I'm starting to feel Anytype's limitations

Salute to the Community! I wanted to write this post for some time now and I’ve postponed it for various reasons whether I found a workaround with current anytype’s feature or it wasn’t vital to my workflow but I feel like now it’s the right time for me to reflect on my one year journey with Anytype. This is by no means a rant, complaint or anything like that. It’s just my experience with Anytype and what I could and couldn’t do without it. I will try to be as thorough as possible while keeping this thread short; also, I will try my best to link posts related to my points, ideas, features, etc.

You have probably seen my movies/series database already. Personally, I really love how it looks and what I can do with it but at the same time, it wouldn’t have been possible without the introduction of inline sets. Also, what you may have not known is that it’s very much incomplete because of the limitations of anytype ( I will give detailed examples in a bit). Also, I want to use or I’m currently using anytype as my fitness tracker, my task manager, my habit tracker, my physics research companion, where I could manager my phd hunting process, etc but as of right now, the limitations of anytype in its current from of object/relation/set structure is seriously holding me back. As a side note, I never clicked with Notion like I did with anytype; I didn’t like its design and couldn’t understand all the hype that was around its database features but I think I’m starting to fully understand and appreciate its capabilities. I’m comparing Anytype to Notion because I think later this year when Anytype goes public, it would be its main competitor. So, without further ado let’s dive deeper into these limitations that I keep mentioning them.

The Argument for Fitness and Habit Tracker

The current method in building a fitness tracker in Notion is to build multiple databases and uses two of the most powerful properties in Notion, roll up and relation, to link those databases to your liking and making it possible for yourself to see information at a glance. To be specific, you can have a exercise database to include items such as “push up”, “pull up”, etc. You can have another databases for muscle groups, and another one for your days. Now, the thing that you can do in Notion that is currently impossible for anytype is linking the data between these databases to showcase your needs. For example, you can track which workout you did for each day of the week, how many sets did you do, etc. Also, you can view your days in a calendar view and view which workouts you have planned or done for that specific day. With roll up and relations in notion you can show as much information span across these databases as you want.

But, as much as I tried I could have a similar thing in Anytype. let me explain. I created an object type called “daily entry” and I created a set of these objects. I wanted each of these daily eateries to be representative of my day. It has a part where I write down my thoughts for the day. I created an inline tasks set and tweaked its filters so that it’s showing my tasks for that day only. Remember, it was possible because tasks usually have a specific “due date”. So far so good, I even linked that task to its specific projects, set its priority and what not. Next, I wanted to show the list of exercise and habits I planned for that specific day and I created two inline sets for each of them. But this was the beginning of the problem. I wanted to be able to track my progress ( I know anytype currently doesn’t support formulas, but I imagine it will be added in the future) during the week, month, etc. But the thing is I either have to create separate objects for every day that I want to exercise a certain workout, i.e. sit ups or the sensible thing to do is set the filter on the inline sets to show only the specific work outs for that day. But since the “sit up” object can’t have infinite “date” relation for each day, I couldn’t do it. Another workaround would be to create a “done” relation for the work out object and add each day that I do this work out in it. This would temporarily resolve my inline set problem but then it would defeat the purpose of “tracking” my progress. You see, the set that I would create of “work out” objects, only has one “done” relation and at one time and if I add this relation to set view, I would end of with a column with hundreds of days and it wouldn’t even show the days completely because of the current wrapping display settings! Moreover, I couldn’t create a view of the week and month for tracking my progress because each of my “daily entry” objects have a “date” relation but my work out set and objects don’t.

The Argument for Movies/Series Database

One of things I wanted to do was to add multiple object types like Academy Awards and display the awards that either that particular movie/tv show or actor/actress/director/composer have won. But again, because of the anytype’s limitations I couldn’t. Let me explain. An object like the Academy awards (Oscars) has multiple categories such as best actors, best picture, etc. So, I can add a “best picture” relation for each of the movie objects to include it but the problem is that every year has different “best picture” being awarded to a movie! So, I can’t then display which year which movie won the best picture. So, for now I abandoned this idea unfortunately. A related idea is discussed here.


