🏠 back to Observable

first-party Collections are getting a bit confusing

I visited https://observablehq.com/@observablehq?tab=collections for the first time quite a while, and there are some things that I find a bit puzzling:

  • There are some collections that appear by default as a tile view but can optionally be switched to a (un-numbered) list view. However, other collections only show up as a numbered list and cannot be shown as tiles. These are clearly considered to be 2 different types by the platform, but to a new viewer there is no indication how or why this distinction exists unless you read the notebook introducing user collections (which mentions ordered collections), and it isn’t clear why the numbered collections can’t be shown in tile view.

  • Some of the collections are hierarchically nested inside each-other (you can put a collection inside another collection, and a sub-collection will have a link at the top of the page to go up one level in the hierarchy), but the sidebar list at left is completely flat (organized alphabetically by name) with nested collections showing up at top level, making the list very long (currently 60 distinct collections) and hard to navigate

  • Some of the nesting doesn’t really make sense. The county maps collection (which includes various choropleth, spike, and dot maps of miscellaneous types of county-level data for the US) is nested inside the elections collection, even though most of the maps have nothing to do with elections

  • The staff picks collection contains no notebooks directly, but instead contains 9 separate nested sub-collections, each with only one notebook inside. This adds a ton of clutter to the sidebar list. It seems like this would work better if each staff pick were added to the same top-level collection and the nesting were eliminated.

  • The featured creators collection links to a specific notebook for each author instead of to the author’s overview page (making it practically about the same as staff picks of specific notebooks), but it isn’t clear how each notebook was chosen: they aren’t obviously the best, most popular, most representative, most clickbaity, etc. examples of the authors’ work (or at least, I think Fil has a whole bunch of more exciting notebooks than the one picked).

Hopefully this doesn’t sound too critical or nitpicky. I think collections are a great and important feature. Personally I can make my way around them pretty well, but I suspect some newcomers are going to feel a bit overwhelmed.


I’m guessing that the single-notebook collections are a crutch to allow for custom descriptions. But yeah, not ideal.