Anytype – local-first, P2P Notion alternative | Hacker News Reader

Anytype – local-first, P2P Notion alternative

anytype.io

414 pointsTTTZ273 comments

(I work at Notion)

The Anytype model is really cool - in a way, they’ve rebuilt Lotus Notes with 21st century E2E encrypted protocols and technology. They’ve built a really solid personal knowledge app with many of Notion’s features - and some clear improvements over Notion.

However they also demonstrate the complexity and tradeoffs of the E2E approach. Anytype has been a work-in-progress since at least 2019. Their docs still state:

> Future versions will allow you to share your work and safely collaborate with others

> There's no browser version of the app. Anytype is a stand-alone software, that works on desktop or mobile devices. There are many points of vulnerability in-browser apps that would compromise our commitment to data security and encryption.

Without these features Anytype is in a much smaller market (PKM) with less distribution than Notion/Coda/Dropbox Paper/Quip/Confluence/…

It is so interesting that you are mentioning Lotus Notes because that is exactly what I was thinking when I was setting up our Confluence pages: it is exactly like Lotus Notes' Knowledge Base template, with some additional funtionality. Crazy how we keep reinventing things.

By the way, and this is just my point of view: we moved from Notion to Confluence because Notion has no free tier for small teams (4 people). We were working on a project for a not-for profit org and their money is always tight.

Since you already give Notion free, have you ever evaluating doing like Atlassian that up to 10 (or 5) people the thing is free?

Google Docs have somehow solved E2E encryption with safe sharing and collaboration. They haven't advertised it much, just a mention here and there as an enterprise requirement or something, it's good how they've implemented it on top of an existing web-first collaboration product.

They do have technical challenges implementing add-on services on top of content - For example, the grammar checker or any content assistance is disabled in E2E encrypted docs, which makes sense.

Surprisingly they support spellcheck: I guess they are shipping dictionaries to browsers and use a local lookup. (Google Docs doesn't rely on browser's dictionary because their editor is not based on `contenteditable`)

Operating systems should be providing spellcheck to textboxes everywhere anyway.

Makes no sense every app rolling their own and not using my operating system dictionary with my own corrections.

Extremely frustrating every Electron app just rolls their own rubbish spellcheck, or doesn't have it at all.

That's an interesting take on things. Would you want your operating system to provide a web browser, pdf viewer, and spreadsheet as well?

To me it makes no sense that software does not interoperate very well. I think most of this is due to the historical evolution of computers and software. A spell checker used to take up an insane amount of memory, a browser is a competitive product, and interfaces for proper interaction have to be standardized worldwide, for which we still lack some kind of governmental body.

It's kinda-sorta possible on Linux distros, you know, with cliphist and all these little programs, but you have to do a lot of custom stitching, adding Firefox extensions to interoperate in such-and-such ways, and most programs don't have extensions. Until mainstream OSes make it a priority, I'll stay in my Emacs corner.

It's amazing how insanely slow OSes can be too install, even without providing some of these things. I feel like getting a new Debian docker image to apt update takes forever now compared to 1995.

I mean, `apt update` is probably more tied to the sheer number of packages (and sometimes repos) that it has to reload in its database than anything about what's actually installed. Maybe also a function of what's installed to check for upgradable packages and check dependency trees, but that still seems like its own problem.

> web browser, pdf viewer, and spreadsheet as well?

Yes, yes, and no.

>Would you want your operating system to provide a web browser, pdf viewer, and spreadsheet

I mean I have little use for an OS without a web browser so yes. PDF viewer, Preview.app is infinitely better than Acrobat Reader so yes again. Spreadsheet I can give or take.

The idea of multiple teams all rolling spellcheckers to build upon the same input boxes is absurd to me. Should be as integrated into the machine as the cursor and caret imho.

Well, neither Google’s cursor or caret are drawn by the system. They want ultimate control, not ultimate integration with the host browser/OS.

Most electron and web editors that use ContentEditable should be able to use the OS spellcheck. Notion uses OS spell checking on web & Mac electron. I think on Windows there’s some weird issue we had to work around but I think we resumed using the OS dictionary, I’m not up to date on that.

Why would you compare to Acrobat instead of Skim?

Yeah, I suspect the grandparent comment meant that spellcheck (and perhaps other tools, like you listed) should be standardized. And thanks to the big cos (Apple and Microsoft) doing full vertical integration of their stack and calling it their “operating system”, that term is increasingly getting used for something more than just “the thing that operates your computational resources and safely exposes their APIs for developer use”.

That said, it would be interesting to see more actual OS development that innovated on supporting/encouraging more standards, without getting too proprietary or centralized.

I wouldn't want the OS to do spell checking, but the OS having a config setting for which spell checker library and settings to use by default and for specific apps would be sweet, would it not?

> A spell checker used to take up an insane amount of memory

IMO they still do. At least anecdotally, with free software shipping with spell check that often makes up a hefty chunk of the file size.

And even ignoring that, for every program with its own spell check, you have all of the redundant spell check data, and then your corrections/additions aren't even shared. Nevermind importing them on a freshly installed OS, syncing them between devices, futuristic things like that :)

I think this is the right solution: Have the API standardized, and perhaps even provide a default implementation out of the box, but fully support replacing it. Just like how, say, most systems ship an image viewer out of the box, but you can replace it with something else because the system has a standard file association system that a replacement can slot into.

Spell check is a universal composable text editing property, great job for the OS None of the other format viewer apps are

I've always thought that textboxes should have some sort of browser-to-os plugin feature, possibly on-demand.

I would love to edit my textboxes using real emacs somehow, even a popup window would be good.

I worked on spell checking for a web app. It would have been really, really nice if there was just an API to interact with the browser spell checker. We couldn't use the spell check you get in a text area, because like google docs, we rendered the text ourselves and had a custom context menu. And we had very good reasons for doing those things.

However, I think it would be difficult to create such an API without leaking information about what words are in a users custom dictionary (which could be used for fingerprinting). And allowing an app to add custom words would probably have some security implications.

Maybe if there was an API to open a spell check popup that was native to the browser and the only info you get is what word (if any) they chose to replace. But, besides having a suboptimal UX, that doesn't help you know which words to highlight as misspelled.

Did they? https://support.google.com/docs/answer/10519035?sjid=6574250...

> Multiple collaborators can work on the same file at different times, but you can’t collaborate in real-time in encrypted files

> Without these features Anytype is in a much smaller market (PKM) with less distribution than Notion/Coda/Dropbox Paper/Quip/Confluence/…

I can tell you that e.g. the whole education market in the EU is waiting for something like AnyType, as soon as they have implemented collaboration, which is very high on their roadmap. As a teacher I just can‘t use Notion or any other US-service based on US cloud services. All claims concerning GDPR are nonsense due to Privacy Shield. I am not allowed to use such services.

Therefore I hope AnyType copies as many good Notion features so that I can leave Notion and its abysmal performance and lack of offline mode for private stuff as well.

Why not use some EU-based product that hosts on EU servers?

That's like trying to find a unicorn on Mars, at least if your definition of unicorn goes beyond a frozen donkey with a mop fastened to its brow...

How about https://fibery.io ?

So, yeah, unless Notion becomes Free Software and actually secure soon, there's the Notion killer.

It would be interesting what you can say about http://fibery.io ?

I tried Fibery when it came out and clicked around some of the newer demos linked from the homepage a few weeks ago, so I’m not super knowledgeable, although I do read the email newsletter sometimes.

I really like Fibery’s plainspoken marketing, but it’s not for everyone. They’ve toned it down a bit on the homepage to be more “professional” now which is a good sign.

I think Fibery’s UX is too complicated for most people. As a programmer I get all the data modeling stuff, but there’s still a lot of complexity in just the UI that I find surprising. Notion gets a lot of feedback that it’s “overwhelming” and I feel the same in Fibery. My impression is that the up-front cost to figure things out in Fibery is still a challenging barrier for the product.

I’m not sure how good Fibery is for writing or on mobile, or if it has any local cache for offline use.

I’m really happy that everyone at Fibery made it through the Ukraine war okay so far.

This is awesome! Love to see more local-first, security-focused apps! Especially ones that are free.

Asking because I'm currently working on a local-first app that I am gaming out how to sync to non-local networks: The docs state that the app syncs to local P2P networks. Does that mean it doesn't sync across the internet? If so, is that because of the cost of maintaining/renting a TURN server? Or is there a technical limitation? If it does sync across the open internet, are there restrictions to that?

Sorry for the grilling! It's just an area of interest at the right time. In trying to come up with a WebRTC solution, I think I've settled on defeat. The local network syncing works like a charm, but I haven't had any luck in trying to get around a TURN server by using an API endpoint to provide routing data. It would be nice if there were other prebuilts than COTURN (something deployable to a deno/node server would be ideal), but with what we currently have, I think I've settled on just pushing the encrypted data through my api server, letting users encrypt and decrypt with well-known hashing, based on some key they provide in each client (exporting client and importing client; they never exchange keys).

Anyway, my pet projects aside, this thing looks really full-featured! Great work here. I use Notion, daily, and this could easily replace it for me. In fairness, I don't take full advantage of Notion, but I doubt most people do either. Seems like you've covered enough of the bases to provide a real product for all but some edge cases, so that's pretty awesome! I'll echo others in lamenting a web app; that's what will keep me from switching from Notion, because its just not portable enough for the work I do. But, otherwise, I'm happy with it so far. Gonna try to plan some stuff in it and see if there are any creaky wheels!

> Does that mean it doesn't sync across the internet?

They currently host their own backup node, so it does sync across the internet, but you only get 1GB of storage (or 10GB if you were in the alpha) on their backup node. Once that's exceeded, additional files are only synced P2P.

They're planning on providing extra storage for a cost and soon anyone will be able to self-host their own backup node.

Right; should have been more specific, thanks!

I meant to ask if the app syncs P2P over the internet, rather than just through their provided storage.

My solution is to have short-lived data so that they can't reach more than a few MBs of memory on my server, but if there's a straightforward way to do P2P across the globe, I'd very much prefer that option. Asking people to believe that I can't look at their (encrypted) data is fine, but preventing me from having their data at all is ideal.

