I'm running the example code in a swift UI app inside a 1sec timer. While I can see load and memory usage updating, the cpu usage remains constant- 7% system, 12% user and 81% idle. While the numbers are actually fluctuating, they only do so at the 5th digit.

It looks a bit as if the raw OS counters are being exported, not the actual load?

• Partially done
• Missing some stats - current charge
• Needs cleanup

• Use for Apple wireless keyboard & mouse/trackpad
• Add on/off & connect methods for creating command line tool to make connecting to magic mouse in clamshell mode easy

• As seen in Activity Monitor
• Seems to be a custom calculation, no API for it
• Maybe something in IOPMLib

• As a reminder to factor this in (I/O Kit)
• It may be there's nothing we have to do. The burden is really on the client with having a privileged helper thats called through launchd
• This is default for iOS, so nothing extra needed

• See beltex/SMCKit#3

• Can't seem to handle kinfo_proc
• See compiler bugs 2259, 2260 - https://github.com/practicalswift/swift-compiler-crashes
• Doesn't seem currently possible, thus we use a bridging header, make the call in C, and return a simplified struct in the mean time

Additional Refs

• https://github.com/practicalswift/swift-compiler-crashes/pull/34
• https://github.com/practicalswift/swift-compiler-crashes/pull/35

Is it possible to make an iOS app with this source, or this is only for macOS?
Thank You!

libtop still has some PPC refs, is this really needed? See pid.c (proc type check).

• Everything should be the same for the most part (32-bit devices are many though - vm stats)
• Need cert to test

Should be an improvement over the raw I/O Kit calls, less prone to break, and more general (support for desktops).

Please update this for MojaveOS

Can't see a way to get around this bug in pure Swift. Anything besides simply initing the struct seems to cause a crash at runtime. This affects the System.uname() method. Only option I can see for now is to call a C wrapper function, but I don't think its worth it. It will likely be fixed soon, as we should be getting new Xcode beta's at the usual 2 week cycle rate for the next while.

Ref: https://github.com/practicalswift/swift-compiler-crashes/pull/76

Carthage is giving me an error while installation as

Skipped building SystemKit due to the error: Dependency "SystemKit" has no shared framework schemes

P.S. I've tried with other projects and they are working fine.

• Use debug macros
• Will add this at the end, don't want clutter for now

• Add install instructions as seen here
• This will setup test dir for us as well
• Note possible problem with iOS 7 frameworks

Hi, are u planning to upgrade the project?

• Related: beltex/dshb#19

I don't know if this is a bug or just me. But using the Battery method charge() doesn't seem to return the correct charge. It's about 5-6 percents lower than what it should be.

Rather than the classic -1, 0, etc. that can be confusing and inconsistent, optionals would be clearer and more "Swifty".

• For listing all of them
• All ops in Activity Monitor
  • Process ID
  • User
  • % CPU

  • Threads

  • Ports

  • Real Memory
  • Real Private Memory
  • Real Shared Memory
  • Kind
  • Sudden Termination
  • Sandbox
  • Idle Wake Ups
  • Energy Impact
  • App Nap
  • Send Bytes
  • Send Packets
  • Received Bytes
  • Received Packets
  • Purgeable Memory
  • Memory
  • Bytes Written
  • Bytes Read

Hi there,
it looks that the result of a call to the usageCPU doesn't update the returned values.
This my my fragment of code (99% identical to the documentation):

    var sys = System()
    let cpuUsage = sys.usageCPU()

    cpuUsageSystem.stringValue =  "\(Int(cpuUsage.system))%"
    cpuUsageUser.stringValue =  "\(Int(cpuUsage.user))%"
    cpuUsageIdle.stringValue =  "\(Int(cpuUsage.idle))%"
    cpuUsageNice.stringValue = "\(Int(cpuUsage.nice))%"