You're reading the public-facing archive of the Category Theory Zulip server.
To join the server you need an invite. Anybody can get an invite by contacting Matteo Capucci at name dot surname at gmail dot com.
For all things related to this archive refer to the same person.
Nick Hu (Aug 31 2021 at 10:39):Hi, it's come to the time where I need to upload my paper to arXiv. I would like to choose the least restrictive license possible (i.e. CC BY), because I believe it makes my work more accessible.
Kenji Maillard (Aug 31 2021 at 10:51):On the second point, the CFP for MFPS 2021 states that
Proceedings
A preliminary version will be distributed at the meeting. Final proceedings will be published in ENTCS after the meeting. ENTCS is open access.
and you can look up informations on the sharing policy if ENTCS on sherpa romeo
John Baez (Aug 31 2021 at 11:04):Certainly there's evidence that putting one's papers on the arXiv increases the number of times people cite them. I haven't heard any evidence that the license you choose matters much. I imagine most individuals downloading papers from the arXiv don't pay attention to the license.
Tom de Jong (Aug 31 2021 at 11:10):Kenji Maillard said:
On the second point, the CFP for MFPS 2021 states that
Proceedings A preliminary version will be distributed at the meeting. Final proceedings will be published in ENTCS after the meeting. ENTCS is open access.and you can look up informations on the sharing policy if ENTCS on sherpa romeo
This is outdated, I think. The website says:
Elsevier ceased publication of ENTCS at the end of 2020. ENTCS was the traditional publication venue for MFPS. Negotiations are underway to establish a new arrangement for MFPS proceedings publication going forward. The new venue will be announced as soon as arrangements are complete. We anticipate the formal proceedings of this year’s conference will be published after the meeting, as has been the tradition for MFPS for many years. A preliminary proceedings will be distributed to participants at the meeting.
and I have an email dated 16 August from the MFPS PC chair saying:
Michael Mislove, the chair of the MFPS Steering Committee is working on a solution for replacement of ENTCS. This will decide the publication venue for the (post-) proceedings. [...] (I expect that this will only be after the conference.)
Nick Hu (Aug 31 2021 at 11:16):Yes, ENTCS has been discontinued, as @Tom de Jong says
Nick Hu (Aug 31 2021 at 11:18):By the way, how am I supposed to interpret the 'License' field on sherpa romeo?
Tom de Jong (Aug 31 2021 at 11:21):Maybe the safest thing to do is to wait for the post-proceedings details and instead of hosting your paper on arXiv now, choose to host it on your website only for the time being?
John Baez (Aug 31 2021 at 14:51):Any journal or conference that doesn't let you put your paper on the arXiv should be avoided at all costs: the arXiv is where most people actually read papers these days. You can use the basic arXiv license if you're worried about tricky issues.
Rune Haugseng (Aug 31 2021 at 21:34):Conversely, I've never understood why I would want to choose a CC license on arXiv - I would like people to be able to freely download and read my papers, but what is the argument for allowing them to create and distribute modified versions of them or (admittedly a very hypothetical scenario for math papers) sell them for profit?
John Baez (Aug 31 2021 at 21:45):There are many different kinds of CC license available on the arXiv. It sounds like you're talking about CC-BY or CC-BY-SA. If you don't want commercial use, you could use CC-BY-NC-SA. If you also don't want people to change your stuff in any way, you could use CC BY-NC-ND.
As to your question: I wouldn't mind if someone created a hyperlinked version of an old non-hyperlinked paper of mine, or made other improvements. But I wouldn't want anyone mucking with the "essence" of my papers, and unfortunately it's hard to define that. I probably wouldn't mind someone selling a volume containing my old papers, since all I really care about is maximizing exposure to my ideas. But I haven't taken advantage of all these different licenses - too lazy.
Nick Hu (Sep 01 2021 at 12:51):I've heard that Quantum and Compositionality require some form of CC, so that they can mirror the articles on their website, which sounds reasonable-ish. But part of why I'm asking is because I have the same queries @Rune Haugseng
Nick Hu (Sep 01 2021 at 12:52):The issue is that the arXiv clearly states that the license cannot be changed after initial submission, so I really need to decide up-front which license I want to use
Nick Hu (Sep 01 2021 at 12:53):I am sure the arXiv non-exclusive perpetual license would be sufficient, but I would like to put in the effort to do better (under my - possibly unfounded - belief that less restrictive license = better)
John Baez (Sep 01 2021 at 12:56):You should read some details of the different kinds of license available on the arXiv and see which one you like. I just checked, and Compositionality claims you need to use a CC BY 4.0 license, which allows anyone to
Share — copy and redistribute the material in any medium or format
Adapt — remix, transform, and build upon the material for any purpose, even commercially.
