(this can be applied to publication in general, but let's just focus on academic publishing)
(also, I might not say something new)
(actually, I am explain a workflow - the technology is already there)
Consider this as notes, and please feel free chip in.
Standard workflow
Standard academic publication works as follows:
Author writes a paper
Author send the paper to a conference/editor
Editor send comments to Author (and author can change)
Editor publish the paper
[Author can keep on doing changes (because why not)]
Author host his copy his profile
Published version:
At point (4) the paper is given a unique identifier
The paper is printed and/or stored on the publisher website/author website/arxiv
Issues (there could me more than this)
Location: Who hosts the paper? Very likely single point of failure - if you self host a paper, will your paper be still on your website in 100 years? domain names are rented!
Pointing to papers: Imagine in a paper, blogpost website, HTML paper, Linked Research, or any other medium, when I want to refer to a paper. How can I do so? A link to a resource that will disappear? Merkle link?
Merkle dag + IPFS solutions
(I mention the merkle dag, since a paper might not just be a PDF, but a folder with an HTML page or files & so on)
Author writes the paper
Author and Editor agree on the final version (they could also decide to sign the paper - but let's leave this for another conversation)
This gets hashed (now this hash represents this point in time of the paper - newer version can always point back in history via POST)
The editor/author/academic institution will store this file (pin it) and serve it via IPFS
A set of academic institution can insure that the file is pinned (using pincoop?)
Now:
Papers have unique ID that represent the state of the paper at publication time
The hash ensure integrity of the paper. In fact anyone can now check if the paper published matches the paper that they have
This is mainly inspired by Linked Research which envision a future in which research could be self-hosted and written in HTML+RDFa. So in those cases linking between papers could really be an issue, but with this infrastructure in place we could change academic publishing.
Possible actions
Transfer Arxiv on IPFS
Run our own Arxip
Interesting challenges
Meta data? (maybe we could be inspired by Linked Research)
nicola
commented
8 years ago
(this can be applied to publication in general, but let's just focus on academic publishing) (also, I might not say something new) (actually, I am explain a workflow - the technology is already there)
Consider this as notes, and please feel free chip in.
Standard workflow
Standard academic publication works as follows:
Published version:
Issues (there could me more than this)
Merkle dag + IPFS solutions
(I mention the merkle dag, since a paper might not just be a PDF, but a folder with an HTML page or files & so on)
Now:
This is mainly inspired by Linked Research which envision a future in which research could be self-hosted and written in HTML+RDFa. So in those cases linking between papers could really be an issue, but with this infrastructure in place we could change academic publishing.
Possible actions
Interesting challenges