> 30 seconds vs 3-4 minutes on my previous laptop (14" M1 16GB) is a big deal. It was more constrained for memory (swapping)
This is why data driven purchasing is key. Running some tests and having some data to show how much time will be saved by a laptop upgrade makes the decision process much easier.
The companies that only decide based on prices and budgets set by someone making blanket decisions for the company always get it wrong.
It’s also possible to go too far on the spending path. I remember some people who demanded brand new maxed out MacBook Pros every generation until someone ran some tests and proved that it wasn’t making any noticeable difference at all year over year despite costing upwards of $5-6K per person. That’s money that could have gone to something else.
Many years ago I failed to convince my employer how buying their programmers less expensive computers with 5400rpm hard drives was an overall loss compared to 7200rpm disk drives.
My argument was that any programming necessarily entailed reading and writing dozens to hundreds files all the time, especially during compile cycles.
Maybe if I'd had data showing the actual time lost waiting?
Or maybe I was a dumb kid trying to justify my desire to have a nicer computer for my daily driver.
> Maybe if I'd had data showing the actual time lost waiting?
This is exactly what works.
Adding analytics that reports duration of test runs to a central server makes it easy. Some developers panic at the thought of this because it feels like spying, but the data is immensely helpful.
This is why data driven purchasing is key. Running some tests and having some data to show how much time will be saved by a laptop upgrade makes the decision process much easier.
The companies that only decide based on prices and budgets set by someone making blanket decisions for the company always get it wrong.
It’s also possible to go too far on the spending path. I remember some people who demanded brand new maxed out MacBook Pros every generation until someone ran some tests and proved that it wasn’t making any noticeable difference at all year over year despite costing upwards of $5-6K per person. That’s money that could have gone to something else.