I know currently anytype is busy with migrating to the new sync infrastructure but I hope after the busy days up until the public lunch, we could start thinking about the future of Anytype, especially the future of object/relation/set structure. I thought with the introduction of “relations as objects” we are moving toward that direction of expanding the functionalities of the current structure but unfortunately it wasn’t so. It would be great if Anytype adopted the roll up+relation feature of Notion but I’m not sure it would be possible structure wise. Instead, I think the idea of adding dimensions to relations could be an interesting replacement. Imagine the relations like objects could be linked with each other. This way, we could add a “sub relation” to an existing relation. Remember the oscars example, if we could link the “date” and “object” relations in a way that I can specify extra information to relations, in this cases adding a “date relation” to “best picture” relation. Or in the fitness tracker example, if we could have a relation like “done” which not only links the “work out” object to the the “daily entry object”, but also we can choose to link another relation/property of the “daily entry” to the work out. This is basically implementing the functionality of roll up/relation property of notion in Anytype.


Other limitations that have been mentioned in the past by the community is having a calendar view, timeline view, inline latex feature, having a link for each block in order to either mention/reference it in other objects or have a synced version of it in multiple objects.

At this point, I’m happily too invested in Anytype to migrate to other platforms. Plus, I still can’t love Notion for some reason :sweat_smile:. So, probably I will continue use Anytype but I have to say, as my workflow and needs grow more complicated, I come face to face more and more this limitations and if these advanced features don’t get on the roadmap, I imagine myself and others like me would start thinking about moving to Notion, Obsidian, Capacities or even tana in the near future.


I totally agree with you.
I like anytype but I feel the limitations

Do not forget that the application would likely go way beyond when a few important features would be there like the API, self-hosting, and openness.

Obsidian is also well liked because a lot of people are making amazing plugins.

When making plugins for Anytype will be possible, a new era will begin :slight_smile:


Great post @dzlg, thanks for the thoughts!
I’m going to say something a bit general about “limitations”, feel free to add anything to this discussion because I find it really interesting!

I used to use Notion heavily everyday, I use Obsidian daily since last year, I tried Capacities and Tana looks really interesting:
They all can’t even compete with AnyType’s vision when it comes to “limitations” because of how different (and better in my opinion) is AnyType’s approach.

The problem I have with the other products it’s that they don’t miss features, they have a design which don’t allow those features I miss. So that’s why I like AnyType, because I see in its vision the potential to have everything I need (the infrastructure and sync protocol, object model, open code…).

Of course now you will get more features and polished experience if you go elsewhere, that’s why I’m still using Obsidian. Only time will tell if this is going to be too much of a bold bet when it comes to giving an actual usable product that can sustain itself in this sea of competitors but in the end I personally care about not getting stuck in bad design anymore so I’ll gladly keep using something else while I wait for AnyType OS :wink:


I like the idea of sub relations. Sounds similar to what’s suggested in Multi-Type Vs. Inheritance.

Thinking back on the first limitation that I experienced in Anytype (categorizing emails), I feel like this is also the solution that I needed. Back then I had to create different relations for different types of email (main, work, personal). With sub relations, all of those can be linked to one relation (email) which makes more sense.

I am also really looking forward to the calendar view. Ever since it was mentioned in one of the Town Halls, I’ve been waiting for it to show up in the roadmap! Improvements on the task management side, as well. A second brain workspace just feels incomplete without these things.

At the moment, my docs and databases are in Notion. But I only started using it some time before I joined Alpha, so it’s more like a temporary workspace for me until my use cases (and syncing) are all doable in Anytype.

On the topic of Movies/Media database, I would also like to add that the lack of grouping in Sets (Support for grouping in set view) is what’s very limiting for me. I have a Comics database in Notion, and it really helps that I can group them based on demographics besides having filters for the types of comics. In Airtable, my research database is also grouped by letters. Would love to see that kind of polishing here in Anytype.

Thanks for bringing up this discussion!


For sure! The future seems very bright for Anytype! Especially if the API is done right to give users Obsidian level of extensibility!


Thank you :blush:

Anytype’s vision and the ultimate Anytype’s OS is just something else :heart_eyes: But don’t forget that we’re not there yet. Given then fact that when Anytype goes public, it should rival and one-up other competitors in order to get traction and get enough support and attention to justify its business model for it to get to Anytype’s OS. That’s why I’m trying to address these issues so that it can be the best PKM, second brain tool in the biz! Right now the only area that it lacks severely are the database features and formulas compared to things like Notion, While it’s vastly superior in other parts such as layout, customizability, overall design and polish, etc and the while object/relation can’t fully captured by even Notion.

1 Like

Yeah my idea for subrelations and extending the capibilities of relations are inline with what has been said in Multi-Type Vs. Inheritance.

Yeah calendar and timeline view are a necessity going forward.

Although I currently don’t use and feel the need for grouping, but I can imagine it would be a huge plus for sets!



