Have you ever been able to solve a problem at work that nobody else could, or solve it faster, or cleaner, or more robustly, by virtue of having learned previously about some random topic which you had no idea would one day be applicable to your work? I think most programmers have, and I'd hope that a good boss at a good organization would understand that - as long as you're accomplishing everything that's required of the job - independent learning is a great use of (a reasonable fraction of your) time that will benefit both you and the business.

I once "solved" a tricky issue with a random bug in an FPGA processing data from DDR. The issue was hardware "malfunction": DDR can flip some bits if accessed in some ways.

No one on the team did know about that, no one believed my fix could work (a one line for a bug we investigated on and off for months). It did. I had read a computer security article about how row hammer can be used to gain privilege or something and made the connection.

During my yearly evaluation, I mentionned it as the most impact that year. I was nearly laughed at. This bug precluded the product from functioning more than a few minutes, and we targeted at least several hours without any glitch. No one noticed that without me in the team, the product was dead for a few more month at least.

I left a year later. The company is dead now.

Wow that’s an incredible fix! What an awesome connection to make.

It depends on what I am learning - if it is required training or continuing education, that is part of my job and I do it on company time. If it is for my own personal projects, hobbies and interests, it is my own time.

The short answer is, it should be company time (to at least some degree). I have development as part of my work plan and, while I'm a bit flexible about things, I have absolutely been spending ~half-days this summer on predominantly classes. Some of those more or less directly feed into other work I do, others less so.

But to your broader point, if I need to step back to do some learning for a work task/presentation/etc. I absolutely consider that to be on the clock.

I learn stuff on my own time. If I need more time and the learning phrase is directly related to the my work then I can ask manager to reduce workload. But usually I don't need to ask. I also believe that self-learning in free time is standard expectation for our industry (software). If I run a company later, I would expect the same for my hires.

IMHO this type of learning can be work as long as it aligns with what the company needs. Dicking around with Rust because you think it might be useful one day doesn't count. There's a grey area for sure.

I also think that it's healthier to strive for setting+hitting ambitious outcomes, and thinking less about literal hours that you're working.

(My blog co-author wrote a post about a very closely related topic - the TLDR is that it's really a manager's job to look for ways to train their team by aligning learning areas with what the business needs. Without that alignment, you'll always have the tension that you called out where it can feel like you're not doing "real work" while learning: https://staysaasy.com/management/2020/08/01/Growth-Paths.htm...)

I think a bit of both to be honest. I don't generally surf the internet while in working-hours, but I always make an exception every Thursday to read the current LWN release.

Computers are my hobby, not just my job, so I feel that I research and experiment in my own time. But I'm reasonably happy to experiment at work, in short doses.