Given what you pointed out, I definitely figure they're not doing WWW P2P syncing, but you'd be surprised the technical stuff you can figure out by just asking someone who spent way too long building it. They're usually happy to tell you all kinds of interesting details (especially when it comes to 'interesting' bugs/workarounds/hacks)!

P2P sync over the internet on our roadmap. We plan to release it later this year. Our protocol already supports it, but we need to implement peer discovery.

Looking forward to seeing if it syncs across ZeroTier networks. They emulate a LAN including multicast so it might work out of the box.

> In trying to come up with a WebRTC solution, I think I've settled on defeat. The local network syncing works like a charm, but I haven't had any luck in trying to get around a TURN server by using an API endpoint to provide routing data. It would be nice if there were other prebuilts than COTURN (something deployable to a deno/node server would be ideal),

The bad news is that if you want something that works in all instances, you need a relay of _some_ sort, because p2p isn't possible/feasible in all cases. Bittorrent (for instance) works around that limitation by simply having many-to-many peering and relying on large numbers, to be reliable. But that doesn't work for 1:1.

The good news is that maintaining a relay (or TURN, with WebRTC), need not be expensive. Yes, you need a server, and perhaps some IP-based rate limiting, but that can handle a LOT of connections and small data.

I created https://github.com/betamos/rdv for this purpose, an extremely light-weight alternative to WebRTC, but for TCP only (BYO identity, auth and encryption). The p2p success rate is, anecdotally, very high. However, you cannot use it from a web browser.

Feel free to reach out (see profile), happy to chat about p2p whether or not you use this project.

Ah, nice! That looks like a method I had tried, for sure. My first setup was to use .NET core to build an API that registered users with session variables and broadcast that session to new users. It required a "host" because they initiated the session and then all other users were given the host's address to negotiate their P2P connection.

That all worked fine but it required a token api, and then having to choose between "polling" and "websockets"; not a particularly appealing choice for me. So TURN servers seemed to be like the ticket, but I just couldn't find someone to say "this is what data a TURN server receives; this is how it determines its response; this is what the response data looks like". Without that, I was having a very hard time building my own TURN server. The only option seemed to be "use COTURN", which is not something I'm interested in. Not because I think it doesn't work, but because I haven't ripped it apart to know how to fix it if something goes wrong. As this is a side-project, I have the luxury of forcing myself to deeply learn each part of the data flow, so it doesn't make much sense to offload something I don't totally understand to someone's library (conversely, I've written an IndexedDB handler, so I use the Dexie library because if it completely falls apart in novel ways, I can still troubleshoot it). So I figured I would wait (it's been a couple of years now), and if no one got around to dumbing it down, I would start inspecting the data flow, myself, and try to replicate whatever COTURN is doing.

So if you can shed any light on how to make a TURN server, I am ALL ears! I do think WebRTC is the way to go so, while I appreciate a suggestion for an alternative (and one that you personally use is great, but one that you personally build AND use is pretty much the gold standards for reccs, so I really do appreciate it!), I don't think it's helpful for my intentions. My project is also web-based (first), so while I'm not averse to a wasm library implementation of TCP connections (mad stuff recently with the curl support!), it's a bridge too far, to start with. In the mean time - like you said - I'm not concerned about the cost of a TURN server. I know that most apps can get by on free-tier cloud hosting limits, when you're talking about those little chirps of data. So I'm happy to build and pay for what it takes, I just honestly don't know how to do it.

> [...] I don't think it's helpful for my intentions. My project is also web-based (first)

Yes, then it's a no-go. No excuses needed.

WebRTC is the only way to reasonably do p2p on the web. This is because of the lack of "raw" sockets and their options (just regular TCP/UDP with SO_REUSEPORT).

> So if you can shed any light on how to make a TURN server, I am ALL ears!

You'd know better in a few minutes of reading the docs, sorry. I will say, pay attention to the requirements - for instance, listening on arbitrary ports, getting the "real" port from the client, even stateful connections etc, aren't always possible on cloud services (the more "managed" the marketing, the more restricted).

> Not because I think it doesn't work, but because I haven't ripped it apart to know how to fix it if something goes wrong. As this is a side-project, I have the luxury of forcing myself to deeply learn each part of the data flow, so it doesn't make much sense to offload something I don't totally understand to someone's library

I empathize deeply with this. In fact, that's the main reason why I built rdv. WebRTC in particular is, imo, a nasty piece of over-engineering. It basically throws together a bunch of separate "protocols" into one. It has way, way too many knobs, and many things are stateful. If you do go down the route of reimplementing even a subset of WebRTC I recommend getting an insurance for your keyboard, one that covers physical violence, (but if you do, I would be very curious about your learnings).

Finally, my words of wisdom would be to avoid p2p if you have the luxury of having a small amount of messages. Just running a simple "relay" of encrypted messages is reliable and cheap. Users are still protected by e2ee.

Why does this smell like a cult? Talking about trust and having illusive pictures and designs..

And why do I need to login for an offline app? Where does the data really go? There is some setting remote storage, and some mb already used, so is in reallity an online-app with some offline-mirroring? But it seems there is not a choice to set the local storage-path, and the saved files are just your usual electron-blob-files. Seems there is no way to really access or control my files externally.

This doesn't give much impression to be trustable.

I think the same. I was trying to find the "How to install the server" but here is no option. It just saves the data in their servers. There is too many wording about security and control, but only in their control.

Here is the link with the instructions https://tech.anytype.io/how-to/self-hosting

That's useless if the client also sends the same data to Any's node. You can't disable that for now:

  Self-hosting only without a backup node is currently unavailable

  The self-hosting-only feature is not released yet. If you don’t want to use Anytype node and instead use p2p sync between your devices, you can block Anytype network traffic (Anytype & Anytype Helper) via your firewall.

From https://doc.anytype.io/d/data-and-security/data-storage-and-...

The idea that you can compile apps with your network configuration, so it's not easy, but it works.

Thank you for the heads-up, we released the self-hosting ability yesterday and didn't update the docs. Now it's been updated; self-hosting is available: https://tech.anytype.io/how-to/self-hosting

Been following/trying Anytype since early Alpha. Congrats on this big shift. My biggest concern was that there was never really enough to promote/allow self-hosting up to this point. Excited to see where things go from here!!!

Thanks! A Docker image is sorely missed.

> Talking about trust and having illusive pictures and designs..

it's one of the key features of the app that most other alternatives lack, it would be weird for them not to mention this

> And why do I need to login for an offline app?

it's not offline only, it can sync to your other devices over the internet

> Where does the data really go? There is some setting remote storage, and some mb already used, so is in reallity an online-app with some offline-mirroring?

it's stays on your devices, unless you pay for dedicated backups of your data

> and the saved files are just your usual electron-blob-files. Seems there is no way to really access or control my files externally.

your data is encrypted, which is why you can't just access it via your file manager

hope this helps :)

> it's one of the key features of the app that most other alternatives lack,

That's their claim, but it doesn't hold up much so far.

> it would be weird for them not to mention this

There is a difference between mentioning it, and using it as a selling-point.

> it's not offline only, it can sync to your other devices over the internet

There are many apps who can sync, most of them have it as an option, without a forced login and hidden sync.

> your data is encrypted, which is why you can't just access it via your file manager

So I can't even access my data and must trust in the app to not screw up, and have a well working export. Which considering it takes Notion and its poor export as inspiration, is not really a good sign

> hope this helps :)

Yes, thanks.

>That's their claim, but it doesn't hold up much so far.

are there any arguments?

>There are many apps who can sync, most of them have it as an option, without a forced login and hidden sync.

We don’t have any forced login. It’s just basic keychain that generated on your machine, that only can use, and can use off-line. It’s a way do deal with encryption. Where you see a forced login?

>So I can't even access my data and must trust in the app to not screw up, and have a well working export. Which considering it takes Notion and its poor export as inspiration, is not really a good sign

we have described our data format, so your data is yours, you can write any adaptor to read/write

> We don’t have any forced login.

Then how can I use the app without creating first an account and space which will sync my data without even asking? You talk about trust and offline-first, but the first thing you do is uploading data and don't even talk about it.

> we have described our data format,

Where?

BTW Don't reinvent menus. I just now realized that File->Close does not actually close the app, like in any other app. And for whatever reason, there was no information about the App still running. Kinda sus..

> Then how can I use the app without creating first an account and space which will sync my data without even asking? You talk about trust and offline-first, but the first thing you do is uploading data and don't even talk about it.

It's not an account you create in a classical way using a server. It's locally generated data wallet ( we need it for encryption and sync). We talk via on-boarding that your wallet will be synced, and that your data is encrypted only with the key generated on your machine. It's for people who don't want to mess with self-hosting. If you want you can make your build from GitHub, disconnect sync, or just self-host yours. This way our promise works.

> we have described our data format

we have a GitHub! here it is https://github.com/anyproto/any-block

> BTW Don't reinvent menus. I just now realized that File->Close does not actually close the app, like in any other app. And for whatever reason, there was no information about the App still running. Kinda sus..

good point!

They seem to have only one "node" so far, because the software is still in alpha? That's what their docs seem to say[1].

  All your data primarily syncs to the encrypted backup node in the current alpha. For alpha testers, the application is always connected to the backup node and cannot be disconnected.

Their homepage could tell this more clearly. It doesn't look like it's in alpha.

[1] https://doc.anytype.io/d/troubleshooting/self-host-your-back...

We are in open beta now. We have a network of nodes not one node.

You should update your docs because this is not what they say. I seem to understand you just started the open beta -- congratulations. You should still mention your product is in beta on your homepage (or at the very least, your download page).

EDIT: How can I see which nodes are in the network? There's only one ID on https://networks.any.coop/ and I don't know what to do with it.

Thank you for the heads-up, we will review the relevance of all the docs.

Regarding the list of nodes: we don't show it in the interface now, but will add more network information in the coming releases with simplified self-hosting. For now, you can see the list of nodes in the code: https://github.com/anyproto/anytype-heart/blob/release/core/...

