Domain For Sale
This domain is for sale.
If you have better use for our domain name than we do, please click here.
Say you’re semi-nomadic and you spend occasional weeks living out of your own home, your romantic partner’s home, your little cabin in the woods, and even sometimes your truck, which is decked out to be a living space all of its own.
We’ll bet your typical grocery list is full of annotations about what location each item is destined for:
• Now what if you need to understand how many of each item you have at each location?
You could use a spreadsheet. We probably have far fewer locations than items, so let’s put locations along the top as column headings, then items down the side as row headings, and at each cross point you put a count of how many items you have.
• Then I ask you to track both haves and needs at each location.
You duplicate the tab in your spreadsheet. You even make a third tab and call it differences.
• Now we want to sub-divide our locations into sub-locations. For example, our home has an upstairs and a downstairs, and we keep paper towels in both.
Getting harder. We have lots of items, so we made them rows, but now we need lots of columns too. Also, rows can be indented to show which ones are sub rows, but that’s hard to do with columns.
• Let’s build modular kits of items too.
Now we’ve got a hierarchical tree of locations and another hierarchical tree of items.
I’m pretty sure no spreadsheet ever made provides a good way to handle this data.
The idea is more generalizable than items and locations, we’re sure, this is where we’re starting. (Our lead developer is indeed semi-nomadic.) But even from this one case we can learn a really important thing:
When aggregating data, sometimes we need to sum down the taxonomy, sometimes we sum up.
Imagine our item and location taxonomies are these very short examples:
- First Aid Kit
- Box of Bandages
- Bottle of Painkillers
Let’s start with items: we’re going to want to aggregate counts downward.
That is, if we have a first aid kit, we can assume we have a box of bandages and a bottle of painkillers. But conversely, we would not want to sum upward: having just 1 bottle of painkillers does not imply we have 1 complete first aid kit.
With locations, the opposite is true: we want to aggregate counts upward.
That is, if we say both our garage and kitchen have a first aid kit, then our overall home does have 2 first aid kits total.
In other words, the math we need to apply when generating reports on our taxonomy depends entirely on human principles.
To a computer, one list of things is usually treated exactly the same as any other list of different things. But when our things are more specific items and locations, we bring with them our human definitions of what those words mean, and our inherent assumptions on how they should be counted.
If we were to naively build an app for taxonomic data where one tree of things is seen as the same as any other, our resulting math would be non-useful. Only by taking into account some real-world use-cases can we design a tool that’s truly practical.
So let us know what other examples you can think of!
We’d ultimately like to make our pivot tables work with everyone’s tree-like data, not just items and locations. There may be more subtle human assumptions like these that we need to be aware of.
~ The Octoboxy Team
This project needs sponsors!
We really want to build this app, but we’re building it only for ourselves. Unfortunately, projects that we’re paid for always take precedence over the un-funded ones.
Please, write to us today if you want to help out, or even join our team!
We need only a little help and this project will come back to life!