6 ms·

Now I am become DOI, destroyer of gatekeeping worlds

73 points by jmnicholson 11y ago

dalke 11y ago
> The injunction against citing DOI-less documents is unfortunate, because people deserve to get credit for the interesting things they say–and it turns out that they have, on rare occasion, been known to say interesting things in formats other than the traditional peer-reviewed journal article. Ain't that the truth. There's a paper that came out a few years ago which said that new method B is 3x faster than old method A. That's well and good. But there were two widely-used free software projects which had done the same thing, blogged about it, and for one case had the old code still available as a #define option. While the published article was based on a proprietary in-house package. I pointed the two previous reports to one of the authors, who had heard of one, but because it wasn't an academic paper, it wasn't worth citing. Grr! Because free software developers really have the time and money to write a paper for the $1,000 per pop open access journal in the field. I've been reading old literature, from the 1980s and older. There used to be a "Letters to the Editor" section, which was sometimes used to point out or correct these sorts of issues. No more. That same open access journal says my only options are to write a comment on their comments section (which unlike an old Letter to the Editor, no one reads, which is not indexed, and is not citable), or submit an entirely new article, at $1,000. The other thing a requirement for a DOI does it prevent citing the older literature, which doesn't always have a DOI. > [Quoting Neuron's author guidelines] "References should include only articles that are published or in press. For references to in press articles, please confirm with the cited journal that the article is in fact accepted and in press and include a DOI number and online publication date. Unpublished data, submitted manuscripts, abstracts, and personal communications should be cited within the text only." This would make certain publications impossible. For example, in my historical research, I tracked down the paper "Ciphering Structural Formulas – the Zatopleg System", Zator Technical Bulletin No. 59 (1951). This was cited by many chemical information papers in the 1950s and 1960s, and in some patents. However, the only copy I could find was in the Mooers archive at the Charles Babbage Institute. (Mooers founded Zator.) The archive controls the copyright, and lets people make copies for research purposes, so long as the archive/box/folder information is included in the citation. Which is why if you look at a paper like http://trafficways.org/ascii/ascii.pdf http://trafficways.org/ascii/ascii.pdf it has citations like: > Thomas E. Kurtz, letter to Secretary, X3, December 21, 1965, Calvin N. Mooers Papers (CBI 81), Charles Babbage Institute, University of Minnesota, Minneapolis, box 20, folder 1. Try doing that "within the text only".
- greglindahl 11y ago
  My PhD advisor had a paper which was rejected by a few major journals, so he decided to keep it as a preprint and see how many citations it could acquire. It has hundreds now. These days he'd stick it on arxiv. Arxiv doesn't appear to issue DOIs, you're supposed to add the DOI from a real journal after the preprint gets published.
JBiserkov 11y ago
http://zenodo.org http://zenodo.org will issue a free DOIs for any GitHub release you tell it to. :) See https://guides.github.com/activities/citable-code https://guides.github.com/activities/citable-code
- techdragon 11y ago
  I was about to suggest the building of this very sort of thing. Glad to know it already exists.
  - IanCal 11y ago
    There's also figshare.com where you can publish data, papers or whatever really. I published code there so my wife could cite it and we do public data releases on there too. Disclaimer: I work for digital science, a parent company.
haddr 11y ago
"The right perspective, I would argue, is to embrace the benefits of technology and seek out new evaluation models that emphasize open, collaborative review by the community as a whole instead of closed pro forma review by two or three semi-randomly selected experts." Can't agree more with that. The problem is that it's hard to change the system that has already been built around gaming this status quo. Maybe some sort of blockchain for publications & peer reviews that could decentralize and break with current world ruled by few publishers? Anyway I have a feeling that any system here could be gamed too...
verisimilidude 11y ago