If you want to dive into the details, information about the underlying protocol and configuration formats can be found there: https://tech.anytype.io

The proportion of philosophy to features in their marketing seems off. They say "trust our code, not our words" and then spend several paragraphs of words talking up their ideals. It feels like something is being hidden or talked around.

Data-independence may be one aspect that's not transparent. Compared to https://obsidian.md/, I can't just use raw Markdown or CSV file with Anytype. I've paid Obsidian $300+ just because it's data independence with a few obvious strings attached. By the same token, if I pursued a local-version of Notion for my team, I would also want the same freedom that local, open-standard files offer.

Anytype's functionality cannot operate using Markdown or CSV, as these formats are limited for our specific use-cases. Instead, we've defined our data format in a protobuf file, which is open and MIT-licensed. You can always compile a version yourself and use your data without relying on Anytype. We hope to include data adaptors in future developments

Markdown sucks in some ways though. Not being able to merge cells in tables? Come on.

To be fair, you can compile the code (including the protobuf format) and implement your own export to any format. It's a bit obtuse, but I had a obsidian vault full of markdown files that are also completely useless as they have a bunch metadata that needs the obsidian plugin to render. This is not a critique of obsidian, but there is a limit of markdown and obsidian is bypassing it using a lot of extra code. If obsidian got deleted from the internet, I would have my markdown files but a lot of the processing would be gone.

If we were to store markdowns only with linking and no advanced tables, do we need any advanced app on top?

Obsidian as I recall, is closed source, and they give some weird non reasons for that, so it’s not radical to assume they’re doing something ‘not in the open’, so not a good contender when talking about owning your data and data independence!

Moreover, the statements just meant you don’t have to take our ‘words’ for it, here’s the proof. And in a different context, developers explained their motivations and what they’re inspired by, to work on this kind of software. There’s absolutely no over the top ‘marketing’ to any type as well, as they explained in different post, they don’t want to market, buy built in customer base with VC money and so on.

I don't understand this license. It's not open-source.

https://github.com/anyproto/anytype-ts/blob/main/LICENSE.md

Co-founder of anytype is here. The most of are repos are open-source. We have our philosophy regarding open source here https://blog.anytype.io/our-open-philosophy/ Happy to discuss concerns regarding our approach.

Yeah no, that's not how open-source works. Your license is effectively a "source available" license and not officially approved by the OSI: https://opensource.org/licenses/

I don't see your license on this list -- https://opensource.org/licenses/

Please don't advertise anytype as open-source.

This is not an open source license:

https://opensource.org/osd/

6. No Discrimination Against Fields of Endeavor

The license must not restrict anyone from making use of the program in a specific field of endeavor. For example, it may not restrict the program from being used in a business, or from being used for genetic research.

He literally voiced a concern and you just completed ignored it by linking a BS corporate speak page.

You are available on Github which doesn't = open-source

Also why only most and not all?

As it’s explained in blog, many open sourced projects are used for commercial gain without doing ’copy left’, the license created here essentially covers those concerns and ensures any forks do ‘copy left’ their changes with all the freedom as in ‘free’ of open source software. I don’t see the problem with this

All the kerfuffle related to ‘official’ definitions, did it help open source projects from being copied and commercialised mercilessly by closed source software (essentially legalised spyware) of corporations? I never understand this obsession.

Software should be open sourced and it should have strict copy left clause!

Relevant idea related to problems with copy left explained here https://www.osnews.com/story/25469/richard-stallman-was-righ...

I've been using this heavily, it's a good tool. Some of how it works is not intuitive coming from Notion but the docs are good enough that I got up to speed promptly. Anytype feels joyful to use in a similar way to Notion but without weight of Notion's mountain of features. It's got the right amount of focused functionality. There's some QoL issues present which annoy me, but they're nitpicks at best.

Some things I would like to see would be alternate card previews for objects, and perhaps some options on organising/separating disconnected graphs, but I would be perfectly happy without, the functionality present is wonderfully spartan and I think I'd weep if it became a feature behemoth like Notion.

This is really nice, but I think the perennial curse of technology people is producing "<app>, but with <esoteric principle>"

People want a thing, not an ideal, and they'll sell their soul to get the thing. At this point, it's actually being surprised by this that is the ludicrous position. Philosophies are a differentiator for a very small portion of population.

If you want them to buy into ideals, you've taken on two problems: Make a thing they want with its own differentiators outside of the noble pursuit, _and_ trojan in your principles.

Normally I would agree with this.

But with Notion in particular, and how heavily they're pushing Notion AI, I could see some c-level exec types being uncomfortable with the idea of Notion (or any managed/hosted service) to store trade secrets, company IP, etc.

When a company is pushing AI features as heavily as Notion is, it raises questions how Notion is planning to (or already using) your private data for fine tuning, model training, potentially scooping up your company's IP in the process.

Those who don’t think it matters already have a choice of great products. We try to build a better alternative to those who think it does. We also think it’s our job to make the ux great, to make sn analogy we think a healthy and tasty food is a better value proposition which for a network is very important.

This is true, and I sometimes wonder if the only true solution is regulation or overhauling IP or such. But I don't see that happening, because proprietary software is too large and frankly libre software isn't asshole-proof enough for capitalism.

My preference would be using this tendency the other way. Make it so good they can't resist it, even if they accidentally eat their vegetables along the way.

You're on a forum for hackers, so most of the people here are the ones that care about the specifics of technology.

> You're on a forum for hackers

I wish. "Hacker" news has nothing to do with that.

Highly recommend using the search box, New, etc.

> People want a thing, not an ideal...

So, welcome to Hacker News, where the particular esoteric principles discussed here are critically important for many of us. This is very much a selling point in this market.

(co-founder of anytype)

we agree, we have an internal mantra "people use products, not protocols". This does not diminish the importance of philosophies. The way we design our product and protocols is based on principles, because of the role software started to play in our lives - the second order consequences of the architectural choices lead to the results we get in our social life. We are believers of fundamental digital freedoms (privacy, ability to connect with those we trust) and importance of user and creator autonomy from the software provider (these freedoms to be governed by us not by software companies). We used these principles to guide our architectural decisions.

At the same time, we fully understand that if we want to build something meaningful we need to do the hardest part. Turn our ideals into the UX that would be attractive on its own. We are focusing on that. Hope to show that the p2p protocols can turned into a product that is fun to use. We are just making our baby steps towards this (not there yet)

Painful to import things from Logseq or Obsidian. It imports one md at a time, and the 'zip' thing does not click me.

It seem to focus too much on individual collections and pages and investing time upfront thinking about how to organize and navigate.

Awesome replacement for some things I had in OneNote, thou. Like an easy personal wiki.

But logseq (and as a second choice obsidian) is taking my heart because the activation energy cost is really low to just start writing notes, the flow seems more natural.

It's really an alternative to notion and maybe onenote, but not to whom just want to write more freely and easily finding things later then reorganizing what it's still important with some plugins and keyboard shortcuts (auto/semiauto moving blocks from one md to another).

For instance, daily notes: I don't want to put a title. Just give me the current date, and ability to see previous days in a few keystrokes, some tags for the blocks I'm taking note, that will help me find the related blocks and pages later.

The P2P sync is what made me try the first place. It's working fine, nice!

I did a similar thing with logseq + syncthing, and all my notes are plain files. I can't find the format anytype is using. The local working folder just looks like a standard chrome/electron folder with everything inside. Unusable for my own backup purposes.

Nice to know about in general, ty!

(co-founder of anytype) thank you for this feedback! Totally agree, that import sucks - very few options. We plan to improve the import and also publish an API to engage our community of contributors to help us build and improve more of them. on titles of daily notes - again, well spotted - again, we have plans to improve :) happy you liked p2p sync - this was the main objective and the next big thing here is multi-player based on p2p

Logseq's journal is the biggest productivity hack we got that we never thought we needed.

Don't know where to put the idea, note or just a scribble? You have a 2022-05-12 page without the cognitive load or confusion.

Let's be cautious where we attribute credit. Daily journaling has been a thing for, oh, centuries.

At the very least Zim has had the same journal feature dating back over a decade. I'm sure others can't point to similar software stretching back at least another decade before that.

I haven't used Zim, but from a cursory look it doesn't seem like it has any way to query its daily journal entries. In Logseq your daily journal might contain a dozen todos, bulleted project notes, and random thoughts all mixed together, but you don't need to look at the day's entry to find them again. Instead you run a query against all your notes to collect stuff where it's relevant.

I haven‘t tried it yet, but damnit, you guys really touched a nerve with the aesthetics of the landing page and the references to personal computer pioneers. It gives me OpenDoc vibes. I‘m so easily manipulated.

It was motivating a few years ago, when I started to see all these references to Engelbart and better thinking and whatnot, but after testing a bunch of projects that amount to nothing, get abandoned or in a state of eternal coming-soon walled garden building, where even some of them already closed (Roam Research, Athens) it got tiring.

And I still don't have that promised knowledge-tool-of-the-future in my hands, just some hype with note graphs, prolix bloggers and their 'creative mentation', yet none of the implementations (not the concepts themselves) have proven to be any sort of breakthrough, or usable enough, where commercial last-generation notetaking and journaling tools managed better at a certain point.

It's time to get real, folks! If they can't deliver on their promises, they don't deserve our attention. I'll stick with what actually works, even if it's not as flashy. Rant over.

Everything about this smells like "we're going to say we're open-source and that you're always fully in control" and then you read the fine print and find out there's some way to rugpull or some other barrier that will screw the user.

At this stage, having seen so many of these things, I'm comfortable disbelieving them outright WITHOUT reading further.

My feeling to: feels like a free sample and then they will lock you in

(co-founder of anytype) the whole purpose of anytype is user autonomy from software provider. The data standard is open. The code is open. You can self-host your network of backup nodes (tech skills required as this is self-hosting alpha released this week). The app is local first with p2p sync (you remove the back up node, you still have your data and still have sync in local networks). Most importantly you control the keys - encryption keys, access to your account - this is not touched by anytype - noone can block you out of your account, no central registry of users, all these. There are two principles for our business model: self-sustinability - support a community of contributors to improve the product and build this positive fly wheel and universal accessibility - anyone can use anytype for free if they use their resources. I hope all of this decisions deserve attention and reveal our intention (they are not accidental things)

If you care to answer then -- what is the thing that you provide that I must pay for? As in, if I don't give you money, then I don't get ....

