1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
|
Here's the wish list:
- Make the stack clickable so that clicing on a process will display
some details about the process.
- Allow user to switch between core and (core + swap) views. Shouldn't
be too difficult -- the infrastructure is already in place as far as
I can make out. Make a command-line option for the same.
- Make a help screen (key "h").
- Fix unbalanced drawables (currently the graphics and text seem to
prefer to hang around at the bottom of the window).
- Make a better icon/pixmap :)
Update: I've been sent a pixmap but have not incorporated it yet
since it'll probably necessitate my learning how to use libXpm. Some
time soon...
- The speed delay due to opening and reading files in /proc seems to
be minimal on my 486/66. However the process does take up a fair
amount of CPU when it's scanning /proc. Maybe I can find out some
way of getting these values from the kernel directly.
- sum(RSS) for all procs works out to more than physical memory when
memory is nearly or completely full. Leave this one for Linus to fix
:-)
Update: Apparently Linus is not to blame, it was my understanding of
the memory values which was faulty. I am not compensating for shared
memory (especially libraries) and hence the values which you see are
pretty approximate. Is there someplace on the web which discusses
process memory usage vis-a-vis the values found in /proc/process-id?
- Get a decent color combination for the bars, labels and background.
R G B is easy but not particularly beautiful. There's also a new
pastel color scheme (define USE_PASTELS in the Makefile) though the
contrast in the colors in that could be better.
- I have reports of gmemusage behaving erratically on systems with
128Mb of RAM. I forgot to stock up on those this quarter, so any
inputs will help. The problem is currently under investigation.
- SPARCLinux gmemusage doesn't seem to read the installed system RAM
correctly. Does stat(/proc/kcore) on SPARCS return something
different on SPARCS?
|
