#PKM proprietary/online products: *exist*
#Obsidian users: "better keep notes local and in a standard format!"
#Logseq users: "right, but also using #FOSS tools is important for the same reasons, Obsidian is not FOSS, on the other hand there is this new app called Logseq..."
Obsidian users: "it's nice that everyone can use what suit them the best :)"
🤷🏻♂
I don't remember if I already asked you, but what's the difference between Logseq's Document Mode and Obsidian? I don't see why one would open another app when in Logseq you can just press `t,d` to hide bullet points and write paragraphs of text in blocks like in any "longform writing" app.
That’s not the only reason I use #ObsidianMD. It produces cleaner, more portable markdown documents, and I can use folders, which are to organization what plaintext is to data. Just as plaintext guarantees that I can always edit my notes in any text editor, a basic folder structure means I’ll always be able to navigate them in any file manager.
### About Markdown:
- I can write a page in Document Mode in Logseq using first level blocks as paragraphs and indented ones as lists;
- if I remember to press `t,o` shortcut to expand all collapsed blocks the `collapsed:: true property` is removed;
- if I don't reference any of those blocks I am free from `id::` properties too;
- I can remove the first two columns of characters from the file to turn first level blocks into paragraphs and get more "standard" Markdown files with a simple Unix command;
- I can concatenate that command to Pandoc that I use to export to PDF so I don't need to actually edit the MD files;
This is more complex to explain than practicing, so for me there is not a reason to split data and workflows between two apps that would be a huge disadvantage.
### About folders:
- Logseq frees me from having to choose a hierarchy, to stick to it and to decide where to place something;
- I can have as many hierarchies as I want to organize pages (and blocks and whatever) and the same element can be in multiple hierarchies;
- I can still organize MD files in folders independently from Logseq that will just respect my folder graph organization; this is useful if there is a set of pages I want to share between graphs (I can symlink a folder) or share a specific folder with other people etc.
Just to be clear mine was a joke about consistency when promoting the product you like
@post Except that one is about your days and one is about the source code of your app.
With #ObsidianMD, your *data* is saved in a cleaner, more portable—and arguably more open—format, and with #Logseq, the source code of your *app* is open.
It’s not inconsistent to place more preferential weight on one than the other in choosing which to use and for what purpose.
Being FOSS and not being FOSS is an objective on/off, while portability of data is a subjective spectrum.
In particular, with Obsidian you introduce special syntax using plugins like Dataview. Logseq has built-in queries using Datalog.
**What's more standard and portable, Dataview syntax or Datalog?**
Also I already said Logseq saves data as Markdown with additional optional syntax that you would add with Obsidian plugins anyway.
I can easily turn the indented lists by Logseq into paragraphs if I want, but those indented lists are still standard Markdown.
About block properties, it is trivial to remove them automatically.
For block reference, again, it is trivial to search in the whole folder for the ID of a block, it can be done manually by text editors or automatically through code.
There is even a LSP (Language Server Protocol) for Logseq syntax that adds functionalities to IDEs that support LSP like previewing a block reference when hovering it with the cursor, like it would happen if Logseq syntax was a programming language. I know Obsidian has one too but the point here is how easy it could be to recover data and UX from a special syntax.
@post Plaintext files with a minimum of app-specific formatting *are* objectively more portable and universally readable. And while the characteristics that make software FOSS or not may be objective, the extent to which anyone personally values those characteristics is subjective.
The reason professional photographers overwhelmingly pay for Photoshop instead of using GIMP isn’t because they’re ignorant or stupid. Photoshop has qualities they care more about than GIMP being FOSS.
Good point, valuing FOSS is subjective but since Obsidian users advocate for digital sovereignty I expected FOSS to resonate with them and even if they won't move away from Obsidian just for this reason I expected to see more appreciation toward FOSS in general and admission that being closed source is a weakness for Obsidian.
Instead Gimp is not an alternative to Photoshop, Gimp is just an image editor while as you said Photoshop is a product for professional photographers and not only.
On the other hand Krita is way better than Photoshop for drawing since it is specialized for that.
Just mentioning one area where FOSS is weak is not fair though. Remaining in creative arts sector we have Blender that can compete with proprietary counterparts and with the recent release of Godot 4.0 we have a professional FOSS game engine.
These are huge complex FOSS projects. Nothing prevents us from having the same but opposed to Photoshop, it just happens not to be the case for now.