You can use anytype for free if you use your own resources (storage, backup). There is an option to selfhost, or just use locally with local sync. You can also become a member of our association and it’s membership based, it will give you backup (if you don’t want to handle it yourself) and some things on top for those who want their spaces to be discoverable in our network. If you use privately for your team/org you may not be interested in paying. That’s ok with us :) joining the association and paying is voluntary. We have a principal of universal accessibility

Then, I guess, the problem is, that it sounds too good to be true

The project seems to come from the crypto-bubble. Bitcoin and DAO are mentioned here and there. So the smell either comes from them being a bit naive and having a poor socializing matching with HN.. or it's a new style of crypto-scam.

I’ve been on the waitlist to try this for close to 5 years and never once received an email. Now, it is available for everyone to try…so I just installed it. Looks nice enough, but it is too similar to Notion for my taste and since I don’t like using Notion I don’t think I would like using this. I’ll stick with Obsidian because I value the flexibility it gives me (running my own code, customizing everything, etc.)

This is how the email system works. We invited everyone from our waitlist ahead of the beta. But we can’t control email deliverability. We know that several % of our letters never reach recipients, but have no influence how to change that. Very annoying. And sorry you didn’t get your invitation

Yeah... no it isn't. Blaming people's email servers is the wrong move here. I've used Fastmail.com for the last 5 years, I keep my inbox at 0 and I look at my spam folder every week. I've never had an issue receiving email before. And by the way, over those years I signed up for the beta multiple times. All kinds of wild shit makes it through to my email, so there's no way your invite got bounced. Good luck.

This looks pretty nice. I use Notion at work and mostly like it but not for personal data because I want everything available to back up on a physical disk. I've been burned too many times with data getting stuck in an inaccessible format when upstream apps get orphaned. Today my primary storage is files on disk in common formats (Markdown, PDFs, JPEGs, etc).

Questions: Is there a way to get at the underlying documents stored in Anytype? Is there any path to a web accessible version for clients too small to host the full dataset? Use case is to host around 50GB of notes, records, and photos in one place.

Currently you can export all files, but there's no way to access them otherwise. They plan on having an API. Right now you're limited to 1GB of storage that they pin on their IPFS node, but they plan to enable self-hosting, i.e., using your own IPFS node.

Roadmap on GitHub: https://github.com/orgs/anyproto/projects/1/views/1

i've been following anytype since 2019, when the initial public release plan was for autumn of 2020 (lol), and since then it has transitioned from being a libp2p based google docs alternative into something more complicated, with relationships, sets, etc, and tbh i just never use it, because it's become too complicated

every time i open the app, which uses electron btw ;(, i catch myself looking at the explainer videos sent out, explaining all the concepts, it's a tool that (for me personally) gets in the way of actually doing anything a lot; maybe there are people for whom this is an optimal setup, but if you're looking for a privacy-friendly google doc alternative, this is not it imo

It really is amazing how little competition Google docs has with free or open tools. I just want something that lets me write a never ending river of rich text documents. I don't want to write markdown source (by default), I don't want a "preview" as I type, etc.--I want WYSIWYG. I don't want to micromanage folders and hierarchies and classifications and tags and whatnot, I don't even want to think about storage or files. Just an app that instantly starts, lets me capture rich text, and search through documents. I've never found anything close to that outside Google docs and it's frustrating.

real

a libp2p based wysiwyg or even a plaintext app with p2p sync + works offline + encrypted + cross device sync + collaboration with other users would be invaluable

Google Docs being free probably has a lot to do with the lack of competition.

Would you need to manage page settings, like margins and tab stops, or are those things more than you need?

If it really is just a local-first, rich text editor with syncing, searchable documents that you need, there might be something on the horizon. But if you need a full blown document editor, that's a bit of a different story.

Anything more than formatted text and inline images changes the scope in pretty exponential ways.

It's not unusual to need several hours of use to understand if a tool is right for you. I use Logseq and Obsidian, and without going to the forums and watching how the features are supposed to be used, I would have continued to use it like Notepad++, becuase out of the box, that's all it looks like it does.

true, i guess i phrased it a little bit wrong

i do understand it now, it's more like i don't have a use case for a graph structure to my documents/files, and since i sort of went exploring anytype while retaining the initial positioning of the app (p2p google documents), i was somewhat disappointed

Sorry but the landing page is super annoying.

Everything gives the impression of being clickable and not even half of it is.

Then I click on "crazy fast loading times" and a slow popups overlay appears.. the slider and the animations are feel deliberately annoying - slow slide.. when a slide is in the center it stops.. takes a deep breath and then is beeing reminded that it has to load it's three arms and bubbles... and all that for explaining in 2 sentence what open source is.

(And the overlay not being dismissable with ESC is annoying)

The UI for the website reminds me of the app itself. It paints its own canvas and is simplified to the point of adding more confusion.

The website is horribly over-engineered and makes it look like a crypto scam. Such a shame

Here's another notion open source alternative: https://affine.pro/

Although a big selling point for Anytype is that it is local-first.

My main gripe with Evernote now, is that in its effort to be a collaborative app, there is a lot of network latency for edits on my mobile device. By comparison, NotesNook is local-first and super fast.

I forgot to mention, affine.pro is self-hostable too... and it doesn't require any account registration whatsoever on their website.

What I mean by “local-first” is not the same as “self-host”. Local-first software are architected in a way where the local device is the source of truth. It usually uses CDRTs or something like git to sync with other peers. It is not a client-server architecture. Uploading files to a personal server is more of designating the server as a peer and regularly syncing with it, similar to how git works with remote repos.

Local-first architecture allows for better control over what you are storing. For me though, local-first architecture lets me take notes with little latency, and I don’t have to wait for timeouts so that the app realizes it is offline. I just want to get to typing in my note quickly and think about syncs later.

Yes, but Affine is local-first. You don't need a container infra or anything, just run locally without an account, data is stored in a user folder.

How does the P2P syncing aspect work with Affine? I didn't see much of anything written about that on the Readme.

EDIT: Aha, https://github.com/toeverything/OctoBase

Indeed, AFFiNE is also local-first! Anytype does have a lot of distinct functionality though, and the P2P model is a big plus there over other options..

That said, AFFiNE also appears to be genuinely open source, while Anytype is only source available.

I am in favor of genuine open source rather than source-available. I also see the OctoBase project, which enables local-first, p2p collaboration using CRDTs.

What about importing Evernote into AFFINE? Do such tools exist?

Very excited to see an alternative to Notion that checks so many of the boxes I care about, especially local-first control of my data.

It would be nice if you would let me opt out of all of your telemetry. I know it's useful for development but I've had to block outbound requests to Sentry, Amplitude, and Graylog already.

I would also like the ability to entirely disable cloud-anything, including backup.

To allow analytics opt-out is on the roadmap for 2023Q3

You can make a fork, disable analytics and make your own build. We plan to make smooth anlytics opt-out later this year.

I've been using this for about a year. It's very polished and been stable for me.

The Apple and Android apps work well too. They've been planning to open source it from the beginning, I'm glad it's finally here.

I'm using it as a very basic note system, nothing fancy like others are doing though.

Can someone explain this is — what it does and what it's used for? I'm not familiar with Notion at all and the page — which basically says you can use it for everything — is not helpful in this regard.

It sounds like a P2P shared note and document app but it's not clear to me.

It is a P2P note taking app with some object-oriented modeling capabilities.

It's much more complicated than that and that's it's biggest downside. I've been trying to learn how to use it but it's quite confusing figuring out their terminology and UI.

One thing that I did with it was make a Recipe type that has any number of Ingredient types in it, takes any number or arbitrary Dietary Restriction tags (e.g. Gluten, Dairy, Vegan), and takes parameters like cooking time and number of servings etc. So not only is a Recipe a note that explains how to make it, but it has a template with the boilerplate information pre-filled so when writing down a new Recipe, it's faster, and it has all that metadata, which allows you to construct Sets from the Recipes.

For example, I can have a document called Quick Vegan Crowd-Pleasers that is a Set of all vegan Recipes with cooking time under 30 minutes and 8+ servings. If I really wanted to go psycho mode with it, I could track every ingredient currently in my kitchen and filter recipes by what I can make with what I have on hand.

You can really do a lot with it. It's just time intensive to set up and also to keep up with entering data.

Think of it as a CMS with custom fields and widgets, for a personal website or knowledgebase.

"Knowledgebase apps" try to reconcile the fact that people (most people? all people?) think in sporadic and unstructured ways with the fact that maintaining data is mostly about codifying it into a structure.

So, instead of forcing a person to think "which app do I need to start getting some ideas sketched out", the app wants people to use it and then do whatever makes sense to them as a means of dumping all of the disparate knowledge into a bag so that it can be recalled, even if unstructured, later.

In an idealized example: if a writer got an idea, they would whip open the knowledgebase app and then...

- they might begin writing a list of characters and start detailing their relationships.

- they might upload a few photos of their local environment for remembering setting details, later.

- they might open a 'canvas' and put together a rough drawing of a new spaceship design they're mulling over.

- they might start working out the economy of their world by drafting an accounting table and assigning the trade networks.

The idea being that if any one of those tasks leads to any of the others (an inevitability), it's nice to just be able to scroll down the page and start on the new data without having to think "okay, which app do should I put this in"?

So, overall, it's about velocity of data recording. Some people just don't work that fast, or would prefer to consider HOW to think about an idea before they want to consider the idea, itself. So, for them, this may seem a bit like overkill. But there are definitely people whom can be far more productive if they can bounce around storing data as it surfaces in their mind, rather than when they have time to categorize it.

Notion is a note taking app. With a shared note feature. (You are right.)

And then it has a ton of more features, almost too many too mention.

Anything apart from simple unformatted text is (in summary) called a component.

It has a wide variety of components. Some of these components have a level of intelligence to them. Some of these components are references and links to information in other notes. Some of them are intelligent components that selectively show information in other notes. Some of them are intelligent components that selectively show information in other components.

Components can be tables, formatted headers, lists, dynamic content, charts, “views” similar to sql views, etc.

That is all I know.

You can think of anytype as a web-site builder. You can create pages (objects), databases, interlink them, create a sidebar. Currently you have only one private space, so it’s more about personal knowledge and productivity. With sharing you can publish a blog, create a community or a private team space. So a site builder with a ux of a note taking (so easy, no skills required)

Note, their idea[0] of "open source" doesn't match the popular Open Source Definition [1]

[0] https://github.com/anyproto/anytype-kotlin/blob/main/LICENSE...

[1] https://opensource.org/osd/

Hence why the "free" in "Free and Open Source Software" really matters. :)

Damn, didn't even think about that.

Not really, someone could just use that misleadingly too, just as with "open source".

"Open Source" already doesn't cover these special licenses, instead, the "source available" is used. Any also acknowledges this actually - the license they are using is called the "Any Source Available License 1.0".

After a bit of brushing up on my acronyms, one could indeed use FOSS or FLOSS to denote that a piece of software is either free (or libre) or just open source [0].

The term "open source" is not referring to software with a free license, but to software whose source code is available to the public irrespective of license [1].

[0] https://www.gnu.org/philosophy/floss-and-foss.en.html

[1] https://www.gnu.org/philosophy/open-source-misses-the-point....

That article you post as [1] refers to https://opensource.org/osd/ for "The official definition of open source software (which is published by the Open Source Initiative and is too long to include here)"

The FSF article directly follow with this, which contradicts your claim: "However, the obvious meaning for the expression “open source software” is “You can look at the source code.” Indeed, most people seem to misunderstand “open source software” that way. (The clear term for that meaning is “source available.”) That criterion is much weaker than the free software definition, much weaker also than the official definition of open source. It includes many programs that are neither free nor open source."

And the OSD also disagrees: "The license must allow modifications and derived works, and must allow them to be distributed under the same terms as the license of the original software."

Terms that can be summarized as "whose source code is available to the public" are called "source available", even by the FSF.

You are indeed correct, should have dug a little deeper (or maybe not doubt the OSD's definition in the first place). Thank you for the clarification :)

GNU also misses the point a bit. With open source, the source is open, but some other general rights are included too, like to restriction to the type usage. Lately, people and corporations made a lot of money on the backs of open source developers, so a new type of license emerged, and this would be the one that really is just about the "open" "source", but to make it distinct from the already widely known term, people call these "source available". Getting back to the topic, Any knows these distinctions too - or at least their lawyer did, because they call their license a "Source Available License"[0]. Source-available however doesn't carry the coolness of what "open source" brings - so on the marketing page, they refer to the project as "open source", which kind of can be argued, since the majority of it is indeed proper open source.

[0] https://github.com/anyproto/anytype-kotlin/blob/main/LICENSE...

Co-founder of anytype is here. The most of are repos are fully open-source. We have our philosophy regarding open source here https://blog.anytype.io/our-open-philosophy/ Happy to discuss concerns regarding our approach.

You deny your users the most basic freedom there is, the freedom to use your software for any purpose without discrimination. This is wrong, and so is your attempt to misuse and redefine the term "open source software".

This is the second time today that someone mistakes open source and free software.

See: https://www.gnu.org/philosophy/open-source-misses-the-point....

I don't blame you, it's a bit schadenfreude on my part because "open source" companies try to dress as free software but aren't

No. Open source is the same as free software. It is a marketing term for free software, that defines the same concept in more practical terms. If you actually read the article you linked, you would have known that.

Open source is not the same as free software.

There are broadly two camps: Camp 1 who advocate for free software & free software alone, and Camp 2 who advocate for "open source" being an all-encompassing umbrella term for a few things, including free software. Those in Camp 1 are typically not supportive of the goals of those in Camp 2. Those in Camp 2 do often try and equivocate the two terms.

No, it really is. Open source is equivalent to free software in everything but definition, and does not include anything that is not free software. There are minor disagreements between different people from the two camps which licenses to accept (e.g. Debian where this definition originated from, does not consider GFDL with invariant sections free, while FSF apparently does).

It really is the whole point of it, define more clearly what criteria must be fulfilled for software to be considered free software.

> Open source is equivalent to free software in everything but definition

Is this a typo?

No, it isn’t. Open source and free software define the same thing in different ways. FSD defines the user’s freedoms, OSD defines what the license cannot forbid. The end result is the same though.

It's not, because Free Software handles practical and ethical advantages as an indivisible unit, while open source focus only on promoting practical advantages.

You are missing the point. The definition of free software focuses on the freedoms of the user, but these freedoms are not easy to verify against a specific software license. The practical aspect of the open source definition (which really is the Debian free software guidelines with the word Debian removed) is that it gives you a toolkit, ten criteria that a licence must fulfil to be considered free software. Importantly, when this definition was created, the alternative, the free software definition, was incomplete and lacked the freedom zero [0]. Even more importantly, that text apparently wasn’t widely known back then, and even Richard Stallman himself liked the DFSG as a definition of free software [1].

[0]: https://www.gnu.org/bulletins/bull1.txt

[1]: https://news.slashdot.org/comments.pl?sid=1129863&cid=268758...

More specifically, their license doesn't meet these criteria of open source software (which are the same as the Debian free software guidelines):

5. No Discrimination Against Persons or Groups

6. No Discrimination Against Fields of Endeavor

You replied to "You deny your users the most basic freedom there is, the freedom to use your software for any purpose without discrimination." and caim that open source isn't about that.

Let's see https://opensource.org/definition-annotated/, _the_ definition for open source, specifically the sections titled "No Discrimination Against Persons or Groups" and "No Discrimination Against Fields of Endeavor":

"The license must not discriminate against any person or group of persons."

"The license must not restrict anyone from making use of the program in a specific field of endeavor. For example, it may not restrict the program from being used in a business, or from being used for genetic research."

So what GP claimed seems to be exactly Open Source's point, no?

The way we crafted it is not clear. Our idea is straightforward: if you want to use the software, you can do so for free, whether for personal use or within an organization. However, if you aim to sell it for profit, you need to contribute to its creation in some way; this is why permission is required. At least, that's the case at this early stage.

That means it's zero-cost and source available, not open source. As it's your software that's your choice, but please don't abuse the term "open source" to describe it. It's no more "open source" than, say, DaVinci Resolve is.

> It's no more "open source" than, say, DaVinci Resolve is.

DaVinci Resolve is not source available. Anytype does not restrict you from forking the code as long as it is non-commercial. And it also lets you fork the code for commercial use, if you take their permission.

So it is simply a non-commercial open source license, with permission required for commercial use. Sure, not OSI Approved License™, but certainly "open source". If this was not "open source", neither would be GNU GPL, because it isn't permissive enough.

This is no different from Qt's dual license except that GPL allows commercial use too. Or, like Creative Commons NC licenses, but for software.

Amen!

Open-Source is a generic term. It's the opposite of "closed-source".

The OSI is on the record on this: https://opensource.org/pressreleases/certified-open-source.p...

The only people who want to push the whole "The OSI's version of Free Software" defines "open-source" rather than "OSI Approved License™" are the Anti-Property GPL folks that never liked the term "open source" anyway, the trolls that want to force other people to work for free, and the people at the OSI

Yes you are right. It's not even open source...

Which is the first time?

Any hacker news link?

> This is wrong

This is a highly subjective take - it might be better to stick to objective dictionary definitions.

This project clearly isn't open source, & shouldn't be advertised as such, but on the other hand the intent here is a common/popular one these days, & its not the first of its kind: I'm surprised no-one has yet coined a term for this relatively new breed of "faux-pen source" or whatever it is.

Our idea is straightforward: if you want to use the software, you can do so for free, whether for personal use or within an organization. However, if you aim to sell it for profit, you need to contribute to its creation in some way; this is why permission is required. At least, that's the case at this early stage.

I was going to come here and post about how this is exactly what I've been looking for, but then I read this.

So, my kneejerk reaction to this deceptive use of open source is to just say "no" and move on. However, I read through your philosphy, and I have a question.

> considering the substantial R&D resources required for the application layer, we believe that businesses and networks utilizing our software for commercial purposes should contribute towards its ongoing development, allowing maintainers to support and enhance the platform.

That seems to be the crux of the concern here. I can respect that. So, why do existing open source licenses not suit you? For example, you could release the software under the AGPL, and still dual-license it as you wish.

Rather than assume bad faith, I'm going to give you the chance to correct yourself. At the very least, calling yourself open source at the moment is deceptive, whether you realize it or not.

The most of are repost are open-sourced with MIT licence. It's clear now, that the way we put words together in the license for clients is not clear. Our idea is straightforward: if you want to use the software, you can do so for free, whether for personal use or within an organization. However, if you aim to sell it for profit(like change the logo and put price tag on it), you need to contribute to its creation in some way; this is why permission is required. At least, that's the case at this early stage.

I don't see the problem with using Anytype License instead of AGPL. AGPL allows commercial use as long as the source code is provided, while Anytype disallows commercial use. Other than that it is the same, and is "open source" regardless. As in, OSI doesn't trademark "open source".

Using the term Open Source for a product when most of the code is under a license that isn’t Open Source feels dishonest to me. The product you build is certainly yours to release however you see fit, but if I’m looking for open source software and I find this I’m going to be extremely skeptical of everything you say.

On top of that, the license itself is actually incredibly restrictive. I’m not a lawyer, but my read of the section on economic value seems very broad:

> does not include uses where the Software facilitates any transaction of economic value other than on Allowed Networks.

My read of “facilitates any transaction of economic value” means that I would be in violation if I used this to keep track of trading cards, made a grocery list, or tried to keep track of what I want to buy my friends for their birthdays. At least it would if I installed this on my home server and accessed it from the couch on mh phone.

