Skip site navigation (1)Skip section navigation (2)
Date:      25 Jun 99 03:39:07 PDT
From:      Jesus Monroy <jesus.monroy@usa.net>
To:        Mike Pritchard <mpp@mpp.pro-ns.net>
Cc:        doc@FreeBSD.ORG
Subject:   Re: [Re: Survey: online documentation 15% incorrect]
Message-ID:  <19990625103907.5722.qmail@nwcst314.netaddress.usa.net>

index | next in thread | raw e-mail

Mike Pritchard <mpp@mpp.pro-ns.net> wrote:
> > Document: Documentation errors in FreeBSD
> > Run Date: 6/17/99
> > Description:  The following is a listing of incomplete, ambigous or
> >               failed references for FreeBSD using whatit(1) and 
> >               man(1).
> >               This survey was conducted using FreeBSD 3.1.
> > 
> > A complete Listing can be found at:
> >
http://www.geocities.com/SiliconValley/Hub/8031/psuedo-indicators/fbsdutils.html
> > 
> > Score:
> > directory    no reference  ambigous reference  other  total  available
> > /bin                    0                   1      1      2         32
> > /sbin                   8                   5      0     13         82
> > /usr/bin               42                  33      0     75        394
> > /usr/sbin               8                   7      2     17        183
> >                                                        ----       ----
> >                                           total total   107        691
> > 
> >                                Final Score  107.0/691.0 = 0.1548 (15.5%)
> 
> I took a look through this list and by my count only about 55 of
> the references are bogus or incorrect or ambiguous. A number of them in 
> your list are producing exactly what they should be.  Some others
> have been fixed in -current already.
> 
    "exactly what they should be" is ambigous in itself. As this is
    the question I was begging. I expect a response as what you
    have outline. 

    My point in doing this was to have a place to cut back
    files on a custom install. I didn't know where to start
    so I took this as the best way. 

    I posted the results in the hopes that someone like
    yourself would make a decent attempt at resonding to
    some of the issues. Plainly many of the problems (or
    misusderstanding) stem for historical reason. 

    The are many fixes I would recommend so I'll list them
    as you have them below.

> There are a handful of man pages with links to other man pages that do
> not show the links in the NAME/SYSNOPSIS section.  I'm going to go fix
> the easy ones right now.
> 
    Thank you.

> A number of man pages/commands are links for historical names for 
> commands,
> which have been replaced by new/GNU replacements and the man pages
> reflect
> the updated versions, not the historical name for the command.  These 
> should probably be fixed, too.  'man cc' is a good example.  You get
> what you need - the gcc man page.  I've never been sure what we should 
> do with these.
> 
    I have a proposal that all new man pages carry a historical
    section in them. I could suggest a format, but I think 
    the matter is open for dicussion and I'll let others
    make suggestions. 

    In any case, the historical section needs to be implemented
    for all utilities includind those depricated.  This section
    should, of course, lead to (or mention) a section of man
    pages that lists the old man pages. This being there
    so if there is any question reference is available.

    Install time issues are a seperate can of worms
    I'll leave unopened at this time.

> One interesting problem your survery points out is that > whatis/makewhatis
> get confused about inetd.conf and inetd.  It lists them both as section
> 5 man pages, when only inetd is section 8.  This is because
> the inetd Makefile MLINKS inetd.8 to inetd.conf.5.  There are
> probably a few other commands that might be doing this.  I'll have
> to look at this closer.  
>
    Sounds like an uninitialized variable to me.
 
> Then there are the commands without manual pages, which is just plain
> wrong.
>
    Yes, I was surprised to find them. 8-O

--
    Another suggestion I'd like to make is that 
    apropos(1) and whatis(1) not perform similar functions.
    This seems to be a useless feature.

    apropos(1), as the word suggests, should return those
    that seem apporpriate; while whatis(1) should be more
    defined, at the very least it should return items on
    a scoring system, not alphabetical such as is.

    Most notably is perl(1) returns far to man items to
    be useful. 

--
    Another point is aliased program. I think we can 
    all recall the amazing discovery that '[' and 'test'
    are one and the same. However, this is not obvious
    even when reading the man pages.

--
    For both of these I would suggest new options
    to apropos(1). One option would be to list all
    aliases, as used in the file system. Another
    option would be to list all depricated facilities.

    This one absolutely needs to be done.
--
    The last suggestion I'll make is to add "section"
    definitions in the man(1). Currently, one must
    know (or guess) the sections. This is useful
    especially with fsinfo(1) and fsinfo(8).


---
"I'd rather pay for my freedom than live in a bitmapped, 
pop-up-happy dungeon like NT."
http://www.performancecomputing.com/features/9809of1.shtml



____________________________________________________________________
Get free e-mail and a permanent address at http://www.netaddress.com/?N=1


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-doc" in the body of the message



help

Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19990625103907.5722.qmail>