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

Bug#565213: apt-listbugs: can't install because it's not installed



On Sat, 16 Jan 2010 23:08:57 +0100 David Kalnischkies wrote:

[...]
> It would be good if someone could have a look at the patch,
> especially at the manpage changes, as it is as always:
> I understand it as i know what i wrote and how it works,
> but your mileage may vary...

I took a quick look at the man page changes, and they look basically OK
to me.

> 
> 2010/1/16 Francesco Poli <frx@firenze.linux.it>:
> > As explained in my message
> > http://bugs.debian.org/565213#37
> > I disagree.
> Yes, make sense - and in the end: It is your package,
> so you can do what you want. ;)

It is (currently) my package and also Ryan's.
But anyway, we are open to constructive criticism!   ;-)

> 
> I didn't know how apt-listbugs behaves so i (wrongly) suspect
> that it has always some kind of output, so a user could
> very easy detect if something went wrong.

It does have some output, but, when there are no RC bugs for the
packages being installed or upgraded, this output is really very few
lines inside the package manager's output.
If the user does not look closely, the absence of those few lines may
really go unnoticed, IMHO.

> Through i guess you have bigger problems if your applications
> "magically" loose their execution flag than a silent listbugs. ;)

Of course one has bigger problems, if his/her applications lose their
execution flag!
But one could find out to have bigger problems, exactly by noticing
that the package manager stops working!


[...]
> A quick & untested idea: Instead of renaming the file how about
> another config file which reverts the config file:
> 
> 10apt-listbugs.conf
> DPkg::Pre-Install-Pkgs::apt-listbugs "/usr/sbin/apt-listbugs apt || exit 10";
> AptListbugs::Option1 "value1";
> AptListbugs::Option2 "value2";
> 
> 11disable-apt-listbugs.conf
> DPkg::Pre-Install-Pkgs::apt-listbugs "";
> #clear AptListbugs;
> 
> ? (at least, this should be valid according to the "spec" apt.conf.5 )

Mmmh, I must think about it.

> 
> The other idea is simply to define the *.disable meaning
> (if dpkg has nothing against it) and ask the others to support it...

I thought that package managers claimed to be compatible with APT
configuration parameters were *already* behaving consistently with
apt-get, when it comes to decide which configuration fragments have to
be ignored!
Now it seems you are saying that this is not the case: this is
worrying, IMHO...
aptitude behaves as apt-get, since it uses libapt, doesn't it?
But what about the other "compatible" package managers?


-- 
 http://www.inventati.org/frx/progs/scripts/pdebuild-hooks.html
 Need some pdebuild hook scripts?
..................................................... Francesco Poli .
 GnuPG key fpr == C979 F34B 27CE 5CD8 DC12  31B5 78F4 279B DD6D FCF4

Attachment: pgpMsL5wyXfhe.pgp
Description: PGP signature


Reply to: