New on Beta: IDE 1.47

@jamesb, the short answer is “no”, the behavior remains the same as before.

Here is the part where users might misunderstand - “Build & Run” associates the new code with all the devices within a model but ONLY restart the selected device so it downloads and runs the new code immediately. If you un-plug and plug (i.e. restart) another device in that same model, that device would also download and run the new code.

Our team is actually working on some changes to make this workflow clear and easier to understand. Stay tuned!

Thanks Kenny, that makes sense of what I was seeing.

Another question, I’ve suddenly started seeing some of my devices are not connecting, and when I make changes to code for those devices and click “Build and run”, the Agent log is not updating. I’ve just hit 110 devices on my developer account, and have 72 devices attached to the model in question. Have I hit a limit for developer accounts?

I’m also getting “No HTTP Handler” when I try to http to the agent. The devices were active a few hours ago, so it’s not due to the agent being inactive for too long

And sorry, one more question @kenny

If you un-plug and plug (i.e. restart) another device in that same model, that device would also download and run the new code.
Does 'restart' include going to deep-sleep and waking up?

@jamesb can you send us details? There has been no outage for either developer or production devices.

There’s no limit apart for your patience on how many devices may be connected to your development account (though note the commercial use exemption… it’s much easier to manage a fleet of devices with the commercial tools!)

…and yes, a deep sleep and wakeup will cause new code to be sent to a device.

@Hugo, looks like the devices/agents in question came back online about an 2.5 hours ago, so they were out for 24 hours. They were on the same network as other devices that were functioning fine throughout that period. Let me know if it’s still worth sending details, given that the issue isn’t persisting (so may be hard to debug - unless you have logs?)

We have logs.