Lookup column provides us with a 1-way “relationship”, as in, we can set a relationship from table A to table B. However, a proper relationship column would be one that automatically creates a column in the referenced table that holds all references to a row in that table.
I.e.: if I create a Parent Relationship to a Table A from Table B, Table A would get an auto-generated column relationship back to A. I could then alter the name on that column afterwards.
This would allow editing this relationship from both perspectives.
Reduce the number of Lookup columns, improving performance. (as this would be called at Edit/Alter lifecycle and be made performant.)
Reduce the number of Automation rules that are currently required to maintain an editable lookup relationship, improving performance.
For the community: There is a “hack” (?) where if you use a “Text” column instead of a “Lookup” column driven through a formula, then you can dynamically add rows to the target relationship through the Detailed View of a specific row.
Hey @GJ_Roelofs, I have been reading your posts today as I saw you commented in a few areas I have visited re: challenges with “user friendly” aspects of Coda naturally linking and referencing data across docs. Today I created what is probably a simple-minded suggestion in comparison to some of your posts, but I think we are talking about similar functionality:
that the Coda product team is working on some solutions you were asking for.
In my particular case, I am probably getting most bogged in Coda with challenges associated with these related issues. I am a manager with minimal technical skills, who has been responsible for selection and implementation, on a basic level, of a Project or Task Management app in my software development start-up. I am pretty easily able to get around Jira, Monday, Wrike, Teamwork Projects, ClickUp, etc., even stuff like TargetProcess and Pivotal Tracker, and I’m fairly adept with AirTable. But it has been a huge challenge for me, not being a developer or db expert, to get Coda to a point where it is user friendly enough and simply “working” as I’d like for me to introduce it to my team.
I am curious, with Coda 2.0 in motion and a growing userbase, how much a user of my limited skill is going to be a target Client/Maker growth-wise for Coda, because right now due to limitations such as the ones you highlight I’m not sure I can manage to accomplish what I’m after. It remains overly daunting for me to get to the finish line pieces of my Team Management doc that I have been working on for months. In order for me to effectively set up Coda, even with the mass of assistance I’ve gotten from their team - which I have to say provides the best support I’ve ever seen - I think I would need some of the functionality you speak of to be present. That is, a lot of relational functionality, including hierarchical set up with concepts like subtasks, that can be used in a more “digestible” way for less technical users as myself. As it stands, even the solutions the Coda team provides to me usually turn out to be too complex for me to take forward on my own without ultimately getting stuck again.
And I did want to point out that Notion, which a lot of people consider the alternative to Coda, does actually automatically offer up creating a lookup in the destination table.
Thanks for the comments @ABp. We are definitely building Coda so that no matter what skill level you have with relational data, you’ll be able to build a doc that meets your team’s needs. The “team hub” use case is one of our primary goals for Coda, so expect to see features that make it simple to get started in Coda and from there, you’ll be able to build more complex workflows in your doc.
Hey @Al_Chen_Coda really appreciate you chiming in, and great to hear! I have faith that you guys will make Coda easier and easier for non-dev types like myself as you evolve the UX.
I have been posting a bunch about this in the community lately, so hope that is helpful to you guys, and sorry if it’s been a bit much. I am very eager to try to master Coda all I can, hence all my questions and suggestions!
@BenDavis Hey Ben, thanks for jumping in. You are right and I have been using this (relatively) new feature a bit, it’s very nice. Stuff like this is what really keeps my faith that you guys know what you’re doing!
I will keep pushing for more and more simplicity with this stuff, but of course you have a growing user base tugging you in various directions for how to develop Coda, so I fully understand much of what I like may never get developed!
And sorry, is there a link or image of this (quoted below)? I don’t see one, but would be curious to get a look at it. I am learning this stuff so eager to get my hands on all the examples I can!
I agree in that this still is not the same as auto-generating the relation. I agree @GJ_Roelofs with the pain points you raise in the very beginning of this thread. The fact that Coda now kind of “suggests” how to set up in the destination table the reciprocal lookup is helpful, but still limited functionality.
I posted probably a primitive suggestion for a column of type “row” that might solve this. I assume as Coda moves along the types of Columns will increase, seems inevitable evolution of added functionality. So perhaps it would be possible to do something with more “wide open” relationships within a new Column format.
@BenDavis, sorry to bother you again on this, but I have been struggling and I thought I was missing something. I just wanted to clarify that it is still not possible, per @GJ_Roelofs discussion here:
aside from the text column/lookup “hack” described.
When you mention "you can create the relationship in two clicks: below:
You are only talking about using the “blue dot” suggestion to add the lookup to the original table, correct? But ultimately you cannot still have both related columns in both tables each set up to 1) add a row to the other table, and 2) automatically show all related rows without manually linking. In other words, the Bi-Directional interaction from both sides, which @GJ_Roelofs has asked about in a few places.
One reason I’m asking is that although @Al_Chen_Coda responded here about a month ago, the original statement Al made that this functionality is not possible is back from August, months ago. I just want to make sure there is no way to do this still without Automation.
And to make sure it is perfectly clear what I’m talking about, I’d like to be able to set up the reciprocal relation between tables as easy as Notion or Fibery.io allow.
I appreciate any time you could take to clarify - again, this has been tripping me up quite a bit lately.
I looked at a bunch of tools before settling on Coda, which was by far the easiest and most powerful for my purposes. But the issue under discussion is one of the weaker portions of the platform.
I would second the suggestion (made by @ABp) that Coda take a look at Fibery.io, for the way it handles the relationships between tables. The UI is really clear and easy and highly functional. Basically, you just say “I want a relationship (1-1, 1-many, many-1, many-many) between these two domains” and then you spell out the details as you go. (Also, they have network representations (boxes and lines) of the relationships, which is a feature I would also love…) It is more table-centric than column-centric, which is a change, but my sense is that Coda is architected as table centric behind the scenes anyway
Thanks for weighing in. I really think this would be a huge boon to Coda, and hope the team is considering implementing it. Something even to start where if you @mention a row or page, that “destination” row will get a reference to that mention, like “mentioned in.” This could evolve into all kinds of additional functionality, but I am really missing that basic capability right now. The existing @mentioning is so easy, it just seems like it’s 1/2 done without the mention getting picked up in the “mentioned” entity.