[2015-01-04] at this start WorkSelf occupied the cpu for about 5 minutes, before I decided to put the system into sleep (for about 2h). After that it behaved normally, that is, used only 1 - 4%:

What is WorkShelf doing in this time?
I know that in the period of reporting internet connection was down (problem at the cable provider). Now it is up again (as this note proves).

Good question, and, as you stated, it was apparently related to your connection being down. Does it happen again if you turn off your Internet access manually?

Jorge, I have now tested:
The problem occurs only if the internet connection is broken at startup of Winstep.
Just disconnecting internet while Winstep is running since a while does not pose a problem.

Thanks for checking this out...

It could be a 'false' case of CPU usage (i.e.; WorkShelf is actively relinquishing processing slices to the OS while it waits for an operation to complete, which make it seem like it is using a lot of CPU when it really is not). To test for this, run another CPU intensive app at the same time to max out the other CPU core and bring CPU usage up to 100%: does the computer feel 'sluggish'?

Jorge, will postpone this test to next week or so and then let you know.

Ok, thanks, I will try to reproduce here too in the mean time.

Jorge, I tested this

1. Start windows with ethernet physically disconnected
2. Disable WLAN
3. Exit Workshelf (which is automatically started with Windows
4. Start TaskManager
5. Start Winstep Extreme and observe cpu usage of it
6. For the period of about 5 minutes Workshelf behaves (max 12%)
7. To load the cpu start a flash movie in FireFox (> 50%)

In summa: the initially reported behaviour of Workshelf seems to be an exception or it happens only when windows starts Workshelf automatically and communication is down at that time. Since WLAN is used for VOIP also i do not switch this off at the router and hence 'no communication' happens only by accident from the cable provider.

I will come back to this if the situation appears again.