1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/python/psutil/TODO Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,95 @@
1.4 +TODO
1.5 +====
1.6 +
1.7 +A collection of ideas and notes about stuff to implement in future versions.
1.8 +"#NNN" occurrences refer to bug tracker issues at:
1.9 +https://code.google.com/p/psutil/issues/list
1.10 +
1.11 +
1.12 +HIGHER PRIORITY
1.13 +===============
1.14 +
1.15 + * #387: system-wide connections (netstat).
1.16 + * OpenBSD support.
1.17 +
1.18 + * #371: CPU temperature (apparently OSX and Linux only; on Linux it requires
1.19 + lm-sensors lib).
1.20 +
1.21 + * #250: net ifaces speed.
1.22 +
1.23 + * (Linux) resource limit get/set - see man prlimit.
1.24 +
1.25 + * Process.name on Windows is slow:
1.26 + http://stackoverflow.com/questions/6587036/
1.27 +
1.28 + * Windows binary for Python 3.3 64-bit.
1.29 +
1.30 + * #269: expose network ifaces RX/TW queues.
1.31 +
1.32 +
1.33 +LOWER PRIORITY
1.34 +==============
1.35 +
1.36 + * #355: Android support.
1.37 +
1.38 + * #276: GNU/Hurd support.
1.39 +
1.40 + * NetBSD support?
1.41 +
1.42 + * DranflyBSD support?
1.43 +
1.44 + * AIX support?
1.45 +
1.46 + * examples/pidof.py (same as 'pidof' cli tool)
1.47 +
1.48 + * examples/pstree.py (same as 'pstree' cli tool)
1.49 + * get_threads() should also return thread names in order to implement it
1.50 +
1.51 + * examples/taskmgr-gui.py (using tk).
1.52 +
1.53 + * system-wide # open file descriptors:
1.54 + * https://jira.hyperic.com/browse/SIGAR-30
1.55 + * http://www.netadmintools.com/part295.html
1.56 +
1.57 + * Number of system threads.
1.58 + * Windows: http://msdn.microsoft.com/en-us/library/windows/desktop/ms684824(v=vs.85).aspx
1.59 +
1.60 + * #357: what CPU a process is on.
1.61 +
1.62 + * thread names:
1.63 + * https://code.google.com/p/plcrashreporter/issues/detail?id=65
1.64 +
1.65 +
1.66 +DEBATABLE
1.67 +=========
1.68 +
1.69 + * [Linux]: process cgroups (http://en.wikipedia.org/wiki/Cgroups). They look
1.70 + similar to prlimit() in terms of functionality but uglier (they should allow
1.71 + limiting per-process network IO resources though, which is great). Needs
1.72 + further reading.
1.73 +
1.74 + * cpu_percent(): current default interval is 0.1 so that by default it will
1.75 + produce a meaningful value. It represents a trap in case the user iterates
1.76 + over multiple processes though, as it introduces a big slowdown.
1.77 + Should it default to 0.0?
1.78 +
1.79 + * Rename connection ntuple's fields 'local_address', 'remote_address' to
1.80 + 'laddr', 'raddr' (note in accordance with http://bugs.python.org/issue17675)
1.81 +
1.82 + * Process per-cpus percent (XXX Windows only?), see:
1.83 + https://groups.google.com/forum/?fromgroups#!topic/psutil/ErrKTxAbu50
1.84 +
1.85 + * Should we expose OS constants (psutil.WINDOWS, psutil.OSX etc.)?
1.86 +
1.87 + * Python 3.3. exposed different sched.h functions:
1.88 + http://docs.python.org/dev/whatsnew/3.3.html#os
1.89 + http://bugs.python.org/issue12655
1.90 + http://docs.python.org/dev/library/os.html#interface-to-the-scheduler
1.91 + It might be worth to take a look and figure out whether we can include some
1.92 + of those in psutil.
1.93 + Also, we can probably reimplement wait_pid() on POSIX which is currently
1.94 + implemented as a busy-loop.
1.95 +
1.96 + * Certain systems (XXX figure out which ones exactly) provide CPU times about
1.97 + process children. On those systems Process.get_cpu_times() might return
1.98 + a (user, system, user_children, system_children) ntuple.