Thank you the comment, it's clear that licence is not clear and we need to improve. Our idea is straightforward: if you want to use the software, you can do so for free, whether for personal use or within an organization. However, if you aim to sell it for profit, you need to contribute to its creation in some way; this is why permission is required. At least, that's the case at this early stage.

Serious question - did you have a lawyer write this? The license text and some of the comments here lead me to believe that it wasn’t done by a lawyer.

If not, you should really talk to a lawyer first about this. Preferably one with knowledge of open source licensing. New software licenses are tricky and should be done by a lawyer and not by random HN comments. (Even if this is an overly well informed set of users on software).

The problem is that proprietary licenses (such as Source Available) are viral: whatever they touch becomes proprietary.

As such, "most repos are open source" (from what I can see: MIT, some forked ones Apache 2.0) is nice, but the end product still isn't open source according to OSD.

There are people who value using "Open Source" for OSD-compliant licenses only (I tend to agree with that notion to keep things clear), but I didn't really want to discuss this: It's your project, after all, license as you wish.

I just wanted to provide a heads-up that the use of "open source" in the header here (and the front page on your site) doesn't match the expectations of a bunch of folks, so they know whether to look closer or not based on that.

I see how making the entire situation transparent muddies the message, but "Everything is Source Available, many parts are Open Source" would already clear things up a lot.

Yep it's clear, you are right!

You used the word "fully", which means part of them isn't fully open source. :(

You should be always skeptic but when someone write "Pure transparency — trust our code, not our words" in an EXTRA BIG font, you should be extra skeptic.

(co-founder of anytype) our main promises are privacy, end-to-end encryption, user controlled keys, self-hosting, p2p sync - all of which should add up to what we can user autonomy from the software provider which we believe to be important. To prove these claims the best way is open the source code. As promises of encryption and ownership stay promises unless you can be sure of it. That was one main motivation and why we think it's worth highlighting.

Is there a way to make sure I am using the code you published and not a different version?

We do releases in the Github Actions CI. So you can inspect the CI logs and published artefacts(desktop/android). Then you can compare the binaries checksums. I would appreciate ideas on how we can make it more transparent

So I see the networking portions of the code are Free Software - that's great. How are people expected to use it if their use does not fall under "non-commercial use" as defined in the client license? Do you expect people to write their own clients for commercial use, or do you offer commercial licenses?

For non-commercial use, you're prohibited from selling it, but using it within your organization is permitted

You do realize this also means that a for-profit company can’t use the software at all, right?

Unmodified, modified, whatever.

I don’t think that was your intent, but that’s exactly how it reads. Or maybe that is your intent?

The idea if you want to sell clients, you need to get a permission form our association.

Why not use the GPL?

Because we want to provide other organizations with the opportunity to offer paid sync services, we needed to incorporate the concept of a network into the license. We crafted the license with that consideration in mind

> Because we want to provide other organizations with the opportunity to offer paid sync services

Why do you think the GPL is not compatible with this?

Exactly…

Plus, they can always offer other licenses in addition to the GPL (or really AGPL would be a better fit for their concerns). It’s their software, so they can license it to anyone with whatever terms they want. (Assuming there aren’t outside contributions, but even that can be dealt with)

> Plus, they can always offer other licenses in addition to the GPL

Yep, that's exactly what I was thinking. They're clearly offering alternative licenses to the one in the public repository. There is no reason the same tactic couldn't be applied with the GPL.

You may employed a wrong strategy.

To prevent paid sync services, you should license your protocols and data formats in AGPL, which requires derivative work (third-party sync services) to be open sourced.

The client app in contrast, should be fine even in permissive licenses.

When will VC companies stop conflating source available with open source? It's even worse than just being closed source as it's trying to pander to a group for money, and without any respect to the actual values of what they are saying.

Our philosophy surrounding open-source is uncomplicated and clear. All essential protocols and data formats are subject to the MIT license. However, considering our clients' needs and with the MongoDB situation as a reference, we must maintain some degree of defensibility.

Our objective is to foster a collaborative atmosphere, co-creating with the community. The size of the community matters to us; when it reaches a significant scale, we want to make collective decisions regarding licensing, reflecting a democratic approach.

Antype, a creation of our non-profit organization, is aimed at sustainability rather than becoming another digital ghost. Our mission isn't simply to exist, but to thrive and make meaningful contributions to the open-source landscape. By intertwining our growth with that of our community, we're setting the stage for a sustainable future.

These are all noble goals, and I don't think many commenters here are arguing that they're not justified. Just that the specific use of the term "open source" is incorrectly applied (to your own benefit).

If a subset of things you do are open source, that's great. Say that.

Yep, you right, we'll correct it.

Thank you! Terminology matters. I really appreciate that you’re willing to make that change.

Your goals sound laudable, but they do not alter the fact that your use of the term “open source” is, deliberately or otherwise, misleading.

> Our philosophy surrounding open-source is uncomplicated and clear.

It does not appear to be the case that your philosophy surrounding open-source is clear. You state very clearly, without caveats, that the product is open-source, which strongly implies that the product is, well, open-source without caveats – and this is not the case. That feels rather disingenuous, if not deliberately dishonest.

It is not unlike printer manufactures loudly proclaiming a page-per-minute value without any note about that rate is only attainable feeding entirely blank A6 sheets out of the device.

Most of our repositories are MIT licensed, which holds significant value. Some of our repositories are source-available. We believe we are building an open-source product, and we reserve the right to define what 'open-source' means for us. Our only prohibition is against competitors making minor modifications and then selling our product. Our position on this is clear. We recognize that our definition may not align with others' perspectives here, and we are open to understanding that. However, labeling our approach as dishonest isn't something we consider accurate. To avoid any confusion for those who uphold the traditional notion of 'open source', we will change the term 'open source' to 'open code' on our website.

It's unfortunate it's not actually going to be Free Software, but being able to at least audit and do non-commercial is... better than Notion, I guess. But I do retract a fair portion of my excitement over the project.

Your "reserved right" to define what a word means to you... puts the work to somehow figure out what you mean onto the reader. This isn't quite nice. The speakers are the ones that should strive to make themselves understood in the first place.

I don't have any love for "open source" since it is just "the part of Free Software that appeases people in suits", but please, use the thing as it is.

While language is by nature subjective and ever changing, be careful defending the right to redefine commonly understood terms so significantly. By the same measure people could choose to interpret anything you say, including your other license agreements, however they like!

Source available, of client and server, is definitely better than no source

I do think open source business model, especially consumer, is difficult to tackle, while protecting the creator

So congratulations and good luck

Are you seriously trying to suggest that source code being publicly accessible is worse than nobody being able to access it? What a ridiculous statement.

You are intentionally misrepresenting OP.

> When will VC companies stop *conflating source available with open source*?

*Misrepresenting* "open source" is the problem. "source available" is the correct phrase

It was worse in the case of early BIOS, which IBM made public in an instruction manual while keeping all relevant rights on distribution. I wonder why they did not publish it with a GPL or AGPL license, where they would keep commercial use to themselves.

IBM PC was released in 1981, GPL 2.0 (which was the first version in wide use) in 1991.

A lot of the comments to this are angry and I can appreciate that. There is obviously some specific nuance to "open source" that individuals in this thread want to maintain and feel that the OSI has enshrined in stone as a definition that no one shall breach.

However, I strongly support the kind of license (in principle) that this software is released under. Source code is available for anyone to inspect, modify for their own use, contribute to, run locally for their own benefit. The main restriction is so clearly obvious: you can't create a commercial competitor. You can't take their code and with minimal effort or minor changes create a competing app and sell it to others.

This to me seems like a completely sane license. So common in fact that creative commons asks two basic questions when they recommend a license: "Allow adaptations of your work to be shared?" and "Allow commercial uses of your work?". In fact, they differentiate this difference with the moniker "Free cultural works" [1] (those that allow commercial use are termed "free").

I'd like to see the same nuance in software licenses. A difference between "open" and "free". That way, we can avoid this bickering in the comments where those who really want completely free software (free from all restrictions including those against commercial use) won't jump down the throats of those who want to open up their source while protecting themselves from competing commercial use.

1. https://creativecommons.org/share-your-work/public-domain/fr...

This isn’t bickering. These definitions have existed for literal decades. There are multiple models for making source code available, and people can choose what license they want. But this isn’t open source. This is source available, free (as in beer, kinda) license.

You don’t have to pay, and you can see the source code. But, in my quick reading, I don’t think you can make modifications, distribute modifications, distribute unmodified versions, and there is a restriction on how you use the software (non-commercial only).

This is the same type of license Microsoft gave certain large (TLA) customers for Windows, IIRC. I believe they called it “shared source,” as in they shared a copy of the source with you, but you couldn’t use the source for more than review. No one would claim that was open source.

There are differences between free, open, and available. This is only the later. No one cares about what license something is available through. Authors get to do whatever they want. People only care when you try to claim one thing, but it is really something else. In this case, the company is trying to use the term “open source” as a selling point of their software, when it isn’t. This license doesn’t even match the definitions they use on their own site!

I’m happy the authors want to make it possible to audit their software. That’s a laudable goal. If they want to restrict usage of the source code to non-commercial use, that’s fine and up to them. Just don’t call it “open source”.

Just because something is free doesn’t make it open. And just because something is open doesn’t make it free (as in freedom or beer). Similarly, just because something auditable and available, doesn’t make it free or open.

The fact that the authors don’t know the difference (or are potentially misrepresenting the difference) will only make the community mad - especially the part of the community that would care about seeing the source code in the first place. If they instead were marketing the project as “source available” for auditing or non-commercial use, this wouldn’t have been an issue.

Your first points: "can’t make modifications, distribute modifications, distribute unmodified versions" appears to contradict the language from their license file:

> Any Association grants you (“Licensee”) a license to use, modify, and redistribute the Software, but only (a) for Non-Commercial Use, or (b) for Commercial Use in Allowed Networks.

> These terms do not allow Licensee to sublicense or transfer any of Licensee’s rights to anyone else.