John Baez (Sep 01 2021 at 13:00):I don't remember getting a CC BY 4.0 license on the archive for the paper I submitted to Compositionality.
Nick Hu (Sep 01 2021 at 13:00):Yes, I would like to use CC-BY because it is the least restrictive out of all the licenses offered by arXiv. However, this doesn't tell me how feasible it is to use this license in general, for conference papers, journal papers etc.
I'm also a bit confused by this requirement for Compositionality, because surely choosing the arXiv non-exclusive perpetual license for your arXiv submission doesn't preclude you from dual licensing your manuscript as CC-BY - why should they care about the arXiv version specifically?
John Baez (Sep 01 2021 at 13:02):I'm a bit confused too. But I don't think they require you to have gotten a CC BY license on the arXiv before you submit the paper. After all, people often put their paper on the arXiv before they know where they are going to submit it, and people often submit papers and don't get them accepted.
John Baez (Sep 01 2021 at 13:03):Also, I don't remember having gotten a CC BY license on the arXiv for the paper I published in Compositionality.
Nick Hu (Sep 01 2021 at 13:03):Yes, probably because arXiv doesn't allow you to change the license post-submission
John Baez (Sep 01 2021 at 13:04):Right. The Compositionality website just says
Authors must also sign our consent to publish agreement, which is based on the CC BY 4.0 licence.
Nick Hu (Sep 01 2021 at 13:04):So your paper now exists in a state of dual licensing where arXiv may use it under very restrictive terms, but additionally anyone else may use it (at least the version published by Compositionality) under CC-BY 4.0
John Baez (Sep 01 2021 at 13:05):I guess so. Luckily I don't care so I don't need to figure this out. :upside_down:
Nick Hu (Sep 01 2021 at 13:06):But this kind of thing affects tools you can build on the arXiv infrastructure, so it's not entirely pointless
Joe Moeller (Sep 01 2021 at 13:07):Compositionality is an arxiv overlay journal, right? This must be why they care about your arxiv license.
Nick Hu (Sep 01 2021 at 13:08):But surely this goes to show that actually you should care about which license you choose when you publish on arXiv?
Joe Moeller (Sep 01 2021 at 13:09):IIRC they required that we give them power to change the arxiv version of our paper, and they uploaded the journal version there themselves.
Joe Moeller (Sep 01 2021 at 13:10):You should care about the license. In my case, I cared the first time I went to upload a paper, realized the default was fine, and then haven't thought about it since.
Nick Hu (Sep 01 2021 at 13:10):Also the arXiv license doesn't cover the case where arXiv goes down or runs out of funding (from my understanding)
Nick Hu (Sep 01 2021 at 13:11):It's hard to imagine we won't have it in 10 years, but what about 100 years, what about 1000 years?
Joe Moeller (Sep 01 2021 at 13:12):Do people today care about the license on papers published in the 1920s? Genuine question.
Nick Hu (Sep 01 2021 at 13:12):Well copyright runs out, but it makes it harder for people to archive papers, and papers from that era are certainly more difficult to find and collate together on one website like the arXiv
Nick Hu (Sep 01 2021 at 13:14):Nick Hu said:
But this kind of thing affects tools you can build on the arXiv infrastructure, so it's not entirely pointless
This relates to what I mean here. I think there are tools which exist which allow you to locally mirror the arXiv offline (say, imagine you were going to live somewhere without internet access for a year), but arXiv licensing makes it unlawful to download papers which aren't under some sort of CC license - so the tool won't download those papers. I don't think it's a hugely contrived scenario
Nick Hu (Sep 01 2021 at 13:15):I mean, there's a reason why arXiv tells you that your paper metadata (abstract, authors, etc.) are licensed under something like CC (iirc) - regardless of how you license the actual paper
John Baez (Sep 01 2021 at 13:28):Nick Hu said:
But this kind of thing affects tools you can build on the arXiv infrastructure, so it's not entirely pointless
I wasn't saying the whole issue is pointless. I just meant: maybe my paper on Compositionality has a perpetual nonexclusive arXiv license and a CC BY 4.0 license, and if I were very legalistic I might wonder how these interact, but I don't actually care too much.
