Post AXpTLdnUghqqo7obmi by anders@thoresson.social
(DIR) More posts by anders@thoresson.social
(DIR) Post #AXmchhslZ1hXrVMfOi by anders@thoresson.social
2023-07-17T11:01:53Z
0 likes, 0 repeats
What I appreciate the most with #tana is that I can write stuff down wherever I’m in my graph.Tasks in a meeting note.Projects ideas in a task. A random thought in a project log. With tags and attributes, what I write down is surfaced in the right place. No other tool I’ve tried had let me work this way. The real value: Less silos in my #TfT. Task. Ideas. Projects. Notes. All part of the same graph.
(DIR) Post #AXmdVWklDRj9pCNOHQ by anders@thoresson.social
2023-07-17T11:10:54Z
0 likes, 0 repeats
I have tried to do the same in #obsidian and #logseq a couple of times. Some of what I’ve done in #tana is technically possible. But at least the way I managed to do it with a lot more friction. And with that, I started to think about file format in a different way.
(DIR) Post #AXmdc0qKFv2Yss4VlY by anders@thoresson.social
2023-07-17T11:12:05Z
0 likes, 0 repeats
Markdown used to be a top priority of mine. But seeing what could be achieved when data was stored in other ways (together with what’s really in those markdown files of Obsidian and Logseq, not exactly human-readable anyway), the results I get out of a tool started to matter more. And with the json export, I feel confident enough that I (with the help of someone more skilled in json) can recreate what I put into Tana.
(DIR) Post #AXmdfme0zIEQ0S6z44 by anders@thoresson.social
2023-07-17T11:12:46Z
0 likes, 0 repeats
(But this is obviously a pendulum, so let’s see where I stand on this in a year or so.)
(DIR) Post #AXmoO7PI9ZEmgR0K5A by spinningthoughts@pkm.social
2023-07-17T13:12:45Z
0 likes, 0 repeats
@anders tbh to me it feels less like a limitation of markdown (the constant arguments over just how one does the keyed metadata aside) and more the software. Logseq gives you tags so you can break silos up that way at least, but where Tana keeps a record of your tags, you would have to work in Logseq to dig up all of your tags… Besides the raw tech infrastructure it‘s the standard UX tooling to discover nodes that really makes this fast, fluid workflow possible.
(DIR) Post #AXmoassjT7mlOnrVI0 by anders@thoresson.social
2023-07-17T13:15:07Z
0 likes, 0 repeats
@spinningthoughts Agree. Probably much more UX/UI than file format. The reason I frame this as Markdown related is because for a long time, the file format has been such a high priority for me that I’ve sacrificed other aspects of the tools I use.
(DIR) Post #AXmp0VdmV8ywqa5V5s by spinningthoughts@pkm.social
2023-07-17T13:19:43Z
0 likes, 0 repeats
@anders I would really like it in markdown or something for other reasons, but yes - the pure raw accessability of the file format cannot indefinitely inhibit functionality and flow. If accessability becomes in itself friction you have an issue.
(DIR) Post #AXmwspxRsUBRB08i6y by EpiphanicSynchronicity@pkm.social
2023-07-17T14:47:57Z
0 likes, 0 repeats
@anders Whenever I open any of my #Obsidian notes in a basic editor, I find them eminently readable.I’m sure #Tana is incredible, but given that their website screams VC-funded and there’s proprietary lock-in, it seems destined in the long term for the process Cory Doctorow @pluralistic has described as enshittification.And I have none of your faith that resurrecting and organizing five or six years of permanent notes out of such an intricate system from a json export will be no big deal.
(DIR) Post #AXn1cSU9saqXW1RsRc by nhan@mastodon.social
2023-07-17T15:40:48Z
0 likes, 0 repeats
@andersWould be curious to see in detail what has changed since 1.5 year ago https://myttl.blog/what-is-the-exit-plan-for-your-notes-2/.Relatively speaking, previously you valued longevity over functionality of your note, and now it's the other way around?