IANAL, but I thought transferring rights was required to redistribute a work. If someone downloads this software from me, I can’t give them a right to use that software. Only the original authors can do that. And even if an earlier clause says that I can redistribute, this cause suggests that I can’t.

As it is, this is a software license that I wouldn’t touch.

But even if I am wrong on this point, the rest of my argument stands. With the restrictions on use, this is neither an open nor free (as in freedom) license.

(Side note: this is why new authors shouldn’t roll their own licenses. Ambiguity is not what you want to see in a license agreement.)

The right you are granted is more specific, so you can redistribute, but the next hop cannot. (I'm only reading what is quoted here in the thread, but I think the important parts are included.)

>The main restriction is so clearly obvious: you can't create a commercial competitor.

I don't see it. Commercial use there is defined as "where the Software facilitates any transaction of economic value other than on Allowed Networks". As I understand, it means financial application.

I think the issue is that they say "open source!" with no further qualification. That's misleading and disingenuous.

I support the use of alternative business models like source available, Business Source License etc. That's fine. But you should accurately describe your licensing. They should have said "source available".

So what is the appropriate term then for "source code freely available, but cannot be used for commercial use?"

I guess probably an unpopular opinion here, but I don't see why "open source" must imply that anyone should be allowed to fork the repo and sell it.

Source Available.

> I guess probably an unpopular opinion here, but I don't see why "open source" must imply that anyone should be allowed to fork the repo and sell it.

Because that is the definition of "Open Source"[0]. As was already said, "source available" is the correct term here.

[0] https://en.wikipedia.org/wiki/The_Open_Source_Definition

Sorry but this is not the definition of "open source", I would argue there is only a conceptual and cultural definition of "open source". What you are linking to is The Open Source Initiative (OSI) Foundation's declaration called the "Open Source Definition".

They are a single organization, that have done a tremendous job at trying to come up with a global and shared legal framework to which people can license code under. They have gone so far to come up with a pretty good definition of "open source", but not the definition.

This would be equivalent to saying that "Freedom" is defined by the US Constitution or the Canadian Charter of Rights & Freedoms. It is not, those are both examples of a legal definition of freedom, but neither are the sole authority for the global and cultural concept of "Freedom"

> This would be equivalent to saying that "Freedom" is defined by the US Constitution or the Canadian Charter of Rights & Freedoms.

The idea of freedom existed before both documents. The idea of Open Source was proposed in 1998 [0], and the OSI was created to define it in the same year [1]. This is not at all equivalent.

[0] https://en.wikipedia.org/wiki/History_of_free_and_open-sourc...

[1] https://en.wikipedia.org/wiki/Open_Source_Initiative

> I don't see why "open source" must imply that anyone should be allowed to fork the repo

Apart from 20+ years of historical use in that way?

You're completely misrepresenting my quote by cutting off the "and sell it" at the end, that's incredibly disingenuous.

MIT license restrict commercialization. Is the MIT license not open source then?

The MIT License does not restrict commercialization aside from requiring attribution. In fact, the text of the license includes the phrase "without restriction":

> Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

> The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

https://en.wikipedia.org/wiki/MIT_License

Open source software licenses have always allowed the licensed software to be resold by others. The very first criterion in the Open Source Definition is:

> 1. Free Redistribution

> The license shall not restrict any party from selling or giving away the software as a component of an aggregate software distribution containing programs from several different sources. The license shall not require a royalty or other fee for such sale.

https://opensource.org/osd/

The "Any Source Available License 1.0" is not an open source software license because it restricts commercial use, but it happens to be correctly named because it is a source-available software license. Source-available software (such as Anytype) is still preferable to proprietary software with little to no source code published (such as Notion), since it is easier to audit software when the source code is available.

Ok I stand corrected on the MIT license, thanks for the explanation. I wonder why so many on HN criticize it then, seems reasonable to me.

Source available does sadly not really mean that much. Just means the source code is available. MIT licensed code and code you can see but are not allowed to use can be described as source available. That is what https://faircode.io/ got created for to solve.

Background: https://github.com/n8n-io/n8n/issues/40

It's not freely available, it's available for a limited set of uses. So it's not open source.

The conventional open source model makes everything free labor for billion dollar companies and hustlers who just take it and slap it up behind a paywall. It's used to underpin SaaS models that are significantly less open or free than closed-source local commercial software.

It'd be nice if the OSI or the open source community addressed this issue head-on, but so far they refuse and insist nothing is wrong. This refusal leads to a proliferation of almost-open-source licenses that just muddy the waters.

If they continue to refuse I think we'll see more and more of this until the definition of open source becomes hopelessly muddy and the whole community starts to wither.

I'm not familiar with Notion. How does this compare to something like Obsidian?

Their philosophies are quite different; Obsidian is much more text and link oriented (it's basically a front-end for markdown), whereas Notion is object-oriented and gives primacy to the kinds of objects that can contain text (blocks, tables, quotes, etc.)

Obsidian can be made to work like Notion and vice-versa, but I see Obsidian as more of a hypertext environment/wiki creator, and Notion as an integrated environment to interface with different ways of interacting with knowledge-focused objects.

I prefer Obsidian largely due to its simplicity, and the fact that it maps really well to how I think. Notion is also incredible for what it does, but it's too much for my brain, and the app is still a bit too slow for the speed with which my thoughts can escape me.

Anytype purports to be like Notion.

Obsidian is plaintext, Notion is richtext. This means, it's simpler to write complex documents in Notion and you don't need to know and handle complicated syntax. Which is good complex stuff like layout, tables, images. Anytype seems to have the same richtext-interface. But with obsidian on the other side, editing can be faster, if you know your syntax, and can type well.

And unlike obsidian, Notion has no direct mechanism for extension, which means there is no way to modify the app, or extend Notion-Documents with new types of data. There is an indirect way, by embedding urls, which comes with some limitations, and ultimatly it means you are depending on other servers, which comes with some more problems.

Not sure how anytype supports extensions.

Anytype aims to be an open ecosystem. We open the source code. Now will announce the contributor program. We will start with translations and gallery of templates, then imports/data adaptors, then design themes and extensions.

"any-sync networks are permissionless. This means that anyone who knows the IP and port of the coordinator node can connect to it, create new spaces, or download existing encrypted objects. Data inside spaces is always secure, as it requires space encryption keys to read it."

From the docs on self-hosting. I am not a security expert, does anyone know if this is still considered safe? Feels weird that anyone can download my data, even if it is encrypted

Surely that’s just the nature of IPFS, right?

Not necessarily. You can add auth at the encrypted block level: https://peergos.org/posts/bats

Here's some project health insights, for those interested.

https://devboard.gitsense.com/anyproto

https://reviews.gitsense.com/insights/github?q=pull-age%3A%3...

Looks like Jira is used for issue tracking, so insights for issues and comments will be incomplete. There also appears to be a good mix of seasoned contributors (2+ years) and new (look for the gift icon in the review tool to see how long they have been contributing).

Full Disclosure: This is my tool

are you aware that the GH org parsing is apparently case sensitive? https://devboard.gitsense.com/jetbrains = blank https://devboard.gitsense.com/JetBrains = works (although it does whine about the number of repos)

Yeah this is something I'll need to fix. Also, the repos limit is to keep resources in check. For example, crunching 5000 repos for Microsoft takes about 10 minutes and ties up 8 out of 20 query threads.

Thanks for the feedback.

Just downloaded it. This looks incredibly polished and smooth! Will play around with it a bit, although I have never been good at taking notes.

iOS version is very well-crafted, but the type on iPad Mini is much, much too small. (I say that as a person who generally prefers small type.)

In terms of PKM applications, believe CRDT is the future。

In response to this, I've developed a "read later" app based on CRDT. The following are my expectations for knowledge management software, as well as the goals I had in mind when designing this application:

1. I want my data to be stored locally so that I can access it anytime and anywhere, thus eliminating the need for expensive hard drives.

2. I hope the software can be purchased for a one-time fee, rather than by subscription. Since all the data is stored locally, this indicates that there are no additional costs for the developers. Therefore, I believe a one-time payment would suffice. At the same time, I'm open to the game DLC model, where enhanced features can be obtained through additional payments.

3. I hope the software can support synchronization so that I can access the most recent data on any device.

I think this is a great application/service. You can have source code, self-hosting, and construct your own network of your devices. Nice work, thank you.

Maybe some kind of P2P network support web browser - without someone's server - would be nice but this does require so much more work, I assume.

Hmm I don't understand whats the benefit of having my notes on p2p network. Usually when it comes to notes I would prefer them to be private. Here I have some vague claim that the cryptographic key is safe enough. Well ok but what is the advantage here?

Also the local first approach here just seems like a local storage on a disk. Not great. Essentially if I move notes from notion to here I am going from a service that has a clear API and understandable storage policies to a model where the data is some binary blob on a p2p network.. Idk this to me seems I would effectively have less control over my data.

By default, your account is created on the machine. You can use it without connecting to any network. We allow you to sync notes in a peer-to-peer way in local networks, or over the Anytype network (or your self-hosted network) via the internet. Currently, you can sync notes among your devices. Soon, you'll be able to collaborate with others using end-to-end encryption. I can't agree that the claim about the key is vague, as there is documentation and code on GitHub. You can check it, and we are happy to hear arguments.

We are genuinely implementing a local-first approach. Why do you have doubts? We have a clear description of the data format, which is MIT licensed. Feel free to write and read. We hope that data adaptors will soon be available with the help of the community. You have full control over your data. Compile your version, and that's it.

Very nice and well developed product. I've been using it for a while! Glad that it's public and has a mobile app now! I've been a heavy user of Notion, finally, there is a better and more secured alternative!

I had been using Anytype for some time but a big disadvantage is it requires a standalone app = Not possible on a corporate laptop. What I use is VSCode with Foam - a colleague uses Dendron. Has the advantage that I can also use GitHub Copilot or local LLMs for autocompletion. Imho the combo is better for pure writing and building quick connections between documents

Interesting to read about AnyType's supposedly poor import of MD files – I imported a folder of MD files (from one of my large Obsidian repositories) with no problems whatsoever, at (very) impressive speed. So not quite sure what's going on there. In short: works for me ;-)

"There's no browser version of the app. Anytype is a stand-alone software, that works on desktop or mobile devices. There are many points of vulnerability in-browser apps that would compromise our commitment to data security and encryption."