Nick Hu (Sep 01 2021 at 13:30):But I'm saying that as a pragmatic, non-legalistic, person I think it still makes a tangible difference
John Baez (Sep 01 2021 at 13:31):Joe Moeller said:
Compositionality is an arxiv overlay journal, right? This must be why they care about your arxiv license.
I'm pretty sure they don't make you do something to your arXiv license when you submit a paper there - Nick thinks it's impossible to retroactively change your arXiv license. And I'm pretty sure they don't make you have already chosen a CC BY 4.0 license on the arXiv, before you submit the paper.
I think they just slap the CC BY 4.0 license on top of whatever else is there.
Nick Hu (Sep 01 2021 at 13:32):I guess ideally we would just be able to 'upgrade' the arXiv license. Maybe someone here knows how to petition for that
John Baez (Sep 01 2021 at 13:32):Nick Hu said:
But I'm saying that as a pragmatic, non-legalistic, person I think it still makes a tangible difference
What difference does it make to you? I'm talking about the arXiv-Compositionality interaction, not the general issue of choosing a license. THAT'S what I don't care about.
Nick Hu (Sep 01 2021 at 13:32):Impossible in the sense that there is no mechanism on arXiv, not impossible legally
Nick Hu (Sep 01 2021 at 13:34):John Baez said:
Nick Hu said:
But I'm saying that as a pragmatic, non-legalistic, person I think it still makes a tangible difference
What difference does it make to you?
I can imagine scenarios where people have limited access to the internet and cannot access arXiv, but maybe some kind of mirror which is only able to mirror arXiv articles which are at least some kind of CC license
Nick Hu (Sep 01 2021 at 13:34):Oh yes I agree the interaction is non-interesting
John Baez (Sep 01 2021 at 13:34):Okay, good.
Nick Hu (Sep 01 2021 at 13:34):But I do think it's important that arXiv reflects the most permissive license of your work, because it's the go-to when people want to look up a paper
Nick Hu (Sep 01 2021 at 13:35):And the issue is that when you upload a paper to arXiv, there are many valid reasons for why you initially choose a more restrictive license, but arXiv locks in that choice (in perpetuity)
John Baez (Sep 01 2021 at 13:36):I am lazy and I've never gotten around to deciding if I want to let people adapt my work, sell it, etc., so I choose the minimal arXiv license - the default option.
John Baez (Sep 01 2021 at 13:36):This is not really more restrictive, it's just giving the arXiv the right to have the paper and not saying anything else.
Nick Hu (Sep 01 2021 at 13:37):It is restrictive from the perspective of e.g. wanting to mirror the arXiv
Nick Hu (Sep 01 2021 at 13:38):It's already annoying when I want to use some software that isn't under some free license, not because of some philosophical reason, but because it means it's likely not going to be packaged up and easy to install (in my distribution)
Nick Hu (Sep 01 2021 at 13:39):If it were really a non-issue then they wouldn't have added those license options in the first place
John Baez (Sep 01 2021 at 13:39):I never said it's a non-issue.
John Baez (Sep 01 2021 at 13:41):If you want someone to be able to put your paper on some other website, or sell it, or various other things, you may want to choose some license that makes that possible. (However, I haven't found out where you can tell which license an arXiv paper has!)
Nick Hu (Sep 01 2021 at 13:43):Compare https://arxiv.org/abs/2108.13745 with https://arxiv.org/abs/2108.13600 (just the two most recent papers on math.CT), and look at the part of the page where it says "Download". That will tell you the license.
Nick Hu (Sep 01 2021 at 13:43):You likely never noticed because the vast majority of papers on arXiv are published under their perpetual non-exclusive license
John Baez (Sep 01 2021 at 13:44):Right. My papers have the tiny word "license" in that spot, and it's almost invisible.
Nick Hu (Sep 01 2021 at 13:45):Yeah, that indicates arXiv perpetual non-exclusive licensing
John Baez (Sep 01 2021 at 13:46):One reason I lazily choose the perpetual non-exclusive license is that if I choose something like CC BY, what happens if I submit my paper to an ordinary journal that wants a lot of rights to my paper? Does this mean I can't legally submit my paper there? I've never figured this out; I'm too lazy.
Nick Hu (Sep 01 2021 at 13:49):I think yes, that is what it means
John Baez (Sep 01 2021 at 14:19):Okay, so I'm right to avoid choosing CC BY or certain other licenses, because usually when I submit a paper to the arXiv I haven't decided what journal I'm going to submit to, and while I'm strongly in favor of open access I won't refuse to submit to any journal that wants some power over my papers.
