checks 0.3.3
8 November 2001
(MTV makes me want to smoke crack)

Needed to be done:
1.	Use some sort of checksum checking instead of just 
	du to see if a directory has changed.  That way if
	a file changes, but the size of the file doesn't change
	we'll still be able to report the change in the directory.  
	Looks like I'll have to recursively list all the files in 
	a directory, then checksum each, add them up, and then 
	wait till the next loop.  Man, that's a pain.
	(Predicted for sometime near v 0.6)
2.	Make sure that a directory actually exists where a user 
	asks us to check.  This way du won't give our user errors, 
	and will overall make this app a little cleaner
	(Predicted for sometime near v 0.4)
3.	I'm debating whether to implement pthreads or fork off several 
	processes if several directories are being checked.  fork 
	raises issues with reading/writing to a preference file, 
	whereas pthreads aren't default on NetBSD, and that's bad, 
	'cause I do a great deal of work on NetBSD.  Stay tuned...
4.	Anyone want to test it on Linux, Solaris, SysV now that 
	it should compile on these platforms?  I know I don't want to.

Changelog:
changes to 0.3.3:
*	checks now parses the .checksrc the right way
*	checks is now also no longer uses fgetln, so 
	we've removed the *BSD dependency

changes to 0.3.2:
*	More intelligent memory management
*	No more temp file being dumped
*	(Almost) proper parsing of .checksrc
*	New n option allows Matt Slack to stop complaining.  :-)

changes to checks 0.3.1:
*	Far more efficient parsing of .checksrc
*	what line in binary
*	A few small speed increases/memory reductions

changes to checks 0.3:
*	Biggish re-write.  Now supports remembering sizes between exit of 
	app and next open.  Can check several folders at once.  Now 
	comes with manpage.  0.3 tends to take about 1/3 less 
	processor time, but checks never took much, anyway...just found 
	that curious.

changes to checks 0.2.1:
*	Got rid of a segfault caused by checks at quitting time.
	Was passing the wrong pointer as an argument in sigtermHandler.  
	Error only happened under NetBSD, so it took me a while to figure 
	out.

changes to checks 0.2:
*	Didn't check if malloc() returned a non-null pointer.  Given 
	that we're not allocating much memory, this shouldn't have been 
	a problem, but it's best to be safe.
*	checks will now delete your temp file at quit.  Good thing.