Rules against citing DOI-less content are easier to understand once you understand a bit more about life within an academic publishing house. An article with a DOI has more longevity. This is mentioned in the post, but I don't think the author fully grasps the importance of this. Plain URLs break down all the damn time; publishers/journals redesign their sites and migrate between technology partners often. It only takes a few Very Important People to flip out about a missing piece of data or a broken link before a publisher will start to put safeguards into the production workflow to prevent such complaints. Mandating DOIs alongside citations is an easy policy toward that goal. An article with a DOI will almost surely be supported by a technology back-end that can keep that DOI up-to-date, generate manifests for (C)LOCKSS, submit materials to archival services like Portico, etc. DOI-bound citations are much more efficient throughout the publishing process. Editors can use DOIs to quickly verify and proof citations. If a publisher's workflow requires sending papers to an XML composition vendor for tagging, DOIs can make that process more automated and accurate. Some of the more sophisticated XML vendors can generate all the citations given only the DOIs. This might not sounds like much, but it's notoriously easy to make mistakes on citations throughout the editorial process, even with NLP and other more enlightened software in the mix. If you ever see an article with a citation list like this... * Suzuki, et al. 2004... * Suzuki, et al. 2006a... * Suzuki, et al. 2006b... * Suzuki, et al. 2006d... * Suzuki, et al. 2006f... * Suzuki, et al. 2006g... * Suzuki, et al. 2007... * Suzuki, et al. 2008b... ...think about all the double/triple/quadruple checking bullshit that goes into making sure all the text and links in that list are accurate, and know that DOIs take most of that headache away. It's substantially more efficient at scale when DOIs are included. Recently I've been thinking about how some of these problems could potentially be solved via new ideas like IPFS. A Handle-based system like DOI might not be as necessary in IPFS (given IPNS), and perhaps you could start creating and citing more diverse sources of content with some confidence in preservation/longevity. It could be a great fit for open access content.
- forgotpwtomain 11y ago
  Correct me if I'm wrong but IPFS intends to be torrent-like in terms of re-distribution - so you can no more guarantee that the article can be found there then with ordinary torrents, which I think is hardly more reliable then a url.
woodandsteel 11y ago
This is a perfect job for ipfs.
IanCal 11y ago
I see there being two main arguments, the first we agree on, permanence. That's really important when it comes to looking things up. The second however is not quality. I really don't understand that point at all. The reason it all sounds so ridiculous as they continue to describe it is, I think, because it doesn't make sense. Anyone can get a doi for anything they want really. No, the second thing is for analysis. DOIs mean you can automatically resolve papers and data. They're unique and unambiguous and easy to detect. Handwritten ones are not always possible to figure out even with a human and certainly not automatically at 100%. Imagine trying to build something like Google when pages don't have URLs but just describe how a person might go and find what they're looking for manually. That's what people are trying to deal with now.
- dalke 11y ago
  I think you interpreted the essay as an argument for or against DOIs, which it isn't. The two arguments in the essay deal with the sentence: "I think there are two main reasons for the disdain many people seem to feel at the thought of allowing authors to freely cite DOI-less objects in academic papers." As you say, "Anyone can get a doi for anything they want really." This essay describes a services for doing exactly that. The essay argues why "quality" is not relevant, which I think you agree with. Switching topics, you write that DOIs are "unique and unambiguous and easy to detect." They are supposed to be unambiguous. In practice, there are errors. Consider 10.1021/c160038a601 available from http://pubs.acs.org/doi/abs/10.1021/c160038a601 http://pubs.acs.org/doi/abs/10.1021/c160038a601 . It's a letter to the editor titled "A Utility Analysis for the MCC Topological Screen System" and supposedly by F. Bruce Sanford. If you look at the page you'll see there are two letters to the editor. Hamilton is the actual author of that letter, while Sanford is the author of a different letter to the editor. There is also supposed to be unique. However, given that "Anyone can get a doi for anything they want really" this means someone can use multiple DOI services to get multiple DOIs for the same thing.
obelisk_ 11y ago
Worth reading. Interesting perspective and useful information, conveyed in good writing. I think having the feedback at the bottom of the page say reviews instead of comments was a nice touch.

73 points by jmnicholson 11y ago

dalke 11y ago

> The injunction against citing DOI-less documents is unfortunate, because people deserve to get credit for the interesting things they say–and it turns out that they have, on rare occasion, been known to say interesting things in formats other than the traditional peer-reviewed journal article. Ain't that the truth. There's a paper that came out a few years ago which said that new method B is 3x faster than old method A. That's well and good. But there were two widely-used free software projects which had done the same thing, blogged about it, and for one case had the old code still available as a #define option. While the published article was based on a proprietary in-house package. I pointed the two previous reports to one of the authors, who had heard of one, but because it wasn't an academic paper, it wasn't worth citing. Grr! Because free software developers really have the time and money to write a paper for the $1,000 per pop open access journal in the field. I've been reading old literature, from the 1980s and older. There used to be a "Letters to the Editor" section, which was sometimes used to point out or correct these sorts of issues. No more. That same open access journal says my only options are to write a comment on their comments section (which unlike an old Letter to the Editor, no one reads, which is not indexed, and is not citable), or submit an entirely new article, at $1,000. The other thing a requirement for a DOI does it prevent citing the older literature, which doesn't always have a DOI. > [Quoting Neuron's author guidelines] "References should include only articles that are published or in press. For references to in press articles, please confirm with the cited journal that the article is in fact accepted and in press and include a DOI number and online publication date. Unpublished data, submitted manuscripts, abstracts, and personal communications should be cited within the text only." This would make certain publications impossible. For example, in my historical research, I tracked down the paper "Ciphering Structural Formulas – the Zatopleg System", Zator Technical Bulletin No. 59 (1951). This was cited by many chemical information papers in the 1950s and 1960s, and in some patents. However, the only copy I could find was in the Mooers archive at the Charles Babbage Institute. (Mooers founded Zator.) The archive controls the copyright, and lets people make copies for research purposes, so long as the archive/box/folder information is included in the citation. Which is why if you look at a paper like http://trafficways.org/ascii/ascii.pdf http://trafficways.org/ascii/ascii.pdf it has citations like: > Thomas E. Kurtz, letter to Secretary, X3, December 21, 1965, Calvin N. Mooers Papers (CBI 81), Charles Babbage Institute, University of Minnesota, Minneapolis, box 20, folder 1. Try doing that "within the text only".

greglindahl 11y ago
My PhD advisor had a paper which was rejected by a few major journals, so he decided to keep it as a preprint and see how many citations it could acquire. It has hundreds now. These days he'd stick it on arxiv. Arxiv doesn't appear to issue DOIs, you're supposed to add the DOI from a real journal after the preprint gets published.

JBiserkov 11y ago

http://zenodo.org http://zenodo.org will issue a free DOIs for any GitHub release you tell it to. :) See https://guides.github.com/activities/citable-code https://guides.github.com/activities/citable-code

techdragon 11y ago
I was about to suggest the building of this very sort of thing. Glad to know it already exists.
- IanCal 11y ago
  There's also figshare.com where you can publish data, papers or whatever really. I published code there so my wife could cite it and we do public data releases on there too. Disclaimer: I work for digital science, a parent company.

haddr 11y ago

"The right perspective, I would argue, is to embrace the benefits of technology and seek out new evaluation models that emphasize open, collaborative review by the community as a whole instead of closed pro forma review by two or three semi-randomly selected experts." Can't agree more with that. The problem is that it's hard to change the system that has already been built around gaming this status quo. Maybe some sort of blockchain for publications & peer reviews that could decentralize and break with current world ruled by few publishers? Anyway I have a feeling that any system here could be gamed too...

