As a sugestion, store the sounds in a soundlib tree, hashed or categorised (boolean (yes, no, true,false, up, down etc.),numbers, caledar(day, date, time etc), state, weather etc) and dont duplicate any sounds then make a sounds tree with virtual categories and sim link to the files needed.
This keeps the directory sizes down and allows for sound sets to be built up with all the words they use in them.
It also allows sounds to be added as needed rather than requiring all sounds to be part of a distribution.


Robert Hajime Lanning wrote:

<quote who="Tilghman Lesher">


Although the OS may cache that information, the userland process
can take quite some time to process a very full directory.  I've had
this happen quite a few times with Linux ext2 filesystems, where the
fileglob * exceeded bash's limit of 32,768 characters.  /bin/ls on
those directories took several minutes before the first results were
given.

I'll additionally comment that the directories I was working with were
not normally that full, but was a side effect of a process dumping
lots of little files into a directory when something went wrong.

On a slight tangent, NT4 had a practical limit of about 300 directory
entries before attempting to process the directory became unbearably
slow.

-Tilghman



A couple of things, searching a directory for a specific name tends to be a linear search through the directory (unless the filesystem uses binary trees, like ReiserFS...), "ls" is a bad example of a command, it is more of a worse case example.

ls will read the entire directory, sort it, then do a stat() on every file
listed.  All of this is done before it formats the output.  So, you have to
wait until it is all done, before you see the first character output.




_______________________________________________ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users

Reply via email to