Stack Exchange
log in

MadHatter

 The Bikeshed

General discussion for opensource.stackexchange.com. Confused ...
MadHatter
Nov 15, 2024 16:54
@PhilipKendall I feel slightly proud of myself for getting that! Clever choice of line on your part, anyway.
 

 Discussion on answer by MadHatter: Ca…

Imported from a comment discussion on opensource.stackexchange...
MadHatter
Nov 5, 2024 12:13
@Darkwing as I said to KalleMP, the matter is capable of argument, and if you want to argue it, do by all means ask a question about it. But please do not try to argue it in the comments field of someone else's answer.
MadHatter
Nov 5, 2024 12:13
@Trilarion did you follow the links down? The specific example they quote is "people living under repressive regimes who want to keep their use of the software entirely secret".
MadHatter
Nov 5, 2024 12:13
@KalleMP then by all means ask a question about it.
MadHatter
Nov 5, 2024 12:13
This would be the age-old philosophical question of whether true freedom includes the freedom to deprive others of their freedoms. Most don't recognise that as a legitimate exercise of freedom, and so free software's refusal to permit its users to deprive other users of those freedoms tends to be seen as legitimate. But I accept that it is a point capable of argument.
MadHatter
Nov 5, 2024 12:13
Ah. This would be the age-old philosophical question of whether a free society allows people the freedom to deprive others of their freedoms. Most don't recognise that as a legitimate exercise of freedom, and so free software's refusal to permit its users to deprive other users of those freedoms tends to be seen as legitimate. But I accept that it is a point capable of argument.
MadHatter
Nov 5, 2024 12:13
@DetlevCM some indeed define it that way, but "accessible" isn't a well-defined term, and "available" even less so. I'd be really surprised if you could produce something classified as "free software" by anyone other than its creator that is closed source - and it certainly wouldn't satisfy the four freedoms if you did. Don't forget that free software and zero-cost software aren't the same thing.
 

 Discussion on question by stevec: Wha…

Imported from a comment discussion on travel.stackexchange.com...
MadHatter
Oct 9, 2024 15:20
I've never been asked for a physical card on check-in. They want a credit card against which they can put the hotel's standard hold, however much that is, but they've never cared whether it was physical. What I have noticed is that they really don't like debit cards, whether physical or logical.
 

 Discussion on answer by Simon Crase: …

Imported from a comment discussion on opensource.stackexchange...
MadHatter
Sep 27, 2024 11:36
@Brandin they're the reasonably free-software-standard disclaimer, indemnity agreement, and liability limitation sections, respectively.
MadHatter
Sep 26, 2024 13:56
Furthermore, the language of ss6-8 is common, well-understood language, the likes of which has withstood challenge before. That of s5f is a lot less well-understood. If any section is to be looked at askance because of possible interpretational issues, it is surely s5f and not ss6-8. s5f adds nothing to the developer's protection, to my mind.
MadHatter
Sep 26, 2024 13:54
As I understand it, Simon advocates that s5f affords some protection over and above ss6-8 because "those only mean what a court says they do". My objection is that that weakness applies equally to s5f. If the license means what it says on its face, then the protections of ss6-8 apply in all circumstances, and cannot therefore be further extended by any other section. If it doesn't, then no other section can be held to offer any particular protection either.
MadHatter
Sep 26, 2024 09:28
@SimonCrase it would be more helpful if instead of saying what argument you are not making, you would say what argument you are. The licence currently contains two disclaimers: one of which is valid in all circumstances, and one of which is only valid in some. Why do you think the second is more protective of the developer than the first?
MadHatter
Sep 26, 2024 09:28
@SimonCrase well, you're definitely arguing that the standard disclaimer may not be valid because it hasn't been adjudicated in this case ("The licence means what a court says it means, not what the author thinks it means") and you're also arguing that the illegal-and-immoral clause protects the developer ("I have a defence: they were outside the terms of their licence"). So it seems to me that you are very much making a logically-indefensible argument, and no particularly perverse reading is required.
MadHatter
Sep 26, 2024 09:28
I don't think you can reasonably argue that a very standard form of licence disclaimer is not valid because it hasn't been adjudicated in this particular (hypothetical!) case, whilst also arguing that a very odd licence disclaimer must implicitly have validity.
 

 Discussion on question by Ethan: Does…

Imported from a comment discussion on law.stackexchange.com/qu...
MadHatter
Jun 14, 2024 06:55
As far as I can tell, the idea behind having a separate offence was to start off with a very clear statement of law that deaths on the roads were not automatically "just one of those things", or "bad luck", but that responsibility, and indeed culpability, might be involved.
MadHatter
Jun 14, 2024 06:54
@MichaelKay yes, see my answer. I'm not sure that it's easier to prove, so much as that there was a fairly-well-documented reluctance on the parts of juries to convict for manslaughter. One might think it was because many of them were drivers themselves, and didn't much fancy being held to that standard, but that would be pure speculation!
 

