1

I wrote a paper that I want to put on my webpage, but that I don't want to publish anywhere. I think that I need a timestamp on it, in case someone decides to plagiarize it. Is GitHub adequate for this?

ff524
  • 108,934
  • 49
  • 421
  • 474
Ken - Enough about Monica
  • 2,453
  • 3
  • 16
  • 32
  • 7
    Did you consider arXiv? – Dmitry Savostyanov Jul 15 '14 at 15:22
  • 1
    Possible duplicate of http://academia.stackexchange.com/questions/712/how-to-prevent-plagiarism-of-my-papers http://academia.stackexchange.com/questions/23367/how-can-i-time-stamp-my-data-without-publishing-it or http://academia.stackexchange.com/questions/435/is-it-advisable-to-put-entire-source-of-my-thesis-up-on-github/442#442 – mhwombat Jul 15 '14 at 15:24
  • @DmitrySavostyanov - One must be invited to arXiv, and if I want to change the paper later, it is my understanding that I cannot do that with arXiv (I must instead upload a new version, and keep the old one as well.) – Ken - Enough about Monica Jul 15 '14 at 15:26
  • If I am not very much mistaken, you can access older versions on ArXiv (see for example this paper). Of course you do not want to constantly revise your paper this way. As for the invitation, if you know other people who have published in your field, there is a high chance that one of them can invite (endorse) you, as you become an endorser quite quickly. – Wrzlprmft Jul 15 '14 at 15:42
  • 1
    It is not difficult at all to be endorsed on arXiv, if you know people in your field. You're right that arXiv provides full version control with all timestamps - in the same way as public github repo does. – Dmitry Savostyanov Jul 15 '14 at 15:46
  • @horsehair Is your aim to be able to edit in case you spot mistakes, or rather - you aim is to have continuous development? – Piotr Migdal Jul 15 '14 at 15:53
  • I don't know anyone in my field. My aim is to have a timestamp that others will believe, in the case that someone claims to have produced my work themselves. I will only make small changes to the document from now on. – Ken - Enough about Monica Jul 15 '14 at 15:59
  • 3
    @Dmitry , I would say ArXiv does limited version control. (For example, you can't remove versions.) For the basic purpose of providing an independent timestamp, ArXiv and Github suffice. Legally, a notary is better. – Not Quite An Outsider Jul 15 '14 at 16:40
  • 3
    Remember that neither GitHub nor ArXiv can guarantee to still be in business when you need that evidence. If you care enough to want a clear timestamp, getting it notarized is a fairly cheap investment. Especially if someone in your school's staff is a notary (fairly common) and offers discount services to colleagues (also fairly common). – keshlam Jul 15 '14 at 17:07
  • 1
    The proposed duplicate does not seem like a duplicate, because it is about data, not a paper. – ff524 Jul 15 '14 at 19:38
  • 1
    I don't know anyone in my field. — Then you have a much more serious problem than time-stamping your work!! – JeffE Jul 16 '14 at 01:24
  • @JeffE - I just finished my Master's, and I generally work in a different field. I know my adviser, but he doesn't use arXiv. Am I expected to know more people? – Ken - Enough about Monica Jul 16 '14 at 06:36
  • 1
    Then you do know someone (even if he's not directly in your field)! Have you asked your advisor? – Christian Clason Jul 16 '14 at 07:40
  • @ChristianClason - Read the comment again please. He doesn't use arXiv – Ken - Enough about Monica Jul 16 '14 at 09:22
  • Yes, but he might know other ways to achieve what you want (certified self-publication) besides arXiv; for example, he could ask your library to publish your paper as a technical report in their digital repository, if such a thing exists. – Christian Clason Jul 16 '14 at 09:29
  • Why do you want to 'put the paper on your webpage', but not 'publish'? If you believe there is a chance of plagiarism, then your paper should be interesting enough to send to a journal, shouldn't it? – Mark Peletier Jul 16 '14 at 19:17

2 Answers2

5

Let's be scientists (It was fun to try out): below is a small repository, where I first tried 'lol' to have a timestamp in the future ... doesn't work; but at least in 'history' I could backdate (bit unlikely that I authored s.th. in git 1980 ;-).

But at least the day that I pushed is set by github.com. In total, though, I wouldn't trust this scheme. ``I forgot to push but look at the authored date, I totally solved it years ago''.

Jul 15, 2014

history 2b5d4208aa Browse code choener authored on Jan 1, 1980

lol fc2a68a571 Browse code choener authored just now

https://github.com/choener/lol/commits/master

choener
  • 633
  • 5
  • 9
  • For plagiarism protection git timestamps may be useless, but exposing code publicly (public awareness, possible forks, Wayback Machine) GitHub may serve its role (vide there was a post that the best way to protect against plagiarism is to sent a draft to many scientists). – Piotr Migdal Jul 17 '14 at 10:24
  • 1
    Absolutely! I only considered the technical 'timestamp' aspect. I prefer to have all my libraries on github and all my preprints on my webpage + preprint servers. (And the community is small enough that it doesn't matter for me) – choener Jul 17 '14 at 13:02
1

Poor man's copyright protection can be done by sending a copy of the work to yourself by registered mail. The timestamp is provided by the federal government and as long as the envelope remains sealed, it is proof of creation on or before that date. Of course, I would also get a notarized or authenticated document as well.

  • 2
    Doesn't work for copyright, so of dubious legal value: http://www.snopes.com/legal/postmark.asp – Christian Clason Jul 15 '14 at 22:23
  • 1
    Unless you are planning on suing somebody for publishing your idea, I don't see why the law has much to do with this. – mako Jul 15 '14 at 23:44
  • It's just to timestamp when you had written some idea in that form. I made no claims about legal value for a lawsuit. Also, the Snopes link only applies to USA and UK. –  Jul 16 '14 at 01:59
  • Technically, this only proves existence of the envelope on or before that date (the postmark is not tamper proof); as long as you have possession of the sealed envelope, I don't see how this would convince anyone who wouldn't already take your word for it. Hence the service of a notary. (But if you visit France, the Soleau envelope is a poor man's notary.) – Christian Clason Jul 16 '14 at 07:16