since, for example, GPL code can be modified and compiled for private use, but must be distributed once the compiled form is released to any other party. Perhaps the combination of "Distribution" and "Private use" adequately delineates this, but this does not seem immediately clear. Regards, Daniel Santos (talk) 06:59, 20 May 2020 (UTC)[reply]

Yes, the inconsistencies are quite jarring. The 'private use' column is particularly problematic because it conflates a number of related concepts that come under an umbrella term more like 'personal use'. This covers scenarios like 'free use', (quotations and for news and instructional purposes all under the Berne convention), fair use, first sale, work for hire and some proprietary uses like product evaluations, testing, trials and secretive development which although might often occur simultaneously carry useful distinctions. My preference would be to strike that column out? User Tammy (talk) 14:18, 12 August 2021 (UTC)[reply]

I agree with you. I've added a paragraph underneath explaining these terms.Gqqnb (talk) 14:31, 13 November 2021 (UTC)[reply]

I have a few questions about the Python Software Foundation License. Obligatory disclaimer: I am not a lawyer, just trying to bumble through...

• It refers to version 3.9.1 but tracking through the references I cannot find such a version, certainly not easily. As far as I can tell from web searches, PSF license version 2.0 seems current and is the only one for which there is a registered SPDX identifier. So I'm wondering if SPDX needs updating or if this table should be better annotated or if the 3.9.1 is just a typo. I don't much care, but I find it confusing. I like that in other rows like MIT, the values are footnoted so that I can go to the relevant source document to understand the meaning.
• In the General Comparison table, it shows a "Yes" in the "Patent Grant" column, but the PSF 2.0 license (PSF-2.0) doesn't even mention patents. If I understand this table (and really I find the patent grant column confusing), maybe "Manually" is the right value to indicate that if you want to add this, you have to do it manually? Is that what "manually" means? I would think a designation like "Not Discussed" or even "Permissive" was better here, but perhaps if someone thinks not they can explain the meanings of these terms in a way that clarifies that.
• The TM field is also confusing. The title is "TM grant" but surely none of these things are granting trademarks. What they are doing is either explicitly saying or not explicitly saying what is probably true in either case, that the fact of the license does not confer a grant of trademark. So it would be surprising if there were a legitimate "yes" in that column. I see the various GPL items have Yes in this column but if I go to the GPL 3.0 license, for example, all I see is "Notwithstanding any other provision of this License, for material you add to a covered work, you may (if authorized by the copyright holders of that material) supplement the terms of this License with terms: ... (e) Declining to grant rights under trademark law for use of some trade names, trademarks, or service marks; or ...". That is not a trademark grant. It's not really anything about trademarks _at all_, other than the indirect permission to make your own trademark rules if you want to, which I'd think is just a redundant statement of what would be true regardless. (If all freedoms have to be expressly enumerated to be true, that's not very free.) But in the context of how things are now, all the PSF does is fail to grant trademark rights (a thing that really all of the licenses fail to do) and remind you as a consequence that you are not free to just use the mark as an endorsement (which is presumably the core of what holding a mark is for in the first place, so I'm guessing doesn't need to be said).
• I notice that there exist licenses where people use a license like this (which is highly specific and not intentionally repurposable) by changing the names in the license. An example is the license for the Python bitarray, which offers itself as the PSF license when in fact it's a modified version of the license changing the names. Is that still customary to call the PSF or is there a modifier one uses like "Adapted PSF" or "Modified PSF" or "Personalized PSF" that both indicates an edit and narrows the nature of the edit to merely a different name? The terms are the same, just the players are different. (It feels like there is an analogy here between internationalization/localization and licenses, where some licenses have been made generic and ready to be made specific, and other licenses have not.) This last bullet point is perhaps wandering from the main topic a bit, but a useful bit of data to know is whether the license can/should be used for any other purpose than its original purpose given its wording is specific to specific individuals/organizations. (OK, the Matching Guidelines document for SPDX seems to answer this question, at least. -- Netsettler (talk) 13:52, 25 May 2021 (UTC))[reply]

-- Netsettler (talk) 19:25, 9 May 2021 (UTC)[reply]

I second this - what does "manually" mean? It should be explained specifically what is meant by that. 128.244.38.5 (talk) 18:55, 20 April 2022 (UTC)[reply]

Is there any reason why this article uses the British spelling "licence" in the title? As of now, the spelling "licence" occurs 42 times in the article and "license" occurs 300 times. Wikipedia does not promote U.S. spelling over U.K. spelling, but it does ask for consistency within an article: WP:COFAQ#ENGLISH. Would anyone have an objection to moving this article to "Comparison of free and open-source software licenses" and changing the 42 variants to be consistent with the 300? Other license-related articles use "license", as seen in Free software license and Open-source license. Iritscen (talk) 16:02, 21 November 2021 (UTC)[reply]

See: https://en.wikipedia.org/wiki/Talk:Comparison_of_open_licences#Merge_propsal — Preceding unsigned comment added by Avoinlähde (talk • contribs) 22:55, 6 March 2022 (UTC)[reply]

Price - someone needs to spell out which allow others to profit from our work, and which prohibit that. 110.143.72.26 (talk) 10:27, 20 August 2022 (UTC)[reply]

The table lists the latest version of GNU Affero General Public License as v2 when it should be v3.

Edit: It seems the table includes GNU Affero General Public License twice due to a change in Author. This seems rather confusing but may be technically correct. ISpyCreativity (talk) 06:50, 12 April 2024 (UTC)[reply]