John Baez (Sep 01 2021 at 14:19):(Now that I'm retired I could start doing that.)
Nick Hu (Sep 01 2021 at 14:25):Yes, I think the best solution would be if arXiv let you initially submit under their restrictive license, and you could 'upgrade' it later after publication. It would be great if we could petition them to allow for this, but I wouldn't know where to start with that.
I think it's possible that submitting your article to arXiv (even with the restrictive license) can violate the rules of some journals which don't allow you to also have your paper on arXiv. I don't know of any, but I can believe they exist. But I guess you just would avoid those journals.
Nick Hu (Sep 01 2021 at 14:26):Nick Hu said:
I think yes, that is what it means
It also might be possible that in such a situation, you can still submit to those journals but with a caveat that you have to withdraw your paper from the arXiv. I'm not sure
John Baez (Sep 01 2021 at 14:27):There are journals that don't let you put your paper on the arXiv. Some of the most famous journals, like Nature, are of this sort.
Nick Hu (Sep 01 2021 at 14:27):Right, some also have embargoes and things like that
John Baez (Sep 01 2021 at 14:28):There used to be many more, and there probably still are in some fields, but in the fields of math and physics we managed to beat them into submission.
So, it's easy for me now to avoid publishing in journals that don't let me keep my paper on the arXiv. In the old days I'd sometimes rewrite the contract provided by the journal.
Nick Hu (Sep 01 2021 at 14:28):I still hear people, in CS at least, discuss the 'issue of open access', so from my perspective it hasn't been entirely resolved
John Baez (Sep 01 2021 at 14:29):Right, and in subjects like chemistry or biology the situation is far worse, though the bioRxiv is catching on.
(For some reason those idiots captitalize the R. I guess they don't write in TeX.)
Nick Hu (Sep 01 2021 at 14:31):Nick Hu said:
I still hear people, in CS at least, discuss the 'issue of open access', so from my perspective it hasn't been entirely resolved
(Maybe someone who is more senior could give me some insight on this? I've never heard of a modern CS publication that isn't freely available on the arXiv so I wonder what they mean by 'open access')
Nick Hu (Sep 01 2021 at 14:31):Probably because it looks more symmetric
John van de Wetering (Sep 01 2021 at 21:29):RE changing license for papers on arxiv: When I got stuff accepted in Quantum I had to reupload a new version of the paper, and when you do that you can decide to change the license on the new version. So I think in that case your old version exists under a different license than the new version. Which makes sense as a possibility. For open-source software the creators can also decide to make their license more free for a newer version (this is what Aleks and I did for PyZX).
David Michael Roberts (Sep 02 2021 at 02:25):The least restrictive license on the arXiv s CC0, not CC-BY.
If you want someone to be able to translate your article into a different language without pestering you or your descendants, or your publisher, then a CC license allowing for modifications is the thing. Who doesn't want their paper accessible to more people ?
Nick Hu (Sep 02 2021 at 08:49):Ah, I hadn't clocked that translation is a valuable reason to use a CC license. Nice
I think CC0 is probably a bit extreme though, and I can imagine (although I don't know) that publishers would take issue with it. Also, putting stuff into the public domain is way harder than it sounds, and I think it's technically impossible in some countries, which creates a whole host of issues
David Michael Roberts (Sep 03 2021 at 05:37):That's what CC0 was designed for: an explicit license that is functionally PD specifically for places where it's not like in the US.
I once got Springer to write me a custom publishing agreement for a paper that was CC0 on the arXiv, as I couldn't sign over copyright :-) Not sure I'd recommend it, but it was an interesting experience.
Dmitri Pavlov (Oct 08 2021 at 16:18):David Michael Roberts said:
That's what CC0 was designed for: an explicit license that is functionally PD specifically for places where it's not like in the US.
I once got Springer to write me a custom publishing agreement for a paper that was CC0 on the arXiv, as I couldn't sign over copyright :-) Not sure I'd recommend it, but it was an interesting experience.
Does Springer force you to sign copyright transfer forms in general? I know that Elsevier does not: if you simply do not respond to the requests, they will publish your article anyway, without signing any forms.