verisimilidude 11y ago

Rules against citing DOI-less content are easier to understand once you understand a bit more about life within an academic publishing house. An article with a DOI has more longevity. This is mentioned in the post, but I don't think the author fully grasps the importance of this. Plain URLs break down all the damn time; publishers/journals redesign their sites and migrate between technology partners often. It only takes a few Very Important People to flip out about a missing piece of data or a broken link before a publisher will start to put safeguards into the production workflow to prevent such complaints. Mandating DOIs alongside citations is an easy policy toward that goal. An article with a DOI will almost surely be supported by a technology back-end that can keep that DOI up-to-date, generate manifests for (C)LOCKSS, submit materials to archival services like Portico, etc. DOI-bound citations are much more efficient throughout the publishing process. Editors can use DOIs to quickly verify and proof citations. If a publisher's workflow requires sending papers to an XML composition vendor for tagging, DOIs can make that process more automated and accurate. Some of the more sophisticated XML vendors can generate all the citations given only the DOIs. This might not sounds like much, but it's notoriously easy to make mistakes on citations throughout the editorial process, even with NLP and other more enlightened software in the mix. If you ever see an article with a citation list like this... * Suzuki, et al. 2004... * Suzuki, et al. 2006a... * Suzuki, et al. 2006b... * Suzuki, et al. 2006d... * Suzuki, et al. 2006f... * Suzuki, et al. 2006g... * Suzuki, et al. 2007... * Suzuki, et al. 2008b... ...think about all the double/triple/quadruple checking bullshit that goes into making sure all the text and links in that list are accurate, and know that DOIs take most of that headache away. It's substantially more efficient at scale when DOIs are included. Recently I've been thinking about how some of these problems could potentially be solved via new ideas like IPFS. A Handle-based system like DOI might not be as necessary in IPFS (given IPNS), and perhaps you could start creating and citing more diverse sources of content with some confidence in preservation/longevity. It could be a great fit for open access content.

forgotpwtomain 11y ago
Correct me if I'm wrong but IPFS intends to be torrent-like in terms of re-distribution - so you can no more guarantee that the article can be found there then with ordinary torrents, which I think is hardly more reliable then a url.

woodandsteel 11y ago

This is a perfect job for ipfs.

IanCal 11y ago

I see there being two main arguments, the first we agree on, permanence. That's really important when it comes to looking things up. The second however is not quality. I really don't understand that point at all. The reason it all sounds so ridiculous as they continue to describe it is, I think, because it doesn't make sense. Anyone can get a doi for anything they want really. No, the second thing is for analysis. DOIs mean you can automatically resolve papers and data. They're unique and unambiguous and easy to detect. Handwritten ones are not always possible to figure out even with a human and certainly not automatically at 100%. Imagine trying to build something like Google when pages don't have URLs but just describe how a person might go and find what they're looking for manually. That's what people are trying to deal with now.

dalke 11y ago
I think you interpreted the essay as an argument for or against DOIs, which it isn't. The two arguments in the essay deal with the sentence: "I think there are two main reasons for the disdain many people seem to feel at the thought of allowing authors to freely cite DOI-less objects in academic papers." As you say, "Anyone can get a doi for anything they want really." This essay describes a services for doing exactly that. The essay argues why "quality" is not relevant, which I think you agree with. Switching topics, you write that DOIs are "unique and unambiguous and easy to detect." They are supposed to be unambiguous. In practice, there are errors. Consider 10.1021/c160038a601 available from http://pubs.acs.org/doi/abs/10.1021/c160038a601 http://pubs.acs.org/doi/abs/10.1021/c160038a601 . It's a letter to the editor titled "A Utility Analysis for the MCC Topological Screen System" and supposedly by F. Bruce Sanford. If you look at the page you'll see there are two letters to the editor. Hamilton is the actual author of that letter, while Sanford is the author of a different letter to the editor. There is also supposed to be unique. However, given that "Anyone can get a doi for anything they want really" this means someone can use multiple DOI services to get multiple DOIs for the same thing.

obelisk_ 11y ago

Worth reading. Interesting perspective and useful information, conveyed in good writing. I think having the feedback at the bottom of the page say reviews instead of comments was a nice touch.