(DIR) Post #AXpSIlG6OFL57rG9LM by ednico@pkm.social
2023-07-17T17:35:30Z
0 likes, 0 repeats
@EpiphanicSynchronicity @anders @pluralistic I agree with you. And also why I was put off using plugins on Obsidian and Logseq - they add complexity and move you away from simple reusable notes for the future. It's funny to think that all the manuscripts saved are in simple format - not sure if in 100 years people will say "oh check out this beauty of a hard drive". Still think that #tana is really solid, but they need to improve their export
(DIR) Post #AXpSImYvXjzXAX4hKC by anders@thoresson.social
2023-07-18T19:49:27Z
0 likes, 0 repeats
@ednico @EpiphanicSynchronicity @nhan First: Really appreciate the pushback here! This is both important – and even a bit fun – to think about, and discuss!
(DIR) Post #AXpSaQCPQa0z1UxA6y by anders@thoresson.social
2023-07-18T19:52:40Z
0 likes, 0 repeats
@ednico @EpiphanicSynchronicity @nhan A couple of thoughts:1. I share your experience here, @ednico. Markdown sound good, but as soon as you start enabling plugins to get features beyond #Obsidian's core, a lot of that value is lost. In practice, a lot of my Markdown files in my Obsidian’s vault are pretty useless when opened elsewhere. And #Logseq is even worse, already out-of-the box.
(DIR) Post #AXpSusQQC4eAM9GFzk by anders@thoresson.social
2023-07-18T19:56:21Z
0 likes, 0 repeats
@ednico @EpiphanicSynchronicity @nhan 2. Answering your question here (https://mastodon.social/@nhan/110730179303483083), @nhan, two things have changed:a) As per 1, the longevity is not in wether you use Markdown or something else as the file format, but wether what you store there is well structured etc. My Markdown files, created over 10+ years, are a mess. What #Tana, and perhaps others offer, is the possibility to alter your data structure as needed. Could be done if you use Markdown as well, but harder.
(DIR) Post #AXpT1bJQDXp1JCC3Tk by anders@thoresson.social
2023-07-18T19:57:35Z
0 likes, 0 repeats
@ednico @EpiphanicSynchronicity @nhan b) And yes, my priorities have shifted somewhat, giving more weight for value creation here and now, and less for longevity. But the latter is still important.
(DIR) Post #AXpTEkeWnjxo2fKv1k by anders@thoresson.social
2023-07-18T19:59:57Z
0 likes, 0 repeats
@ednico @EpiphanicSynchronicity @nhan 3. @ednico, while we have manuscripts that are hundred of years old, we have less of the working material from the creators. I have started to think about my own tools and processes the same way. #Tana is a great thinking environment, that helps me produce things that are used outside Tana. So even if some of my working material would be lost, what it helped me produce will (perhaps) still be around.
(DIR) Post #AXpTLdnUghqqo7obmi by anders@thoresson.social
2023-07-18T20:01:12Z
0 likes, 0 repeats
@ednico @EpiphanicSynchronicity @nhan 4. In a thread with @kepano, I recently asked about the metadata work being done in #Obsidian. If it becomes easier to work with metadata in Obsidian, it's quite possible that the pendulum swings again. :)
(DIR) Post #AXpc2ZGAEXCkotCnrM by joonhyeok_ahn@pkm.social
2023-07-18T21:38:32Z
0 likes, 0 repeats
@anders What is the edge over Obsidian? I haven't used #tana at all so wonder what are the advantages if I switched over to it.
(DIR) Post #AXpcOFLcwi9jR268y8 by anders@thoresson.social
2023-07-18T21:42:29Z
0 likes, 0 repeats
@joonhyeok_ahn The big difference to me is how I am able to write ”where I am” in my graph. If I am in a meeting for project A and realize a todo for project B, I don’t have to switch context. I can still jot that todo down in the meeting notes for A, tag it as a todo, and assign it to project B. And have it show up in project B’s task list.
(DIR) Post #AXpcUwY3EzgD9k6QS0 by anders@thoresson.social
2023-07-18T21:43:43Z
0 likes, 0 repeats
@joonhyeok_ahn Some of it can be done using queries and plugins in Obsidian. But at least to me, it’s too hard to have the frictionless experience that Tana allows for.