I agree with you but I personally care more about polishing, having a good workflow in the app rather than new functionality.
This would also apply to the “getting enough support” part because people should like the app for its basic features and stick with the project for its vision.

But as I said on the Multi-Type topic you linked the object model needs to be clearer, so we can at least know what we will get in terms of functionality. On this I completely agree.


I agree with you @kerbless. The focus is on delivering an app that is as bug-free as possible. Then we can start brainstorming on what features Anytype needs going forward, which features would be practical, and which features are going to be helpful longterm. At this point, it wouldn’t make sense to add a ton of extra features because it would likely break the app - that wouldn’t be a good thing and Anytype would crash before getting off of the ground. Personally, I can’t list any extra functionalities that Notion or Capacities have that Anytype doesn’t have, or if there are any, they certainly aren’t trade-offs. The focus should be on the vision - software takes time to build and it will be a while until we reach the last milestone in Anytype’s trajectory, but it will all be worthwhile. That’s why I’m super supportive of Anytype and won’t give it up because I realize that if we all remain dedicated to the vision, it will end up where we want it to go in the end. That’s the great thing about Anytype - we are a community! Unlike proprietary, closed source software or software that you have to pay for, we as community members actually have force in steering the way that Anytype goes. We can recommend and suggest features, and they get implemented. In the future, any developer will be able to contribute to the code source when they open the repos on GitHub. We are all building and shaping Anytype together. However, for now, the team has an excellent start that already has features that make it a winning choice against the alternatives. We just need to keep supporting the team and encouraging them in the process of creating Anytype!


I think linking relations in the way you mentioned is an interesting idea. For now, I was imagining creating many subject-home pages with inline sets where the task of creating objects is somewhat automated and includes the various relations and sub-relations automatically, but in your case, that wouldn’t work for a daily tracker functionality. I am a new user, so I haven’t run up against the limitations quite yet. Your movie database was a great inspiration to me, thank you for sharing this with the community!

I can’t imagine a way we move forward without formulas and a rollup-like functionality. Also, the timeline and calendar features will be a game changer for many reasons, I am more looking forward to this for the implementation of an historical events overview than a daily tracker or project manager. We are in two different places mentally at the moment, I am mostly grateful for the current functionality and having a blast refining my usage of the app, while envisioning what will be possible in the future as new capabilities are unlocked for us to use. Please hang in there, you are an asset to the community and I want to see the types of projects you design in the future!

I would like to ask your input if you would humor me. I am working through different variations of setting up something akin to wikipedia, that can cover all things in a generalized fashion, but also offer specific usability for individualized topics. Also, I want to incorporate the ability for future collaborators to be able to communicate back and forth on their thoughts on whatever detail, event, or subject they’d like to, by creating what I’m calling “claims” or “arguments” as objects, and relating them to the applicable topics. This would also function to refine the encyclopedia, as information is added/corrected over time, given some ability for the contributors to be able to interact with the objects and vote via a relation input or some other means.

I fully recognize that this functionality will require something akin to rollups, and also probably buttons, and many other small improvements in order to function properly, but my question relates to an overall organizational design problem. In what way could you imagine organizing all things/people/events/thoughts in the world into a single, coherent, and standardized system? What problems jump out at you that you think would need to be solved? Do you have any object types (going for the minimum number possible) in mind that would be necessary for a sufficient level of organization to emerge between objects that would make useful the information present in a system like this? My first instinct is that there may be designs that offer a handful of ways to view the same information, and that these types would be used to differentiate on the user side, not really for the input. For example, everything has some kind of date attribute, a physical object in the world was designed/invented, created, put on the market, used, re-designed, mentioned in an article, etc. So, all objects have some kind of timeline / calendar relevance, and I imagine(hope) that a standardized view of information could be used for eternal things such as “the sky” or “hydrogen” as could be used for an upcoming/past event present on your personal calendar.

I have begun to create most relations as object-type relations, I am assuming a notion-like function will eventually come that will allow self-referential sets to be automatically created using templates, and this will be necessary to create new objects that are systematically linked to the appropriate related objects upon creation. I understand my question is a very wide and open-ended one, but any input, small or detailed, would likely be helpful, as I am very early in my design process.

Thanks again for sharing your movie database, and I understand if you do not have the time or desire to respond to such a lofty question. Good luck with your projects, I look forward to seeing what you may share with us in the future!


Anytype’s stability and smooth day to day operations for the end user is and has been the no.1 priority. No debate there.