 Discussion on question by Silkly: Acc…

Imported from a comment discussion on opensource.stackexchange...
MadHatter
May 3, 2024 15:43
So I think you accept that it's perfectly lawful to release binary-only software under eg the MIT licence, but you're asking why permissive free licences are even allowed? Why don't we all prefer strong copyleft licences? Or are you asking how someone who releases binary-only under MIT can refer to their software as "open source"? I can tell you're upset about this (and I don't blame you!), but I'm less clear about what your question is.
MadHatter
May 3, 2024 15:43
Could you take a look at this question and its answer, and let us know whether or not that answers what you wanted to know? If not, could you clarify what in your original question remains unanswered?
 

 Discussion on answer by MadHatter: De…

Imported from a comment discussion on opensource.stackexchange...
MadHatter
Nov 15, 2023 22:07
I know it's concatenated; you told us so yourself, though you took three days to resolve the ambiguity. I didn't say it wasn't. But since you don't seem to follow my argument in bulk, I'm taking it one step at a time; therefore, I repeat the question. Removing the name is of no help in regard to my original comment, though removing the nonce would be.
MadHatter
Nov 15, 2023 22:07
I do not accuse you of designing your own hash function; what worries me is that you're rolling your own crypto scheme with it. The nonce is my concern, but you misunderstand me, so let me take this step by step. Would you acknowledge that with an XORed nonce, after you unblind yourself, anyone else can trivially come up with another name and nonce pair that combine to the same input bit pattern to sha256, and that this would provide a difficult-to-rebut attack on your copyright registration plan?
MadHatter
Nov 15, 2023 22:07
I entirely agree. Problem is, you've picked a scheme where a pseudonym is mapped from the combination of your real name and a nonce that you choose, and I'm trying to point out that the addition of the nonce weakens the strategy fatally in certain circumstances - enough that I'd think twice about using it the way you intend to, unless you're a genuine cryptanalyst who can say with some certainty that's it's as secure as you think.
MadHatter
Nov 15, 2023 22:07
@user3368561 agreed, but it suffers from the concomitant problem that it possibly can't be disclosed without weakening it. Don't focus too much on the security of the hash algorithm (which I do not question) when evaluating the security of your cryptosystem built around it. Long-term security at the expense of short-term security isn't a good trade-off.
MadHatter
Nov 15, 2023 22:07
Sorry, are you saying that the less effective scheme should be preferred because it's simpler? And I'd like to know a bit more about your cryptanalytical credentials before accepting your word that concatenation doesn't suffer from similar problems (and note that we don't yet know the OP meant concatenation, see ambiguity documented above).
MadHatter
Nov 15, 2023 22:07
It's easiest to show in the case of an XOR: if the string to be hashed is the XOR of a name and an arbitrary nonce, then constructing another nonce that combines with a chosen alternate name to give the same input to the hash function is trivial. As I say, I don't know for sure that it's similarly flawed if name and nonce are combined by concatenation, but I am fairly sure they are for both OR and NAND. Why take the chance, when a known-robust scheme is available?
MadHatter
Nov 15, 2023 22:07
It's also used for NAND, and, as you say, a number of other things, though Wikipedia suggests that the double vertical bar is used for concatenation. Nevertheless, the OP's strategy still relies on providing the input to the hash function in order to prove ownership, and I'm reluctant to say for sure that that's no help to an attacker who can choose alternate inputs. Why take the chance, when there are other strategies that are known to be resistant to such compromise?
 

 Discussion between user334639 and Mad…

Imported from a comment discussion on opensource.stackexchange...
MadHatter
Oct 5, 2023 16:04
@user39559 "I'm finding it very poorly redacted" I do not think that word means what you think it means (with apologies to The Princess Bride).
MadHatter
Oct 5, 2023 07:42
@Brandin that seems to me like a very good framers-intent reading. I'm fairly sure it's what the creators of the license had in mind, even though I don't think it's what they wrote.
MadHatter
Oct 4, 2023 20:11
@Brandin fair point, but then what the hell is "the licensed material in modified form" that is different from "adapted material"? This just reinforces my opinion that the CC licences are not well-crafted, certainly compared to the GPL, and it makes me happy that they're off-topic here.
MadHatter
Oct 4, 2023 06:01
I don't think we can resolve this. You are still insisting that 3(a)1(c) is only an attribution requirement, because it's in a section titled "Attribution" even though on the plain face of it it's a licensing requirement, as at least two other users have confirmed. I'm sorry you don't like my answer, but I don't think you're going to get a different one, and certainly not one that accords with your reading.
MadHatter
Oct 3, 2023 20:18
(I'll leave aside that nearly no licence requires you to release code, but merely specifies the conditions under which it must be released if you choose to release it.)
MadHatter
Oct 3, 2023 20:18
You say "The license does not require me to release the modified version in any specific license". What then do you think it is that the licence does require to be licensed CC BY-NC? You've ruled out the original work, since you say that none such is conveyed when you convey a modified work.
MadHatter
Oct 3, 2023 20:10
The licence says that if you share the licensed material, in either modified or unmodified form, you must indicate that it is licensed CC BY-NC, which I maintain means that it must be so licensed; I stress, in either modified or unmodified form.
MadHatter
Oct 3, 2023 17:35
Then what do you think the "original material" that's being distributed and therefore, in your view, has to be licensed under BY-NC, is? Your previous comment makes it sound as if you think there is no original material and thus 3(a)1(c) refers to the empty set whenever non-verbatim redistribution occurs.
MadHatter
Oct 3, 2023 17:35
Then I think you've answered your own question ("someone could make a tiny change to the work and release under whatever license they please"). Under your interpretation, it's only their tiny change they can release under "whatever license they please". If on the other hand you take the view that the whole derived work is the licensed material, then it must all still be CC BY-NC. Neither of us thinks they can make a tiny change and release the entire derived work under a new licence, which was what you asked about the possibility of.
MadHatter
Oct 3, 2023 17:35
We're going to have to agree to differ on this one. I can't see any other way to read "indicate ... the material is licensed under this license" without also actually licensing it under the specified licence. If you can, good for you, but I think you're in the minority.
MadHatter
Oct 3, 2023 17:35
Brandin says it more eloquently than I could've! I'm afraid licences aren't like newspapers; you can't just read the section headings and think you've got the gist of it, you've got to read every single word carefully. Even if s3a were titled "Choosing a dancing partner", 3(a)1(c) would still require that CC BY-NC terms apply on distribution of the material.
 

