Dense on purpose
SolidFire’s Volumes grid packs thirteen columns into forty-pixel rows. It asks a lot of a first-time user and gives a daily one a whole screen of volumes to compare, and choosing which of them to design for first is the real decision.
The Volumes grid I designed at SolidFire has thirteen columns at forty pixels a row. It’s dense on purpose.
SolidFire sold all-flash storage, and a single cluster could hold thousands of volumes. The Volumes view is where a storage admin spends the day, and part of that day is answering an application team whose database just got slow. Which volume, and why? So each row carries capacity, efficiency, snapshots, current I/O, latency, throughput and quality of service, and a scan down the list finds the volume worth opening.
That has a cost. Thirteen columns at forty pixels a row ask a lot of a first-time user, and there’s no gentle way in. They also give a daily user a whole screen of volumes to compare at once. That second half is why I’d make the same call again.
Who goes first
Every dense tool comes down to one choice: which person you design for first.
The first-time user and the daily user want opposite things from the same screen. The first wants room, fewer columns and the detail a click away. The second wants everything in view, because comparing is the job, and you can’t compare what you have to click to see.
The first-time user is new for a few days. The daily user is there for years. A grid built for the first visit charges the expert a click or a scroll on every visit after that, and for a storage admin, the visits that matter most happen under pressure, with someone waiting on an answer.
So the grid is built for the person who opens it every morning. Density is how an expert sees a lot at once.
Dense has to be calm
A dense screen only works if most of it is quiet. In the Volumes grid, color stays off until it means something, so a page of calm rows lets the one in trouble stand out.
The other half is not losing your place. The previous UI opened a volume on a screen of its own. In this one, a volume’s throughput, IOPS and latency open under its own row, and the rest of the list stays where it was. Six thousand rows deep, losing your place is the expensive part. The drilldown pushes the rows below it down instead of opening over them, so the list moves, and nothing it held is hidden.
Both of those make density cheaper for the expert. Neither makes it easy for a newcomer, and I didn’t try to. A row that tried to explain itself to the newcomer would be a little worse for the admin every day.
The mistake that runs the other way
I see the opposite trade more often than this one: a tool people use all day, designed to look good the first time someone opens it. Generous spacing, four columns, the rest behind a click. It demos well. Then the people who live in it pay for that first impression on every visit for as long as they use it.