But as I mentioned, the things I touched upon are not necessarily extra features as much as core philosophy/functionality for building blocks of Anytype, i.e. relations/blocks/sets. Capacities is an interesting and fast developing alternative but it’s not at the level of Notion or Anytype. As for features, Notion has quite a few functionalities that Coda, Anytype, Obsidian or Tana currently doesn’t have.


I appreciate it :blush:

Thanks for the kind words :blush: Don’t worry my love for Anytype is not going anywhere anytime soon. As a matter of fact, for the past week I was looking for inspirations and ways to implement my ideal habit/fitness/task manager setup combined with journaling in a daily entry type and as much as I tried, I couldn’t do it with Anytype’s current tool :sweat_smile:. That was when I decided to write down this post and publish it.

As for your question regarding, to be honest I can’t say I completely understood it :sweat_smile:. Can you please repeat it and maybe give me an example/use case for the goal you have in mind?


Yes, I will say that having extra features to handle more specific workflows and scenarios will be nice in the future, but I’ve somewhat been holding out for making feature requests because I want to do it at a time where the devs will have the ability to implement the feature (I don’t want to bog down the developers who are already experiencng a large amount of incoming bug reports). I also have possibly not realized the full potentional of the aforementioned knowledge management tools because I didn’t play around with them too much. Notion was too stubborn for me to use for very long, Capacities had UI flaws and didn’t solve any extra problems that Anytype didn’t solve already, Coda was too restrictive, Obsidian is far too basic, Tana is in its early stages (so I haven’t used it), and, ultimately, I have grown to really appreciate Anytype for its slight ruggedness which makes it highly customizable. Also, this is all said from the perspective of a student - Anytype agreeably works better than great for hardcore (which I don’t get even close to) notetaking and general task management. Anytype is getting better everyday and I’m sure most of the features we request (within reason) will be added to the app eventually. But until then, it’s simply great to have a tool that (at least for me) pretty much beats its competition in multiple aspects.


Sure. Let me try. I have many pages of notes, some implementation attempts, and varying moments of clarity and confusion on the subject so forgive me for being unclear. Instead of trying to explain it in full, I will lay out some of the functions I want the system to be able to do. Keep in mind, this is not something that is possible today or even in the near future with the current features of anytype, but it should be at some point.

User A creates an object that is an opinion about a specific subject. For example “I believe that quantum computing will not have much of an impact on cyber-security because the community has been anticipating the changes and will be ready to implement new security protocols when the time comes.”

This statement relates to many things, which in this system, would be objects. Just a handful of them would be Computing, Quantum Computing, Encryption, Technology, Cyber-security, etc. You could think of many more relations if given enough time, some broader then others. (these may not be objects in the end implementation, but as I currently imagine it, it would work best as them. “Everything is an object”)

Also, this statement could have many forms of evidence detailed in the body of the object. It could reference an article, a published paper, or it could even reference a “claim object” this user had previously submitted to the system. It could also contradict or challenge a claim made by another user, and this can be indicated in the body of the object as well. Another detail about this specific claim is that it is a prediction, and can be useful to track in the future as this issue unfolds.

I want to use a framework of claim-explanation-evidence to encompass many(all?) other types of objects as well as a logical claim as detailed above. For a “table saw” object, the claim is essentially the name of the object, the explanation a description of it, and the evidence is the various references you can attach that relate to the tool.

There are also events, upcoming on the calendar, past events throughout history, etc. These are their own objects, as I’m sure you are familiar within your own system of using.

So, when you open an object, I want the following to be available for the object.

  • All “claims” made by users about the object, sorted by most recent and other methods. Claims can be intellectual or casual, they are just expressions of a thought of some kind.
  • All “claims” that this opinion contradicts that are currently in the system.
  • All “people” that are mentioned or directly related to the object. A person mentioned in a news article, the inventor of a tool, someone attending a calendar event, a person that works with a specific tool (paintbrush - DaVinci) etc.
  • A timeline of the past, present, and future events related to the object. The invention of a tool, previous iterations of a recurring event, news articles published that relate, previous claims made about a specific object sorted by date.
    (A note about opening a “user page” is that this information sorted by the user will build a profile of various beliefs, predictions, interests, etc. This will allow the system to function as a much more effective means of getting to know people based on their statistical similarities to you, for one example.)

Now, this would quickly become very detailed and the usability of the above information for a lot of objects would suffer substantially because of the quantity of information available. I plan on addressing this by sorting the data by various relations, and I have many ideas on how this may be possible, including a “self-moderation” system I want to try and implement that filters out the noise that many contributors would bring. One idea I have for keeping the system homogenized is by creating and updating a dictionary/glossary that tracks tags, definitions of terms, abbreviations, slang usages, etc. I imagine many dictionaries made by many different users that over time are combined and standardized in order to keep the decentralized system workable and able to be integrated with each other.

Given the enormity of this project, the all-encompassing nature of the information, and how ridiculous it may sound to try and pull off a workable system like this, please tell me all the ways it could fail. How would you go about organizing such a system? How would you moderate it? Segregate it by topics?

I do not expect to pull this off alone, I only hope to create a basic framework that hopefully other people are interested in contributing to, and allow it to grow over time. I am working with my own information now to try and test what I can with the features we have available, and hope to refine it over time and be able to have some meaningful progress on it by the time collaboration is available. Please do your best to tear down my hopes and tell me any problem you think this system has, it’s limitations, and of course, ways you think something like this could work.


Thanks for explaining in such detail! I’m glad you know what exactly it is that you want. Funny thing is that Microsoft Loop is by no means ready to compete with the like of Anytype or Notion but it’s comment/reaction ability to each block/loop component could be a key factor in you design system.

Anyway, let’s get back to Anytype. What you just described is somewhat similar to a system I designed for my physics research workflow. I will try my best to explain and draw parallel similarities for your case.

I’m on Mac and recently I come across a PDF reader app called skim which was the last piece of the puzzle for my research system. You see I created different types of objects including “research paper”, “lecture note”, “text book” and a template for faculties/researchers. What I do is I use simple pages for my different topics which you can also use for yours. Then, I would create and organize each topic based on what I feel right at moment and since we are not restricted to top-down folder structure we can mention various pages in different areas, so that’s no problem.

Then, I have put in place several key words such as [RP: Reference to Research Paper], [FP: From Research Paper] and so on to keep track of every piece of information that I either write it myself or borrowed it from these sources, then I would highlight these keywords and link them to their respective source which I created in Anytype. Then, In each of these sources, for example a research paper, I would write down the highlights from its PDF and use skim to generate deep links which I can link in Anytype so that whenever I click on it, it opens that specific part in the PDF.

Then I would use various relations to build inline sets in order to list all the publications for a specific person. you see where I’m getting at?

You can create several types of objects such as comments, claims, etc. with relations like “Person”, “related projects”, etc and then either write down the whole comment in the object that you’re tracking its development and then link it to the source comment, or just create a simple table with a row with that person’s name and the word comment which is linked to the object, or what I would do, is to use inline sets and filters to put all the comments made on that object sorted by time or any other factor that you want.

This way, you not only can create different filters for you comment/claim set, but also can use inline set to showcase all the comments a person made regarding everything in his/her own object page using inline sets.

As a last piece, hopefully when the local graph view is introduced in Anytype, or right now you can just use the Flow tab to keep track of all the related people, topics, comments, claims, etc with regards to your object.

I hope this can give you an idea to further refine it for your personal use :blush:. let me know if you want clarification in any area of it.


Thank you for the detailed response. I am encouraged by the design of your system, it makes me feel more confident in the possibility of something like this functioning in a shared and decentralized ecosystem in the future. I have to admit, I had a little anxiety about sharing this idea so early in my implementation because it seems like such an ambitious goal, but I had a feeling you were the right person to run this by. I have had a rough outline of this in my head for a while, and came across Anytype from a Youtube comment on one of the review videos on project management software options that are available now. I bet it was from one of the forum users, the comment was something like “What about Anytype?!?”

we are not restricted to top-down folder structure we can mention various pages in different areas

Exactly. This is central to this working imo. In the real world, everything is connected, we are all 7 steps away from Kevin Bacon.

PDF use skim to generate deep links which I can link in Anytype so that whenever I click on it, it opens that specific part in the PDF.

I came across similar functionality in my research in software designed for students to manage their research data, I forget the name of the app but it was a widely used one I’m sure you’re familiar. One way to describe this system in a nutshell is that it is a decentralized peer-review system that can encompass the entire world, and has little to no barrier to entry. I anticipate when the API is released that a PDF notation and linking capability will be available to us and this is exactly the type of implementation I will incorporate into the system.

Then I would use various relations to build inline sets in order to list all the publications for a specific person. you see where I’m getting at?

objects such as comments, claims, etc. with relations like “Person”, “related projects”, etc

Yes, I understand. One issue is that I want this functionality to be available across all objects, regardless of their type. For now, I am working in a limited data set and I will naturally expand into more broad types of objects and see where it takes me. One open question is where to find the balance between creating types and relations in order to organize the various types of information to be stored and accessed on the system.

An interesting thing to think about is how the ecosystem might regulate itself once the information is published and connected together. I imagine various moderators will emerge that can build their reputation in their respective fields of expertise, and conflicts between moderators can create paths for people to take to decide what their particular experience is going to be. Once rollups and formulas are enabled, I imagine tags can be utilized to create comparison tables that can differentiate between different moderators’ ideas, and also can bring to light interesting conflicts that reflect an interest of some users to quiet or censor some particular piece of information from spreading in the system. This is hard to fully imagine without collaboration being implemented, but nonetheless would be central to a community like this being able to be function in a useful way.

Thanks again for the input. As I continue down this path I will share my progress, I’m excited at the prospect of something like this coming to fruition.


Again, Thanks again for your kind words :blush:. I’m glad you find it useful and inspiring for your system. I get what you are saying about anxiety :sweat_smile:. I, myself am a bit of an idealist and although I created this post to talk about limitations and possible future past, I haven’t stopped thinking about the ways I can implement different scenarios with the current tool! And, I believe I made a little breakthrough with regards to that Oscars thing in my movie database :sweat_smile:. I’m still working on it but maybe I would share it with the community in the future.

This sounds interesting :thinking: I’ve searched and tried a couple of them like Zotero and research rabbit and consensus but I’m curious which app you are talking about :thinking:.

I see what you are saying. I have to admit; I, too spend an unhealthy :sweat_smile: amount of time coming up with the most optimized solution for my workflow with the least amount of types and relations. I find it most helpful to look at others creations and get inspired and also, not add any type or relation unless it’s absolutely necessary. Also, templates can be quite powerful! For example, right now I think I have around 6-7 templates for my type of human: Actor, Author, Director, Composer, Faculty, etc.

I agree.

you’re welcome :blush: Happy building!

1 Like

Zotero was one of them, I’ll have to look into the one I was referring to, it may have been that. I will get back to you on this. A significant part of what I am trying to build includes the integration of published papers, so I did quite a bit of digging into these applications looking for either inspiration or a program that would do what I’m looking for.

This confirms one of my thoughts, specifically with the human type, and also with what I’m calling “organization.” You have businesses, charities, government agencies, etc. The rabbit hole goes deep (public companies, private companies, technology companies, holding companies, LLCs, INC, etc), but templates has to be the way to go. This way, you have a fighting chance to make usable the object relations by restricting the object types upon creation to “organization” instead of trying to figure out all the specific types of groups that may or may not be relevant to that particular relation. Currently my object types are unrestricted, so you have to know what you’re looking for when you select the objects that become relations. And also, when doing this, you can pre-select your relations through the template process, so each sub-type is automatically indicated. This is the only workable way to indicate sub-types that I have come across.

One final thought I would like to share is that I’ve begun using a collection to create a dictionary/glossary for all of my relations/templates/types. This way as I move forward, I can reference back to the collection and keep my system consistent. You can utilize the filter system to create different views that are useful, like types, tags, templates, etc, and add a standardized “instructions for use” relation to them all, so you can remind yourself (and maybe one day other users) how to correctly create new objects, templates, types, and relations. With the collection, you can add a relation, and then go through them one at a time editing the values, so adding this particular relation would be the easiest this way.

Just typing this all out to you and communicating the ideas is extremely helpful for me in concretizing the ideas I have. Thanks again. I hope to see your solution to the Oscar problem soon!


Interesting! Published papers are an integral part of my workflow and right now I have entered over 300 of them.

I guess that’s one way to look at it! I, personally haven’t realized how to embed collections into my workflow :thinking:

Happy to chat :blush: You can DM me anytime you want for deeper analysis. Yeah, the Oscars, I think I finally settled on an approach which serves my purpose, it’s beautiful to look at and not a hassle to manually add everything. It still can’t beat the idea of linking relations with each other and having multi-fold relations but it’s something :sweat_smile:. I will share it while it’s popularized enough with Data :wink:

1 Like


I found out this post Ability to limit the scope of a relation, which is not too different to my idea of treating relations as object in a way that you can either link them or use them in a hierarchy; i.e. relation as animal with subcategories like dog breed.

Or for example in a workout scenario, you have a relation called “exercise”, you can not only choose “bench press” as an object for it, but also you can limit/relate other relations(aspects) like “reps”, “sets”, “duration”, etc to it.