Is there a free-world [0] equivalent to GHTorrent [1] for storing git ephemera [2]?
[0] https://developers.reverseeagle.org/replace/github
[1] https://ghtorrent.org
[2] https://investigating-archiving-git.gitlab.io/updates/define-scholarly-ephemera
CC: @VickyRampin
PS: @resist1984 You quoted a very good URL for the ethics of github recently (IIRC) - could you remind me (us) of the URL please? (It was incomplete though, without mentioning the racist action of excluding people from some countries.)
@resist1984 @VickyRampin @switchingsoftware @codeberg
Excellent :))). Thanks!
That's now in the footnote on page 15:
https://upload.disroot.org/r/PaJyxAQD#MTpx0hT9hFx0qqJ6gF90NNgWen4FUPvO97O7kvwWuqs=
Commit dc06405 (or a variation on it) should appear in the final
https://arxiv.org/abs/2006.03018 version, i.e. a postprint, if my pull request https://codeberg.org/maneage/paper-concept/pulls/2 is accepted.
PS: The first link in the git.sdf.org list is archived at:
https://web.archive.org/web/20200819132217if_/https://infosec.exchange/@bojkotiMalbona/104637098084869887
@boud @codeberg @switchingsoftware @VickyRampin i love the way you did your footnotes, using superscripts for hyperlinks at the bottom of the page & bracketed references to MLA/APA citations at the end. A also love the code listings; great idea to color the comments gray inside of a verbatim environment. Is the tex file published anywhere? It would be cool to use the attachfile pkg to embed a copy of the latex code in the pdf.
@VickyRampin @switchingsoftware @codeberg @boud i think it would go along with your thesis. Couple anomalies: the superscripts use the same numbering system as the citations, and there are conflicts (citation 38 can be confused with superscript 38). Starting citation numbers at 100 would fix that. It's good that you give an archive link on 38, but archive.today is a cloudflare site that redirects to another. I suggest archive.org
attachfile: Nice idea, though I'm not convinced. I've posted it as a possible task: https://savannah.nongnu.org/task/index.php?15984 (coincidence in the number!)
Citations vs footnotes: I fully agree, but this is constrained by journal style.
Archive.today vs Archive.org: thanks for the pointers on the ethics. I was aware of archive.today using GAFAM captcha :( . You're right about the short URL - my choice was based on the #TyrannyOfConvenience .
@boud @VickyRampin If the attachfile pkg is incompatible, another approach is to have the makefile do a "pdftk attach_files" command. I think a PDF gets more widely distributed and outlives the *.tex that it came from, so embedding the source in the PDF ensures that the source inherits the longevity of the PDF in a reproducable way.
@boud @VickyRampin Consider this failure case reguarding a study at Carnegie Mellon by Dr.Cranor: https://web.archive.org/web/20170329100626/dl.acm.org/citation.cfm?id=2911988 They attempted to distribute a zip file w/all the raw data but they botched it. The zip only contains the same PDF file that sits next to the zip. When really there was no need for the zip. the PDF should have included the files
@VickyRampin @boud that's not just a case of missing tex source code, but that project involved writing a bot that crawled the web for banks' privacy policies. I would love to see that code and the raw data, but it was lost. If it had been integrated into the PDF, it would have persisted.
@boud @VickyRampin btw here's the link to the zip file: https://web.archive.org/web/20210613163219/https://dl.acm.org/doi/10.1145/2911988
@VickyRampin @boud i often embed the tex files in the PDF produced by the tex code for ad hoc collaboration, particularly if I don't know if the recipient knows what to do with a tex file. And in some cases the other person extracts the tex files, makes some edits, and sends them back to me.