This is the open source "competitor" to Roam?

One way to put it @U04V70XH6

I'm curious what warrants the scare quotes.

Because I don't know enough about either project to know whether they are truly competitors or not.

Besides Roam and Athens, are there any other similar solutions out there?

Reading up on this stuff, I'm reminded of Notion. https://www.notion.so/product

Not quite the same problem space but there's definitely overlap.

A few other notables for discussion: • https://www.remnote.io/ • https://logseq.com/ • https://obsidian.md/ • https://github.com/foambubble/foam • https://dendron.so/

You can also zettelkasen with org-mode https://blog.jethro.dev/posts/zettelkasten_with_org/

Does Athens support exports?

there is also https://twitter.com/codexeditor

https://www.patreon.com/codexeditor

re Athens exports: you can have a db backup, so I guess that counts as export

I think athenresearch is pretty nice and lovely! A good showcase for another thing possible with clojurescript! well done @UMA337M3P!

Where do I find some intro docs for Athenresearch? E.g. to learn what ((block references)) is? Is there any support for formatting (e.g. markdown)? thanks!

Ok, found the Welcome page in the app 🙂

Athens was recently accepted into Y-Combinator too. It'll be neat to see how they approach open source with commercial success which is a big topic in the zeitgeist

looking to make seamless interop between datascript<>markdown. currently just datascript 🙂 @U0545PBND @UA2R84M28

support a subset of markdown right now @U0522TWDA. it's likely we go for full markdown but also many users are non-developers, in which case a WYSIWYG would make more sense for them

One thing I did not get - can I post links to external content, as I would in .md do with [my title]() ?

yes

I had made a visual to help explain zettelkastens more https://clojurians.slack.com/files/U8WFYMFRU/F01N8D0SX5F/zettelkasten.pdf (still a rough draft) and there’s some other discussion about it in https://clojurians.slack.com/archives/C03RZGPG3/p1613078474271900.

@UR0KX1MRT I use Notion for my zk but org-roam is another great option for emacs using org-mode.

I'm curious - what might be some of the use cases? I can only think of somehow losing deps.edn.

@U2FRKM4TW For babashka scripts this could be useful if you want to port your script to the JVM and you want to compile it with GraalVM for example

Ah, nice, thanks!

This can also be useful when you want to compare the newest deps in your .m2 with the ones you already have. Or when you want to migrate your git libs to mvn ones :)

I imagine it could also be useful if you use clojure.tools.deps.alpha.repl/add-libs to hot-load dependencies and you want to save your state into a deps.edn file.

This can also be a useful addition to clojure-lsp. Type (:require [clojure.tools.cli]) and it will be interactively added to deps.edn. cc @UKFSJSM38

Looks valid, not sure what's version would add in deps.edn though

oh got it, that's deps-infer responsibility, really nice

I found another nice use case: you want to port a lein project to deps.edn (but don't want to rewrite the project.clj yourself)

or a boot build.boot file (which is arguably harder because of code instead of config)

oh, that I could use on clojure-lsp itself 😆

Extra use case (and long wanted one!): making implicit dependencies explicit e.g. maybe I'm actively using math.combinatorics in my project, but it's in my classpath because it's a transitive dep of some other dep.

Another use case: see which deps you are actually NOT using in your original deps.edn

Although this can give false positives when you wanted to pin a particular transitive dependency to a particular version or include some plugin that some other dependency (or your code) uses dynamically.

if you wanted to pin that version, why not include it explicitly with that version?

yeah, false positives can happen, this is probably why you should manually review the output

> if you wanted to pin that version, why not include it explicitly with that version? That's what I meant. Your project is using lib [email protected] which uses [email protected] You know that [email protected] has a bug that's been fixed in [email protected] that's backwards compatible with [email protected]. Naturally, you include the [email protected] in your deps even though your project never uses it directly. Perhaps, deps-infer deals with that already by analyzing not only your code but also the code/dependencies of your dependencies. It's not obvious to me given the README.

@U2FRKM4TW I see, yeah, in that case I think you should put a big fat comment in your deps.edn saying why you included that dep

You can --analyze all of your own code + dependencies, but then you will get a list of suggestions for all transitive deps as well

The use case I have is, I know the api for a number of libraries well, so I can sit down and write a clojure program that uses them, and then suddenly I have make a deps.edn to run it, and I have to Google around to find maven coordinates.

Which is really dumb since the information is actually all already on my local ssd

my local m2 is over 2.5 gigs and I must have been copying it forward across 2-3 laptops because it contains artifacts like the fork of clojure we used at work briefly 8 years ago

https://clojars.org/mantle/clojure

Oh god, you made me recall how we had ~45GB .m2 dir for a legacy project. And that's on a fresh install. What a nightmare it was to work with.

yeah, so so much of the information about libraries is already there

@hiredman you would just type it without a REPL connection?

yes

I think in theory I could do that but without the REPL I just feel naked in a thorny forest