What does this even mean? How does having a mobile app mean more security than a web app?

Are they worried about browser extensions? Or is this just an excuse not to host on the web?

I see this software has tables - are these able to be used like excel sheets?

I see this as a necessity for software like this. It's the one feature that Notion has that the competition doesn't seem to have, and it's absolutely crucial to everyday work for so many people. It seems to be a blind spot for programmer-types, I'm guessing because they see Excel as limited?

I don't think Notion has formulas for tables. I think you mean formulas for databases.

The formulas for databases like Notion has is actually the second-most requested feature right now and is tagged as Acknowledged, so hopefully we will get it soon.

Coda has the most powerful tables I've seen of the Notion alternatives, and was a fair bit cheaper for my team since they don't charge per viewer or editor but per "creator". But unfortunately it's not open-source or local.

They have a pretty good [writeup](https://coda.io/blog/productivity/tables-not-spreadsheets) on why "spreadsheets" were avoided in favor of more traditional "tables" (a la RDBMS).

(co-founder anytype here) tables are simple now, no formulas. Plan to add formulas to data-bases (what I think you mean by notion having it) - it's one of the top requested features on our forum

Is non-Electron client not on the Roadmap? That's a big missing piece in the puzzle of great note taking alternative

Great, I really want a offline version of Notion, and asked the official guys from 2019 to now. Thanks for your sharing!

Not to write the classic Dropbox comment, but can't you just do this with any of the notion alternatives based around local markdown files (i.e. obsidian, logseq). Just have local e2e encryption on your laptop and point the app at a shared folder with in-transit encryption (i.e. Syncthing).

You can, but it is definitely not a download and go situation. I use Syncthing for Obsidian and it generally works well. However, I was recently in need to use a VPN on both sides of the "sync" and it choked hard.

Made me miss onenote...

To be fair this looks like an open source project that aims to be download and go some day, not something there today

Yes but it's a lot of hassle to set up on Android. No idea about iPhone. And if you think that's fine for you, cool. But now imagine trying to set that up for a shared knowledge system. It's just not feasible.

Oh absolutely this sucks. Just not sure Anytype is any better. It doesn't support sharing, for example. Logseq + SyncThing just works on Android, you simply point both at the same folder. Presumably that's impossible on iPhone.

> It doesn't support sharing

Well this release doesn't but it's kind a beta by the look of things. Team wiki/company knowledge is definitely on their roadmap and one of their major selling points.

I tried this briefly about a year ago but didn't see myself using it. They had an hour long induction session over zoom to teach how to use the tool before they sent the download link iirc. It was too complicated.

Moved on to Zotero and text/plain files for my notetaking and journaling needs ever since.

I like the Notion and Confluence UIs.

Are there tools for importing from Evernote?

Does this have mobile apps?

And I assume, if I want to back it up somewhere, I can set up a peer that would sync from my devices?

A UI specific for journals is a huge selling point for me. That's like 80% of my use of Evernotes.

One thing I found amazing about Notion was how nicely my Evernote was imported in.

Can I do the same here

not yet, we have very limited import options at the moment (notion, markdown, csv and protobuf). This surely limits our activation > not easy to start. We are working on more import options and also plan to publish an API and hope our community helps to build more options together with us

Do you plan to create native desktop app? Android app looks and feels good, but the desktop one is terribly slow (especially scrolling), probably due to being written in TS... I used Linux AppImage version if that matters.

In contrast with many others crappy things (Notion, Standard Notes, Joplin, Notesnook, Obsidian, etc, etc, etc, I tried literally dozens of notetaking apps, including some self-hosted stuff), Anytype is, at least, barely acceptable.

It can be used to write notes and share them between devices, it can be used in offline mode, it is cross-platform, it has acceptable navigation and more or less good editor with proper code highlighting. Also, you don't have to share your content with server owners.

The UX isn't that nice at this point (especially the graph view which is totally useless if you have at least hundreds of objects) but, again, that's something acceptable and it's free (at least for now).

So firstly - been an Anytype beta user for a while here, and it's great to see the launch! And how far Anytype has come on recently - a previous frustration admittedly was that the app and data sync could feel quite sluggish at times, and the latest release is way snappier and feels more stable! Excited to see how the project develops.

I did want to echo some of the points already raised regarding licensing.. From the outset, when I heard of the Anytype project some time ago, the whole thing that got me excited about it, and why I've been following it so attentively for so long, was the promise of an "open source, privacy-first, p2p, Notion alternative". While I'm sure we can all respect the licensing model you've settled on from a commercial perspective, and being source-available with select open source components is far more welcome over a fully closed model, it clearly does not make the project as a whole open source. While that's quite disappointing, it's positive to see you're acknowledging this and tweaking how you present the open source nature of the project to accurately reflect!

I did just want to ask however if you might have considered a licensing model similar to a project like Sentry? They also fully open source select components under permissive licences (MIT/Apache), but for key commercial elements they instead adopt the Business Source Licence (https://open.sentry.io/licensing/). This is very similar to the terms of your "Any Source Available License" in allowing personal use but restricting commercial use, with the very key caveat that it only enforces these restrictions for a period of 36 months, after which point the licence for those elements reverts to Apache 2.0.

Such a restriction would in theory protect commercial interests in a similar manner (only code from more than 36 months ago, or whatever delay preferred, is fully open sourced), but make the project far more open in spirit, and critically allay concerns Anytype will otherwise attract as a non open source project. For example - what happens if Anytype go closed in future once people are already bought in / if the organisation or development ceases at some point / if some decision was to be made in future which compromises user privacy or security, etc! A key element of open source is that trust doesn't just come from saying "trust me, honest!", but also from the licensing model - if the project disappears or is somehow compromised, it can be forked. Such a licence would allow this (with a delay!)

Might you have considered such a license, and if so would be curious your thoughts on why that isn't your preferred approach?

Looks nice, was disappointed with the analytics/telemetry however...

Also i found the website to be quite chaotic. I'd prefer something better organized and informative.

hey everyone, Zhanna - co-founder of anytype here (with my co-founder Anton also here). If you have questions or feedback, we are here to answer and listen :)

on one hand it’s good that we have so many note taking apps.

it could also signal about a huge gap in the platform that’s not being addressed

is it sync, ux , storage , collaboration

the fact that we are pivoting note taking an productivity trends every couple years is interesting

it could also just be another trend like micro services , nosql, orm , oop– something people do when they need to procrastinate getting work done

Don't bother to download if you are unwilling to create an account with their service.

Account creation happens on your device without an internet connection. Just the encryption key is generated. Anytype does not have a central registry of users. So i think you are misinterpreting it. We have no central registry where you create an account

Marketed with "Trust our code" , but doesn't share it?

Here is their github with repos

https://github.com/anyproto

Those are client libraries, not server

Here we are https://github.com/anyproto/any-sync

dang these types of website layouts are bad...how can i trust that your app is useful if can't even make a simple easy to read website telling me about your app?

It's a matter of taste, and ours differs. You can establish trust by going to GitHub, checking the code, and compiling it yourself. In my humble opinion, trust and taste aren't closely connected.

>Computer pioneers

Nice to see Ted Nelson getting some recognition.

That website looks horrible.

vim. cat. less. grep. find. wc. git. (and so on...)

life is better (TM)

kids: get off my lawn

How do you make money

I'm sorry to hijack the thread but as someone forced to use Notion as a team wiki, please improve the performance! I can't emphasize this enough. The Notion subreddit is filled with complaints: if that's not a wake up call to every single PM at Notion, I don't know what is.

Stop shipping features until the perf is a joy to your users. It's mind-boggling to me that companies would prio anything over fixing perf outside of fixing bugs.

Notion is the single biggest frustration of any workplace tool I've ever used in my entire software career and it all comes down to its performance.

> Notion is the single biggest frustration of any workplace tool I've ever used in my entire software career and it all comes down to its performance.

Even worse than JIRA, or MS Teams? That sounds horrible indeed.

Not in my experience. Notion is faster than Jira. But we don't have super massive wikis, hundreds of pages at most.

Yes, worse but it depends on what type of page you're working in. If it's of medium complexity and up, it's worse (a few media elements, maybe a database, etc.)

If it's just text, it's faster than JIRA or Teams.

I have the same problem, I understand the enjoyment of sales / marketing team when it comes to using Notion, but for engineering team, it's just horrible, so slow.

I can vouch for it being worse than JIRA, which I was absolutely not prepared for when I moved to a company that uses Notion for its ticketing system.

They could try the WebKit strategy [1]; a strict policy against performance regression:

> We adhere to a simple directive for all work we do on WebKit: The way to make a program faster is to never let it get slower.

[1]: https://webkit.org/performance/

yeah i hate the slow perf too, but their success in the market unfortunately i think validates the "features over perf" strategy

(also, as slow as it is now, it's MUCH faster than it used to be, if you can imagine that!)

I went from being a notion advocate to a hater. Perf, mobile apps, and pricing weirdness have just made it clear it's no longer a high quality product but another enterprise-y Jira type product.

The early notion product was really performant, the mobile apps were ok enough and it had good ideas.

Now my last remaining notion sites sit on my to do list to be migrated to an alternative. The mobile apps are trash. Working with notion is painful.

What is your alternative?

Most things moved into Obsidian with databases that needed to stay databases (versus tables) going into NocoDB (an airtable clone that can use sqlite).

I still haven't found a replacement for "online thing that anyone can edit easily". Some wiki systems are close but I don't want to give up some of the richer component types like kanban cards.

Notion is definitely quite slow

But I also think a lot of companies in the same stage as Notion have fallen into the trap of shipping features (namely AI features) ahead of fixing perf bugs

We detached this subthread from https://news.ycombinator.com/item?id=36799739.

Reply on news.ycombinator.com