Community/Members papers section

[FCLC]

Based on hpc-social/blog#3, idea is having a sort of mini arxiv for the HPC community that serves to hold papers, posters, talks etc. in a common directory.

Ideas for organizing in a clean, accessible way is to make use of the existing work behind the ar5iv project that converts latex posts into web compatible HTML5 posts.

example and information on Ar5iv:

https://ar5iv.labs.arxiv.org/html/1910.06709
https://ar5iv.labs.arxiv.org/
https://blog.arxiv.org/2022/02/21/arxiv-articles-as-responsive-web-pages/

Converts Latex to what is basically an HTML5 compliant webformat.

For hosting: posters, papers and articles we're fine (the format has the bonus that it works super well for things like screen readers on the accessibility front 🙌🏻)

For talks I don't see a world where we're hosting our own full video CDN unless the funding is significant. Youtube or the likes is more than fit for purpose and probably better in terms of bringing in new people to the community, especially students.

[vsoch]

Pinging @davidbeckingsale - we actually just were talking about this today, and maybe it could be a collaborative effort! I think we'd want to take a shot at not just giving a home for "non sexy research software that other journals to be unnamed would not call research software" but also a way to rethink credit and value. I had a server-based idea around citelang, which asserts that the source of truth for software should be in package managers, of course I haven't built anything yet. Of course I could! <3

So I definitely want to think more about this idea and (maybe?) if we can come to a cool idea can make it over the holiday break.

[vsoch]

I guess my thing is that mastodon (at least the UI on my computer / phone) is just kind of slow? it's not pleasant to use.

[FCLC]

wow, sorry, phone posted in the wrong thread 😅

[vsoch]

haha no worries, I got it!

[alansill]

There are several people (underfox, Tadashi Ogawa, others) who routinely post reviews or summaries of articles, papers, even patents) that they find interesting. These often have interesting commentary on the points that make the item(s) interesting. Given that things like Academia.edu, Research Gate, etc. exist where users can post their publications for others to see and follow, how would we differentiate such a service and add useful features (for example, commentary or question/answer sections, etc.?)

[vsoch]

Interesting, although arguably if I've never heard of it (let alone use it) it's not meeting the need that I have.

[alansill]

Um, ouch.

[vsoch]

I'm sorry I didn't mean that to be ouch, I just haven't heard of this until now!

[alansill]

NP, just an organization I've been volunteering with for the past 20 years or so! ;-). It's one of the sponsors for hpc.social btw. (We're in the process of modernizing our infrastructure and a partnership with hpc.social also made sense.)

In this case I wasn't so much pointing out the specific infrastructure, which was pretty much specific to grids, as the problems of hosting "live" documents when you want them to be the basis for work in he community and how that organization managed things to allow documents to be proposed, commented on, reviewed, published, and maintianed with options for revision and/or replacement. The full process is described in https://ogf.org/documents/GFD.152.pdf. It's also a good example of an open participant-driven organization.

Returning to your original concept, a totally open live document could just be a web site, I suppose. If you want process around that for how to propose, vet, review, and publish updates that could mean new tooling but the principles woud be good to establish ahead of time. Since the formation of grids, the process of publishing open specifications for people to work from has been expanded greatly by other organizations including just putting things out on repositories as individuals to follow (this comment is specific to specifications but applies in general to any document for which you want an organized process for revision).

I like the idea of live documents but you need a process for revision if the changes to the document will produce changes in how people refer to or use its contents. At some level, any document-production software will face the same issues. In summary, it depends on what you mean by "publication" and whether a work product that is subject to revision can have a place in the chain of references for a topic or for how to do something if it is subject to change at any time, and if so, how to manage that revision process.

[vsoch]

Makes total sense! I think what I’m imagining is more scoped to live publication - the “living documentation” is an idea I came up with in grad school that generally says something is continually being updated (but in this context not explicitly docs).