How to get CPU serial number using HISE?

[CatABC]

Is there a way to get the CPU serial number using HISE?

[Christoph Hart]

HISE | Docs

(docs.hise.audio)

[Dan Korneff]

@Christoph-Hart is it possible to have access to more parameters on the system ID generation?
I'm getting system ID changes when people perform OS updates, but there seems to be ways to only use the CPU/Motherboard to generate the ID instead

◆ getUniqueDeviceID()
static String SystemStats::getUniqueDeviceID	(		)	
static
This method returns a machine unique ID unaffected by storage or peripheral changes.

This ID will be invalidated by changes to the motherboard and CPU on non-mobile platforms, or performing a system restore on an Android device.

There are some extra caveats on iOS: The returned ID is unique to the vendor part of your 'Bundle Identifier' and is stable for all associated apps. The key is invalidated once all associated apps are uninstalled. This function can return an empty string under certain conditions, for example, If the device has not been unlocked since a restart.

[CatABC]

@Christoph-Hart Will the system ID change after reinstalling the system using this method? Is there a way to bind it to the hardware ID? In this way, even if the system is reinstalled, the system ID obtained will be the same

[ustk]

@Dan-Korneff @CatABC set JUCE_USE_BETTER_MACHINE_IDS=1 so it is more OS update resilient

[Christoph Hart]

What @ustk said, the default machine ID generation is a bit less robust and will change with major OS updates, but with this preprocessor it will be a little bit more stable.

Your copy protection system must cope with the scenario that computer IDs can change though.

[Dan Korneff]

@Christoph-Hart said in How to get CPU serial number using HISE?:

Your copy protection system must cope with the scenario that computer IDs can change though.

My protection system and I are in therapy. Hopefully we can cope soon.

[Oli Ullmann]

@ustk
Do I have to use this when I compile HISE, or do I use it when I compile the plug-in?

Thank you
Oli

[ustk]

@Oli-Ullmann Both, otherwise the ID created will not be the same during test and production.
Once you use this flag it's forever for compatibility reasons
It should be set to true by default imho, but I think it's for breaking old project compatibility that it's false... @Christoph-Hart ?

[Oli Ullmann]

@ustk
All right, that makes sense. Thanks for the info! :-)

[Dan Korneff]

I'm not sure what I'm doing wrong here, but HISE and my compiled plugin can't agree on a "better" machine ID.

I've compiled HISE like this:

and added this to my project:

When I launch the plugin, it detects that the ID is different and I update my license file. Then when I load HISE, it still thinks that the ID is different. What am I doing wrong?

[ustk]

@Dan-Korneff Have you tried to set it in the 'Modules` tab instead? That's what I do because I always think some pre-processor aren't working, that might be the case for this one...

[Dan Korneff]

@ustk I've changed the preprocessor in the module like this:

Still the same result

[Dan Korneff]

It appears to be something with the exported plugin. The machine IDs generated with JUCE_USE_BETTER_MACHINE_IDS=1 are the same as if it's set to false. At least I know where to start looking

[Dan Korneff]

Ok... I've obviously have not had enough coffee today. I'd probably have to export the plugin after I rebuild HISE for something to take effect.
Ignore my stupidity for now.

[ustk]

@Dan-Korneff yeah it can be troublesome to be sure both are working (hise and plugin)

[Dan Korneff]

I'm using JUCE_USE_BETTER_MACHINE_IDS=1 in my new project. It's mostly working well, but I have a couple Windows 11 users that have to reactivate their license every few reboots. Is there a way to list the devices that are used to create this number so I can debug what is changing on their system?

[ustk]

@Dan-Korneff Ouch! As it doesn't respect what the name says, that is very unfortunate!

[Dan Korneff]

@Christoph-Hart I think this is the issue:
https://github.com/juce-framework/JUCE/commit/120f9266dc881c81371109c69b37a0a246accdeb

Is this fix already implemented?

[Christoph Hart]

@Dan-Korneff welp it‘s time for a JUCE_EVEN_BETTER_MACHINE_IDS macro…