Also notice that many of us already can work with 100% FOSS, while it's impossible to go 100% proprietary.
For example there is no proprietary Web engine to my knowledge; proprietary Web browsers would still be based on Blink or WebKit.
@post I agree that there are great open-source apps and I often use them. I’m writing this in an open-source Mastodon client. Open source has real benefits. I just value the freedom to use whatever software I find most useful and enjoyable over the freedom I’d have to modify or fork the source code if I were a developer.
Btw, Krita is better compared to Illustrator than to Photoshop.
Are you sure about Krita? It's more for raster graphics like Photoshop than for vector graphics like Illustrator
@post You’re right, I’m thinking of Inkscape.
@post does FOSS software produce better plaintext markdown files?
Is this supposed to be a provocation? 🧐
Just in case you don't know: Logseq stores data in Markdown files just like Obsidian but they are standard Markdown indented lists using dashes.
To some extent one can even use Logseq and Obsidian with the same folder.
It's up to the user to avoid special syntax and this is true both in Logseq and Obsidian (the latter's strenght is the ecosystem of plugins that often adds their own syntax).
Also notice that `#hashtags`,`[[wikilinks]]`, and YAML headers to store metadata are not standard Markdown.
@post I’m not sure what we’re arguing about? I know what Logseq is.
Then maybe you don't what FOSS is?
@post oh are we arguing about if it matters id your markdown editor is completely open source?
No, not really for any reason I can think of. But if that’s super important to you.
Maybe we both know what all these things are and disagree about how much it matters to our actual work or the usefulness of the product.
It doesn’t. Semantically, whatever - do what makes you happy. It’s not a risk to your plaintext markdown output either way.
Sed, awk, and grep are OSS, I’m pretty sure
@post it would take 5 minutes with standard *nix tools to reformat and strip out or convert tags between markdown syntax recursively in a vault.
Obviously tags from obsidian plugins aren’t going to render in MarsEdit.
If you mean Unix tools, those are for plain text, not for Markdown that has tables, indented lists etc.
On the other hand #Logseq not only provides handy HTTP/JSON API, but you can also access the AST with command line tools (using Babashka) that basically means the files are already parsed for your convenience and you can use whatever other tool including the Unix ones to interact with it.
Here there is an example of command line tool (it doesn't depend on Logseq being running nor being installed):
@post I'm talking about sed, awk, grep, tr... etc.
And yes, for markdown that has anything because markdown IS plaintext - even the bullet points.
You're just wrong about that. Markdown is plain text, even the bits that aren't english words. It's not compiled, it's not binary - it's ascii text and it can be manipulated with standard tools if you know what you're doing.
But I get it, you like logseq. That's cool.
OK let's see your Bash script where you reorder the elements of a Markdown table in alphabetical order using sed, awk, grep etc.
It seems to me that it is you who like Obsidian too much to admit FOSS is as much important as standard formats.
Also people who like Unix tools should be happy to see what you can do with Babashka and Logseq's AST.
@post sure. Strip the delimiters, split the fields into an array - sort it, and then print it back with the delimiters.
It’s ok. You can like logseq.
I have nothing against it.
You are the one seriously replying to a joke 🤷🏻♂
You didn't tell me how you located the target table in a MD file that contains something else and how actually replace it with the new one. It's a program on its own, not just some piped Unix commands.
It would have been better if you said "there Markdown editors other than Obsidian and even Markdown command line tools and libraries".
The fact that you mentioned Unix tools for plain text files suggest that you are talking in a theoretical way with no actual practical experience with parsing Markdown. Maybe you don't even know what an AST is.
No offense, but you're taking the stereotype of my joke to a new level.
@post dude. This is a stupid argument.
Can I format a markdown table and sort it in bash, yes absolutely. I worked as a linux engineer for 15 years doing stuff just like that.
I’d use python myself, but absolutely could be bash… in probably < 10 lines.
But the reality is that I would never put a table in markdown in the first place. I’d use a database like a grownup.
I don’t actually give a shit about your work or what tools you use to do it.
Please address whatever issue you have before joining a social platform, bye.
@post I’m *really* liking Logseq, but for different things than I use Obsidian for.
I’m finding Logseq ideal for dealing with the messiness of daily life—for things I need to remember and refer to now and in the coming days but won’t necessarily need long-term, which is different from the way I use Obsidian.
Rather than alternatives, I now see them as complementary tools with different though overlapping capabilities—more like Photoshop and Illustrator than Photoshop and Gimp.