 Discussion on question by Tooster: Li…

Imported from a comment discussion on opensource.stackexchange...
MadHatter
Sep 7, 2023 10:40
@Brandin do you want to write that up as an answer?
MadHatter
Sep 7, 2023 09:59
Otherwise I'm going to have to do a giant pruning session later.
MadHatter
Sep 7, 2023 09:59
This is a fairly active discussion, so I hope everyone's OK with me moving it to chat.
MadHatter
Sep 7, 2023 09:58
Only if you are either the sole rightsholder, or you have a CLA in place with all other rightsholders. Are either of these the case?
MadHatter
Sep 7, 2023 09:58
That is true. Then I think you want dual-licensing (LGPL to all, proprietary non-copyleft with mandatory credit clause to paying customers) which will require a CLA from all contributors. How does that sound?
MadHatter
Sep 7, 2023 09:58
What's wrong with a permissive licence, like MIT or BSD3?
 

 Discussion on answer by Johns-305: Ho…

Imported from a comment discussion on travel.stackexchange.com...
MadHatter
Aug 5, 2023 06:22
It also slightly seems to be overlooking the principal element of a Boeing 787's carbon footprint, which is the fuel it burns over those fifty million miles.
 

 Discussion on answer by Thomas Owens:…

Imported from a comment discussion on opensource.stackexchange...
MadHatter
Jul 25, 2023 10:05
@ThomasOwens I completely agree that the novel content in the collection is copyrightable. I don't think I've ever contended that point, so I'm not sure why Feist is coming up, which speaks only to the copyrightability of collections. However, I think we're starting to go around in circles in these comments. You believe a collector of third-party CBS3 works has the right through CBS3 to licence his/her entire collection under terms (s)he chooses, and you've written an answer saying so. I don't, and have done likewise. Time will tell whose point of view commands more community support.
MadHatter
Jul 25, 2023 10:05
@ThomasOwens that cannot be the correct interpretation, since if it were, you could then adapt the works out of the CBS4 collection, and have them available under CBS4. If it were the intent of the authors of the licence that you could do that, they wouldn't have bothered writing s4a as they have, but would simply have said that CBS3 works can be relicenced under CBS4 on demand. Since they instead wrote a lengthy s4a showing how simple copies of the work must remain at CBS3, your interpretation cannot be correct. And as I keep saying and you keep ignoring, the licence says not on its face.
MadHatter
Jul 25, 2023 10:04
@ThomasOwens then you seem to be ignoring the bit of s4a that says when the work is incorporated into a collection, you can release the collection under any licence you choose apart from the work itself. That is, the filigree bits can be under CBS4, or GFDL, or BSD, or CC0, or anything you like. But the works in the collection stay under CBS3. It's really very clear about that.
MadHatter
Jul 25, 2023 10:04
If there's nothing copyrightable to the collection except the works, it seems we agree those have to be under CBS3. So assume there's enough novel expression to it that it is copyrightable. I repeat my question: by "the collection" do you just mean that novel expression, or that plus the original works?
MadHatter
Jul 25, 2023 10:04
What do you mean by "the Collection" in the above comment? The entire collection - works, rubric, and all - or just the filigree around the edge that binds the work(s) into a collection but does not include the works inside that collection?
MadHatter
Jul 25, 2023 10:04
You write "You can release the collection under any license that you choose". No; see CC-BY-SA-3.0 s4a: you may release the collection under any licence you choose apart from the Work itself.
 

 Discussion on question by Programmer1…

Imported from a comment discussion on opensource.stackexchange...
MadHatter
May 25, 2023 15:24
I am really depressed - but not very surprised - to hear that a modern course about IP and licensing didn't cover the four freedoms, and the whole class of licences that give them. Hopefully we've been able to open your eyes a bit!