(KainX)
Basic cleanups on the config stuff. I made sure that all epplets using the
config subsystem are using atexit(), and I added the same re-prioritizing
code that my epplets have to E-Cpu and E-NetFlame so they wouldn't be as
CPU-intensive. E-NetFlame has also been upgraded to v0.3 with bandwidth
patches from Shawn (the author), and it also has a config file option for
which interface to monitor (default "eth0") for those who might want to keep
an eye on ppp0.
SVN revision: 1037
(KainX)
Okay, don't kill me. I had to change the API again because it was broken.
Epplet_load_config() now takes no parameters. I also renamed some functions,
which the following command should help you do:
perl -p -i.bak -e \
's/config_data_with_def/config_def/g; s/_config_data/_config/g;' \
<files>
That all goes on a single line, BTW. And replace <files> with the names of
your .c files. You must now specify any defaults by calling
Epplet_query_config_def() instead and passing the default to that function.
I also added the ability to right-justify label text by specifying a negative
x coordinate. See E-Time for a sample.
SVN revision: 989
(KainX)
First off, I removed some unneeded variables from E-MemWatch. I also fixed a
reporting bug in E-Disk.
But most importantly, I fixed Epplet_change_label() to avoid changing a label
to the exact same string. This is the best place to handle this situation,
because it saves resources (we already have both strings to compare) and, due
to the nature of epplets, the probability is very high that this function
would only (or at least primarily) be called from a timer callback. So it
simplifies client code this way too.
SVN revision: 983
(KainX)
E-Disk, an epplet which gives you a numerical and graphical view of your disk
I/O. It also keeps persistent bandwidth data, so the longer you run it, the
more accurately the percentage bars will portray your I/O bandwith usage.
SVN revision: 969
(KainX)
Added Epplet_query_config_data_with_def(). See the header file for syntax.
For some reason, the compiler didn't like it when I spelled out "default."
SVN revision: 964
(KainX)
Lock files now store PID's so that stale checks can be made. It seems to
work here, but I'll need lots of testing.
PLEASE remove all .lock files from ~/.enlightenment/epplet_config/*/ and close
all running epplets before running any new ones!
SVN revision: 963
(KainX)
Alright, I've got the config stuff working as planned. Epplet_Init() has the
same parameters as it used to. If you use config files, make sure to call
Epplet_load_config() AFTER calling Epplet_Init() but BEFORE attempting to
access config data.
Also note the function name changes: Epplet_save_config(),
Epplet_query_config_data(), and Epplet_modify_config_data().
SVN revision: 951
for instance accounting, every epplet has its own directory in
~/.enlightenment/epplet_config/<epplet-name>, transparent config file
handling is provided. Users only need to query for settings and change
settings. Config settings are saved automatically on exit, which
means that users MUST call Epplet_cleanup() before exiting. Every
epplet instance has its own config file. Also added a mechanism for
defining default settings.
Updated all epplets accordingly.
Look at ConfigTestEpplet.c for the details. It shows what instance it
is running as and keeps a record on how many times that instance
has been run in the config files.
SVN revision: 949
(KainX)
Added 7of9 mode to E-Biff just for raster. :-)
Also fixed some bugs with the config file stuff. If you're going to parse a
config file, *honor* it for pete's sake! =P
SVN revision: 885
but I haven't tested it enough yet.
The idea is to run
Epplet_load_config_file("~/.whaterver");
assuming that the config file looks e.g. this:
a_number 123
some_word hello
my_string hello world
and then query for the keys in the first column using
char * s;
s = Epplet_query_config_file("my_string");
if (s)
{
/* s is now "hello world". */
}
Hope that's a good idea.
Cheers,
-- Christian.
SVN revision: 838
Michael Jennings
Carsten Haitzler
Mandrake
Christian Kreibich