Kevin Buzzard (Oct 10 2021 at 11:07):In the late 1990s I once didn't sign an Elsevier copyright form (I wrote them an email saying which parts I was not happy with) and I got a grumpy email back from a publisher there saying "if you propose to super-distribute your paper on the internet then why are you even bothering to publish it with Elsevier?" :-) I withdrew the paper the same day, and sent it to JAMS, who accepted it :D
David Michael Roberts (Oct 11 2021 at 02:51):@Dmitri Pavlov It was more of a web form that wouldn't allow one to proceed without just hitting a button saying "I agree to all Springer's conditions". I didn't wait to see what happened if I just didn't do the thing. I had to go through several email exchanges to get to the point of avoiding agreeing to said conditions.
Dmitri Pavlov (Oct 11 2021 at 03:02):David Michael Roberts said:
Dmitri Pavlov It was more of a web form that wouldn't allow one to proceed without just hitting a button saying "I agree to all Springer's conditions". I didn't wait to see what happened if I just didn't do the thing. I had to go through several email exchanges to get to the point of avoiding agreeing to said conditions.
Did the email that invited you to complete the form say your paper will not be published without completing this form?
David Michael Roberts (Oct 11 2021 at 11:48):This was quite a while ago. If you are super interested, I can try to dig it out
Dmitri Pavlov (Oct 12 2021 at 14:01):David Michael Roberts said:
This was quite a while ago. If you are super interested, I can try to dig it out
I am actually very interested, since I am also trying to avoid any copyright transfers. If you manage to dig it out and forward it to me (including the original email from Springer asking you to complete the online form), I would be most grateful. So far my experience is that Elsevier, International Press, MSP, AMS do not require copyright transfers.
David Michael Roberts (Oct 14 2021 at 01:23):@Dmitri Pavlov here's the main part of the email.
Thank you for publishing with Springer. This message is to let you know that your articleArticle title: On certain 2-categories admitting localisation by bicategories of fractionsDOI: 10.1007/s10485-015-9400-4has gone into production. Before we can send you your proofs, we have to ask you to provide some additional information. Please go to the following website (you may need to copy and paste the URL into your browser): http://www.springer.com/home?SGWID=0-0-1003-0-0&aqId=2864078&checkval=a7b5ec58e978da8595a8739e28b61d1aPlease indicate if you would like to:order Open Choice, i.e. publish the article as open access. The published version will then become freely availabe for anyone worldwide in exchange for payment of an open access charge.order paper offprints or e-offprints of your article upon issue publicationorder poster of your article with issue cover page, article title and the authorshiporder printing of figures in color in the journaland totransfer the copyright of your article (if you do not order Open Choice)In order for the publication of your article to proceed you must go to the above website and complete the request. The entire process should take about 10 minutes.You can help us facilitate rapid publication by returning your answers within 2 working days.PLEASE NOTE: This link expires WITHIN 5 DAYS after this e-mail has been sent to you so please make sure you complete the request before this date
David Michael Roberts (Oct 14 2021 at 01:24):I then got daily reminders for three weeks while I tried sorting it out, usually looking like this:
Just a friendly reminder that we have not yet received your answer to the message below. This means that your article
Article title: On certain 2-categories admitting localisation by bicategories of fractions
DOI: 10.1007/s10485-015-9400-4
cannot be processed any further.
Please go to the following website and enter the required information as soon as possible so as not to delay the publication of your article.
http://www.springer.com/home?SGWID=0-0-1003-0-0&aqId=2864078&checkval=a7b5ec58e978da8595a8739e28b61d1a
PLEASE NOTE: This link expires within the next few days. Please make sure you complete the request as soon as possible
David Michael Roberts (Oct 14 2021 at 01:28):So I assume they would literally keep doing this until I did something. Given that I'd not seen the page proofs, it would be risky for the publisher to just dump the paper online. What did Elsevier etc do?
Dmitri Pavlov (Oct 14 2021 at 03:16):David Michael Roberts said:
So I assume they would literally keep doing this until I did something. Given that I'd not seen the page proofs, it would be risky for the publisher to just dump the paper online. What did Elsevier etc do?
For Elsevier, the process of producing and approving proofs is completely parallel to (and separate from) the process of completing the “rights and access” form. For the latter, they only send you one or two emails and that's it. They also more-or-less say that if you do not complete the form, they will publish your article anyway operating on the (legally justified) presumption that somebody who submits an article to a journal must want to see it published there.
David Michael Roberts (Oct 14 2021 at 05:12):I can send you the agreement I ultimately signed, if you want, in case you want to argue your case at some point. My argument was that my paper was licensed CC0 on the arXiv, so the work was functionally in the public domain, so I didn't own copyright in the work to sign over. Maybe not watertight, but it worked.