(DIR) Post #AXpck0svOpLKwx3Heq by anders@thoresson.social
2023-07-18T21:46:27Z
0 likes, 0 repeats
@joonhyeok_ahn Expressed in a more general way: I find it much easier to creat relationships (with semantic meaning) in Tana. That means that I DO create more links between notes/thoughts/etc in Tana, which in turn adds value to my thinking/working processes.
(DIR) Post #AXph7NXNkAEPgNN9TU by joonhyeok_ahn@pkm.social
2023-07-18T22:35:28Z
0 likes, 0 repeats
@anders Sounds like it fits better into your workflow! That's nice. I feel like everyone has different preferences for #pkm and that's good.
(DIR) Post #AXpkF5uhS711CaeME4 by stevepaulson@mstdn.social
2023-07-18T23:10:25Z
0 likes, 0 repeats
@anders @joonhyeok_ahn As a new Tana user I do find it easy to create these relationships but it is so far not clear to me that it is any easier than doing it in Roam which is still my primary TfT#Tana #Roam
(DIR) Post #AXpwkDaq9GLtRmy5OS by nhan@mastodon.social
2023-07-19T01:30:32Z
0 likes, 0 repeats
Here's the thread Anders mentioned for those who want to know morehttps://thoresson.social/@anders/110702317457856052
(DIR) Post #AXrPQ1DLYT3h5kewvA by alexl@pkm.social
2023-07-19T18:26:30Z
0 likes, 0 repeats
@anders @ednico @nhan I think what could solve everything once and for all is:1. A (graph?) database that runs locally (or remotely but with API to access it)2. Local or remote apps to manipulate that data (including a Tana-like UI)3. A *local* service that mount a virtual file system with notes (in MD or whatever) in folders. The folder structure is derived by traversing the graph of notes using some directives given by the user. For example "new hierarchy starting with <years>, then..."
(DIR) Post #AXrdKx3J04yk5tROD2 by anders@thoresson.social
2023-07-19T21:02:31Z
0 likes, 0 repeats
@stevepaulson @joonhyeok_ahn I have never really used Roam, so can't comment on that. :)
(DIR) Post #AXresXsFDeqnnwqkLI by anders@thoresson.social
2023-07-19T21:19:48Z
0 likes, 0 repeats
@alexl @ednico @nhan Interesting idea! Wonder what the odds for some standardized file format is.
(DIR) Post #AXrhPoIxLnUsEEKkcK by alexl@pkm.social
2023-07-19T21:48:08Z
0 likes, 0 repeats
@anders Do you mean if the "virtual" notes should be Markdown files?I think the DB should have a rich schema with things like headers, tables, lists, tasks etc¹ and how they translate to file formats could be a user configuration: maybe Markdown, Djot, Asciidoc or HTML files. These files would contain only what is supported by that format, the rest (generally metadata) would remain in the DB.1: the GUI used to take notes could be WYSIWYG, Markdown, spreadsheets or whatever input method
(DIR) Post #AXtG8xPtuUlv7UF0ue by spinningthoughts@pkm.social
2023-07-20T05:17:54Z
0 likes, 0 repeats
@alexl @anders I wonder if the graph database could basically store a - well not an AST, a "Concrete Syntax Tree" to borrow from Tree-sitter‘s terms. Then you use the Pandoc model of input and output translations to and from that common data model. You could implement it as JSON-likes in a document property graph database. Would generally be pretty powerful.
(DIR) Post #AXtG8yHmg9sDobS2ts by anders@thoresson.social
2023-07-20T15:52:03Z
0 likes, 0 repeats
@spinningthoughts @alexl Ok, now you lost me in technical details. :)
(DIR) Post #AXtHvbyUWIX6lJBVEe by spinningthoughts@pkm.social
2023-07-20T16:12:01Z
0 likes, 0 repeats
@anders @alexl :blob_thinking_smirk: :blob_rainbowheart: Sorry, this just fell on some fertile mental ground and I wanted to see Alex' possible feedback to the spitball. :D