[BusyBox] next list of docs inconsistencies

Robert P. J. Day rpjday at mindspring.com
Sun May 2 13:45:43 UTC 2004


(coreutils directory, one directory at a time)

chroot:

	- might mention that it's illegal to chroot to "/"

dd:

	- synopsis line broken weirdly over two lines, check table format

env:

        -, -i   start with an empty environment
        -u      remove variable from the environment

	1) "-u" option doesn't emphasize that it takes a var argument

echo:

	$ echo "Erik is cool"
        Erik is cool

	1) example violates self-referential coolness level :-)

fold:

	-b      count bytes rather than columns
        -s      break at spaces
        -w      use WIDTH columns instead of 80

	1) "-w" doesn't emphasize it takes a width value

id:

	-c      prints only the security context (???)
        -g      prints only the group ID
        -u      prints only the user ID

	Does the id command really recognize the -c option (selinux)?  
	It's not obvious from the source, but perhaps I'm just reading it
	badly (the call to "is_flask_enabled()"?)

install:

	install [-cgmops] [sources] <dest|directory>  ...

	Fails to document the "-d" option.

	Spelling error in source install.c:
		"coulnt create"

mkdir:

        -m      set permission mode (as in chmod), not rwxrwxrwx - umask
        -p      no error if existing, make parent directories as needed

	1) -m doesn't emphasize it takes a mode as a value

mv:

	Makes no mention of "-i" or "-f" options.

rmdir:

	Online doc makes no mention of "-p" arg.  Shouldn't it?

sleep:

	sleep [N]... 

	This gives the impression that the sleep duration is somehow
 	optional.  AFAICT, it isn't, so maybe this should say

	sleep N

	P.S.  the way "usleep" does it.

uudecode:

	uudecode [FILE]... 

	Uudecode a file that is uuencoded.

	Options:

        -o FILE direct output to FILE

	1) Synopsis should mention possible [OPTION}, no?

yes:

	yes [OPTION]... [STRING]... 

	Repeatedly outputs a line with all specified STRING(s), or 'y'.

	1) OPTION?  What OPTION?


A couple more pedantic general observations:

1) In a number of cases, a command accepts either "-r" or "-R" for 
recursive behaviour.  Currently, I've seen this represented by either
"-R, -r" or "-r or -R".  Why not just "-[Rr]", for brevity?  And
generalize this for other options if other examples exist.

2) For clarity, it would be helpful to have command and output examples
with the typed command in bold, and the output in regular, as most books
do.  Yes, it's picky, but it would make things easier to read.



rday



More information about the busybox mailing list