what was up with ntp again?
René J.V. Bertin
rjvbertin at gmail.com
Mon Mar 23 08:16:36 PDT 2015
On Monday March 23 2015 10:59:33 Brandon Allbery wrote:
>> Would it make sense to see if it works on OS X?
>Sure, if it does.
Erm, it'd make sense to see if it works if it does? Does build, maybe? :)
>I do wonder if the inexplicable delays I sometimes see in app start-up, or
>> apps being sig-stopped by something aren't due to discontinuities in the
>> system clock. (Yes, I have app nap
>SIGSTOP, I'd be tempted to wonder if that's taskgated refusing to let it
>run until it's been sufficiently checked for code signatures, etc.
Oh, I also have set taskgated to allow everything. Anyway, this happens to applications that use to run normally. The one I see it most often with is Bresink's Hardware Monitor Lite (I usually notice it because the menubar keeps claiming my CPU is at 97 degrees C long after the fans have settled down...) but the other day it happened to a normal (= non agent/LSUIElement) app. Curiously they're not listed as stuck or frozen by the Activity Monitor when this happens (so I'm not 100% sure it's a SIGSTOP that paused them), but a SIGCONT will get them going again.
Apparently app nap did use SIGSTOP/SIGCONT during an early implementation, but since I never had a 10.9 beta installed I shouldn't have a shadow of that code on my system.
And then there are applications that apparently aren't being throttled or app-napped, but still run much more slowly than they should; the background code parser in KDevelop is a prime example of that.
I realise it could be it has nothing to do with the clock, but with those newfangled memory features supposed to reduce swap usage. They're very good at that which is nice for people with small SSDs, but I think I prefer the days when over 8Gb of swap meant I really had to close or kill a bunch of tabs in my browser :)
More information about the macports-users