What interesting discussions as of late! I think @dusjagr and @naikymen both make good points. Whatever we decide to do, I think we need people who are willing to put in the sustained work needed to tend this digital garden.
I am also interested in @Sarah’s suggestions as to how a loose, international community like GOSH can find the people to come together for this gardening. Happy to follow up in early 2023.
As for the directory of interesting OScH projects that @NKArranz mentioned, I think it would first of all be a good showcase of why GOSH is important. If we successfully create this directory, it would be good to give it some visibility on GOSH’s website. One possibility is to use this directory as a starting point for knowledge management. Once we have it set up, we can start thinking about how we add other bits of knowledge to this garden. On the other hand, maybe it will be simpler to keep this OScH directory independent.
Also somewhat randomly, I was skimming through the Open Scientist Handbook by Bruce Caron, and it’s got a chapter on idea gardening, though it’s more about how ideas cross-pollinate and how current academic institutions are not conducive to great ideas…
…
How do we move from an approach entirely based on temporary projects to an approach based on community-based sustainable infrastructure?
What kinds of social and technical infrastructures could support the Knowledge Commons?
…
There are free alternatives to Publish. 0. 1. 2. 3.
The Obsidian graph interface is similar to @Paul 's link above.
@jarancioabove mentions Zotero, which I agree is a great resource. I generally just drag notes/annotations from Zotero into Obsidian (which auto-generates a link back to that note), but there are other integrations. 1. 2.
If I can use it (albeit poorly), it can’t be too difficult. I’m not necessarily pushing for obsidian, but perhaps some git system here that leverages links between .md files, tagging, graphs, etc. with a simple jekyll site might be more active and easier/nicer to navigate than a wiki.
When I get to it this summer, I will collate the open source hardware information into
something akin to The One Card Deck to Rule Them All that I generated several years ago at https://engineeringunleashed.com/card/2167
I could definitely use some help organizing all that information!
Jim Brenner
Really love how comprehensive @eric’s post is about Obsidian! It looks like an amazing tool, though I’m concerned that it’s closed source which not only does not respect my digital freedoms, but also risks costly vendor lock in and forced obscelesence.
I wonder if there’s a way to achieve what Obsidian can do with TiddlyWiki as @DrBrian suggests? I’ve known about TiddlyWiki but hesitated when I learned that it uses its own mark up language instead of Markdown. Though I also know that TiddlyWiki is highly extensible, and maybe there’s a way to make it work for me.
Also looking forward to @jbrennerFIT’s “One Card Deck to Rule Them All”, which is a curious name!
Logseq is a good open-source alternative to Obsidian. I’m a happy daily user. The main difference between the two (besides open vs closed source) is Logseq is block-based (in the outliner sense) and Obsidian is page-based. They each have their pros and cons for that reason.
Anyhow, happy to answer any questions folks might have. Feel free to DM.
Here’s a related and useful web standard: Robust Links
As you know, hyperlinks on the web are subject to link rot, where the pages they point to might change or disappear over time. Robust links is a proposed standard for the <a href> </a> HTML tag for hyperlinks to embed an additional link to an archived version of the target page. The archived page could, for example, be a snapshot on the Internet Archive’s Wayback Machine.
I love this idea! Something to keep in mind, though it’s not supported in Markdown or other mark up languages at the moment.
Here’s an interesting open source tool called Danswer, which “allows you to ask natural language questions against internal documents and get back reliable answers… You can connect to a number of common tools such as Slack, GitHub, Confluence, amongst others.” Another discussion thread is on Hacker News.
Sadly it currently uses OpenAI LLMs to works its magic, which is not open source, but plans work with other models in the future.
I probably won’t start using this right now, but such a tool could help partially solve the discoverability problem of a large knowledgebase/digital garden by helping you query it using natural language. I.e. “Tell me where in my notes I wrote about X.”