Cross Docs - Delicate Flowers and Major Blockers

My team has two major goals that depend on cross docs:

  • Make it so team members of varying Coda skillsets can quickly “tap into” data from our main docs, such as our Library (knowledge base) and spin up their own coda docs and published sites using this data.
  • Make it so we can easily share data across all of our main docs (which we call “Hubs”)

We have some major problems in achieving this goal:

  • Cross Docs are too burdensome to set up. After bringing in a table:
  • Users have to hunt and peck to find all of the dependent look up tables. If we could drag and drop tables from one doc to another and then the platform would ask us “do you also want to sync these relevant tables that are being used as lookups?”
  • All look up or select list columns change to text columns and must each be set up again. All column types should be preserved.
  • All conditional formatting is lost. Conditional formatting should be preserved.
  • Users must manually change the default sync settings for each table from its most restrictive settings - “manual sync”, limited rows, etc. We tend towards having the highest refresh and most open table settings in our docs, and would have the option to default to that.
  • Cross Docs are “delicate flowers” too “fragile” for the robust growth we have planned. Changes to the source table like altering the name of a column do not ripple out to the Cross Docs and their formulas. We have a very complex cross doc system with different “owners” so this is very perilous for us and makes having distributed ownership of docs quite difficult. Having “smarter” change management ripple out from the central table would be marvelous.

Excellent points! If I may add:

  • Manual syncing can only be done by hitting the sync icon for a particular table, there is no way to control it (or multiple tables) with a button
  • When a user doesn’t have the right token loaded, the table doesn’t warn them nor give instructions, it just loads eternally
  • Two way cross doc has many gotchas
  • It’s slow
  • this, reiterated:
  • There is no easy way to export only a subset of rows to another doc, which means that if I have a doc with customer information in it, I can’t easily export only the information relevant to that customer

@Federico_Stefanato, I feel like you would also have some insight on this

Hi @Ben_Sutor: great post indeed. I like the metaphor of delicate flowers :slight_smile:
@Connor_McCormick, yes, I do have some thoughts in regards.

On a wider perspective, I think that the paths of cross-references and permissions are tightly bound together; perhaps different from how they are now.

Source of truth, data segregation, accessibility and - even - lineage are topics that sooner or later any data-driven application needs to tackle at some point.
However, I am aware that this requires a significant effort and it’s probably necessary to go with intermediate solutions.

What I love about Coda is that it allows so fast development that you reach any application’s critical points really quickly.
The drawback is that we need to be patient for the game field to be enlarged.