[eecc]

Ah cool… I also took a stab at something similar several years ago: https://github.com/ecausarano/heron

Also at the time I was considering IPFS.

But I guess the real trick is implementing a WOT to implement peer review and filter out the inevitable junk that will be published

[westurner]

"Help compare Comment and Annotation services: moderation, spam, notifications, configurability" executablebooks/meta#102 https://github.com/executablebooks/meta/discussions/102 :

> jupyter-comment supports a number of commenting services [...]. In helping users decide which commenting and annotation services to include on their pages and commit to maintaining, could we discuss criteria for assessment and current features of services?

> Possible features for comparison:

> * Content author can delete / hide

> * Content author can report / block

> * Comments / annotations are screened by spam-fighting service

> * Content / author can label as e.g. toxic

> * Content author receives notification of new comments

> * Content author can require approval before user-contributed content is publicly-visible

> * Content author may allow comments for a limited amount of time (probably more relevant to BlogPostings)

> * Content author may simultaneously denounce censorship in all it's forms while allowing previously-published works to languish

#ForScience

[westurner]

FWIW, archiving repo2docker-compatible git repos with a DOI attached to a git tag, is possible with JupyterLite:

> JupyterLite is a JupyterLab distribution that runs entirely in the browser built from the ground-up using JupyterLab components and extensions

With JupyterLite, you can build a static archive of a repo2docker-like environment so that the ScholarlyArticle notebook or computer modern latex css, its SoftwareRelease dependencies, and possibly also the Datasets can be run in a browser tab with WASM. HTML + JS + WASM

[hugoroussel]

Exactly the problem that I have. Many edge cases too.

[kloch]

What do you mean by junk? Spam and Abuse, or scientific papers that the reviewer simply doesn't like?

[eecc]

Well, as anyone can and will publish, it will be stuffed with junk, porn, trash of all sorts, so my plan was to implement a filter to ignore any data published outside of one’s WOT.

Furthermore, a user searching for “reviewed” data and papers would normally filter for items with enough “endorsement” metadata items signed by known WOT actors.

I haven’t figured a mechanism to prevent “review rings”, although being totally transparent it should be easy to spot them.