Bug 17017

Summary: [PATCH] [SMALL] Clarify inetd's logging behavior
Product: Base System Reporter: Doug <Doug>
Component: binAssignee: Sheldon Hearn <sheldonh>
Status: Closed FIXED    
Severity: Affects Only Me    
Priority: Normal    
Version: 3.4-STABLE   
Hardware: Any   
OS: Any   
Attachments:
Description Flags
file.diff
none
inetd.8.diff none

Description Doug 2000-02-27 09:30:02 UTC
	The inetd.8 man page is not clear about how, where, and
under what circumstances connections and attempts are logged.
This patch clarifies it.

Fix: Apply the following patch. This patch applies to both
-Current and -Stable, in both the CVS and reality. :)

How-To-Repeat: 
	man inetd
Comment 1 Sheldon Hearn freebsd_committer freebsd_triage 2000-02-29 17:29:09 UTC
Responsible Changed
From-To: freebsd-bugs->sheldonh

I'll take this one. 

Comment 2 Sheldon Hearn 2000-02-29 17:33:41 UTC
On Sun, 27 Feb 2000 01:22:34 PST, Doug@gorean.org wrote:

> +Either wrapping option will cause all connections to be logged to the
> +.Dq auth
> +syslog facility.

All connections?  Are you sure?  I use -wW and I only get _refuseds
connections logged via syslog(3).

Ciao,
Sheldon.
Comment 3 Doug 2000-03-01 06:01:21 UTC
Sheldon Hearn wrote:
> 
> On Sun, 27 Feb 2000 01:22:34 PST, Doug@gorean.org wrote:
> 
> > +Either wrapping option will cause all connections to be logged to the
> > +.Dq auth
> > +syslog facility.
> 
> All connections?  Are you sure?  I use -wW and I only get _refuseds
> connections logged via syslog(3).

	Hrrmm.. good point. I have always used the -l so I just added the -wW.
With -lwW successful and unsucessful logins are both logged to auth. I
think the change from daemon to auth for -l is what threw me, and makes
it that much more worth documenting. 

Good catch,

Doug
-- 
"Welcome to the desert of the real." 

    - Laurence Fishburne as Morpheus, "The Matrix"
Comment 4 Sheldon Hearn 2000-03-01 07:27:12 UTC
On Tue, 29 Feb 2000 22:01:21 PST, Doug Barton wrote:

> With -lwW successful and unsucessful logins are both logged to auth. I
> think the change from daemon to auth for -l is what threw me, and makes
> it that much more worth documenting. 

So, um... care to reword?  I must admit that I'm not really seeing any
inaccuracy in the manual page, just a lack of detail.  Am I right in
thinking that it's only the lack of detail you're trying to address?

Ciao,
Sheldon.
Comment 5 Doug 2000-03-01 08:02:53 UTC
Sheldon Hearn wrote:
> 
> On Tue, 29 Feb 2000 22:01:21 PST, Doug Barton wrote:
> 
> > With -lwW successful and unsucessful logins are both logged to auth. I
> > think the change from daemon to auth for -l is what threw me, and makes
> > it that much more worth documenting.
> 
> So, um... care to reword?  I must admit that I'm not really seeing any
> inaccuracy in the manual page, just a lack of detail.  Am I right in
> thinking that it's only the lack of detail you're trying to address?

	Yes, exactly. People who are seriously interested in getting the logs
regularly (couple times a month) send the "I enabled -l for inetd but it
doesn't log" questions to the mailing list. I shudder to think how many
just give up. 

	Attached is a diff to the latest.

Thanks,

Doug
-- 
"Welcome to the desert of the real." 

    - Laurence Fishburne as Morpheus, "The Matrix"
Comment 6 Sheldon Hearn 2000-03-01 08:27:42 UTC
On Wed, 01 Mar 2000 00:02:53 PST, Doug Barton wrote:

> 	Attached is a diff to the latest.

Thanks!  I've committed a slightly mangled version of your patch.
Please take a look at it, because you are quite involved in contributing
to the project and you may as well see what was "wrong" with your diff
now. :-)

Basically, your patch broke two rules of thumb:

1) Lines should be short enough such that typical mailers would not wrap
   them (i.e. <= 72 characters).  When a sentence is too long to fit on
   a single line, try to break it on sentence fragments, particularly
   immediately following punctuation.

   This makes diffs easier to review later.

2) Sentences should start on a new line.

   The whitespace that separates a sentence from the preceding sentence on
   the same line is called a hard sentence break and tends to degrade
   the spacing in typeset versions of the document.  Hard sentence
   breaks also tend to make for bigger diffs later.

   Generally, it is acceptable to leave alone hard sentence breaks
   consisting of exactly two spaces alone in existing documents.  These
   should be left as is in patches that do not directly touch the
   adjacent sentences.

I'm really hoping that someone will stick this _somewhere_ in the
guidelines for contributions.  I'm just not sure _where_ exactly it
should go.

Ciao,
Sheldon.
Comment 7 Sheldon Hearn freebsd_committer freebsd_triage 2000-03-01 08:28:22 UTC
State Changed
From-To: open->closed

Committed and merged to stable, thanks! 

Comment 8 chris 2000-03-01 19:20:06 UTC
On Sunday, February 27, 2000, Doug@gorean.org wrote:
> >Category:       bin

> Index: inetd.8
> ===================================================================

   For future reference, the proper catagory for these reports is
`doc'.

-- 
|Chris Costello <chris@calldei.com>
|**FLASH** Energizer Bunny arrested, charged with battery. 
`----------------------------------------------------------