keen to learn what you do, here, @U3X7174KS 🙂

i think i've already shared my 'models' stuff with you https://rostnl.notion.site

how it all connects with your question.... i don't know, i just know that somehow it does 😄

It definitely does! 😊 Let me chew a bit on how, specifically. :thinking_face:

Actually. @U0509NKGK, are you happy with your current setup? I guess you get a lot for free by leaning on Notion rather than building your own thing from scratch. If you'd started from scratch again now, would you use Notion again? If yes/no, why? What's working well for you now? Where do you feel friction?

I presented my own setup a month ago: https://m.youtube.com/watch?v=JSMcK5strRo&t=10m16s

I'm still limping towards open sourcing a minimal viable Clojure port of my knowledge management thing. Maybe in 2023! 🙂

@U07SQTAEM I’m keeping my fingers crossed! Your work on Clerk and your comments on Clojureverse and Twitter have helped me build my own system — thanks for that. I bet people would be interested in your current workflow, even though they can’t use it yet. A possible format could be a scicloj meeting (like the youtube video I linked to above), possibly with a host asking questions to drive the conversation.

Imagine something between logseq and clerk, where you can input code or text and it is all treated equivalently and in a transactional temporal database with a variety of interfaces connected to it, repl, visualization, etc

i suspect this may be somewhat heretical but i don't 'note-keep' like those who use logseq/Roam et al, nor do i journal. i'm pretty frugal with my writings-down. this is probably why Notion and our monorepo (which is the only place i code) are enough for me 😄

i am quite focused on shepherding a shared knowledge + work space for a team though, so Simple As F is really quite important

thanks for the video share @U3X7174KS, going to watch carefully!

> Imagine something between logseq and clerk, where you can input code or text and it is all treated equivalently and in a transactional temporal database with a variety of interfaces connected to it, repl, visualization, etc > @UK0810AQ2 How do you imagine one would work with such a system? Plaintext (ala clerk), an outliner (ala logseq) or something else? I've been missing temporality and dataviz/"automatic synthesis" for my own Roam Research graph. So I've got two systems. The plaintext / html / public system, and the more messy private Roam graph.

Why limit yourself to one input method?

a REPL can have many ways to read and print, no reason to not use the appropriate one for the problem

Perhaps you could suggest some input method that treats code and text "equivalently", so we know what you mean by "equivalently". Likewise, you might expand on what you mean by "an artificial separation between code and comments" in this thread https://clojurians.slack.com/archives/C035GRLJEP8/p1671972398118759 so we have a basis for discussion.

by equivalently I mean "just data". We represent our code with data literals. We can manipulate, query and reason about it as such. Text has structure, too. Hell, HTML already does that. The problem begins where comments and docs, and in Clojure this is true for line comment, comment blocks and docstrings, are dead code. Why should a reference to a var name in a docstring ever point to a var which does not exist? I want to be able to toggle input methods based on the input type, and that everything will get stored as structured data in the end

You can write your text as hiccup and use a projectional editor to provide a different "input type" today. :man-shrugging::skin-tone-2: However, if you want the same file to remain easily editable as text, you run into questions about what syntax will indicate the text's structure (including which parts are meant to be executed). As I said in the other thread, my preferred approach to this problem is Pollen ("the book is a program"): https://docs.racket-lang.org/pollen/

Must have missed pollen in the other thread, from a cursory glance looks 👌

I'm just suffering a bit from Unison envy

@U3X7174KS Hi. I'd be interested in such a channel. Could be interesting to hear more about dbx as it evolves and I'm happy to give updates on how logseq is becoming simpler > The community around a specific tool for thought tends to accept that tool as a dependency without further consideration Not always. At logseq it's possible to make apps that live independent of the editor app. Logseq's core is nbb compatible so a node app/script can take in logseq markdown files and turn them into edn or a datascript db. Some examples at https://github.com/logseq/nbb-logseq/tree/main/examples

https://www.thalia.de/shop/home/artikeldetails/A1003277169 and https://www.adlibris.com/se/bok/the-unabridged-journals-of-sylvia-plath-9780385720250 are the two bookshops. The German one seems a bit cheaper.

Sure, could do this. The German one takes 2-3 weeks delivery, if that’s an issue.

Weeks!

Let me get back here tomorrow morning.

Thanks!

Seems I can get it in Norway as well.

delivered in 5 - 16 days.

I’d be happy to help out.

I wonder if this is a good startup opportunity.

FWIW, Amazon US also offers it but it's sold by allnewbooks and "usually ships in 6 to 7 days"

Id like to avoid customs. Brexit has cut off a good source of used English books :(

Amazon NL says delivery between 16-19 Jan

(That’s to my address in NL, btw. I was offering to buy and post 🙂 )