Why a real machine beats a simulation every time
Simulations teach students to pass the simulation. Real environments teach them to do the job.
Simulated labs are tempting. They are tidy, they cannot really break, and they grade themselves. But students are quick to notice when an environment is on rails, and they learn to follow the rails rather than understand the terrain. The skills do not transfer, because the thing they practised on was never the real thing.
The messiness is the lesson
Real tools behave in ways a simulation smooths away. A command fails for a reason you have to diagnose. A config file is in a slightly different place than the guide said. A service will not start until you read the logs. Wrestling with that mess is the skill, and it is exactly what a first job will throw at a graduate on day one.
Give them the real thing, safely
The historical reason for simulations was safety and cost: you could not hand every student a real, powerful machine connected to the internet without risk. Cloud labs remove that excuse. StudentLabs gives each student a genuine machine, isolated so they can break their own setup and nothing else, and switched off when idle so the cost stays sane.
- Students use the same tools they will use in industry, at the same versions.
- Mistakes are recoverable: rebuild from the template and go again.
- Confidence comes from having actually done it, not from having watched it.
You do not learn to drive in a car that cannot stall. The same is true of the machines we ask students to master.