[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#700943: RFS: stdsyslog/0.03-1 [ITP] -- log a program's output to the system log



On Wed, Feb 20, 2013 at 10:00:43PM +0100, gregor herrmann wrote:
> Control: tag -1 + confirmed moreinfo
> Control: tag -1 owner !
> 
> On Tue, 19 Feb 2013 17:26:34 +0200, Peter Pentchev wrote:
> 
> > I am looking for a sponsor for my package "stdsyslog" - a tool that runs
> > a program and sends its output (on the standard output, standard error
> > and possibly other streams) to the system log.  This will be an initial
> > upload to close my ITP bug #699347.
> 
> Nice idea, and the package looks good to me.

Thanks! :)  And thanks for the review!

> Just two minor questions:
> - Why don't you run the test suite during the build? I tried with
>   TEST_STDSYSLOG and it worked.

Well, I imagine - though I could be wrong here - that some people might
not be too happy seeing syslog messages coming from a package build; not
to mention these syslog messages making their way into IDS reports and
suchlike :)  And yes, I do realize that the Debian package build
infrastructure probably won't care (but still there is a chance that the
DSA team does examine the system logs of the build cluster machines in
detail).

Of course, another part of the reason is that these tests do not really
test the *functionality* of stdsyslog - all they do is make sure that
the stdsyslog binary *accepts* the combinations of command-line options.
For a real functionality test it would be necessary to examine the
actual system logs after running stdsyslog to make sure that the
messages really got there - but I believe that this is way beyond the
scope of stdsyslog's test suite :)

Still, if you think that it would be a good idea to run the test suite
during the Debian package build, I'll make it so.

> - Any specific reason for "debhelper (>> 9)" instead of ">="? Not
>   that it changes anything, I'm just curious about the
>   not-so-often-used option.

I guess I kind of got used to this sometime in the early days of my
Debian packaging... I honestly cannot remember which packages I looked
at, but I do remember that I picked this up from some real-world source
packages.  As you say, it doesn't change anything; I could switch it to
">=", especially as this would make it easier to depend on more specific
versions if needed (although, actually, since I'm in the habit of
depending on versions with a tilde at the end to also handle backported
versions, it would still make no difference - >> 9.0020~ would match
exactly the same set of *really uploaded* packages as >= 9.0020~).

> (BTW: A pristine-tar branch in your git repo would be nice; at least
> for git-buildpackage users, maybe not so mach otherwise :))

Oof, right :)  I'll do that tonight or early tomorrow.  It's just that
this is the master repository for the stdsyslog "upstream" sources, too,
so I didn't even think about pristine-tar in this case :)

G'luck,
Peter

-- 
Peter Pentchev	roam@ringlet.net roam@FreeBSD.org p.penchev@storpool.com
PGP key:	http://people.FreeBSD.org/~roam/roam.key.asc
Key fingerprint 2EE7 A7A5 17FC 124C F115  C354 651E EFB0 2527 DF13
If wishes were fishes, the antecedent of this conditional would be true.

Attachment: signature.asc
Description: Digital signature


Reply to: