Bug#862485: fwsnort mustn't set iptables rules when purged

2017-05-13 Thread Adrian Bunk
Package: fwsnort
Version: 1.6.5-3
Severity: critical
Tags: security

The #861999 fix adds the following on purging:
  grep -v FWSNORT /var/lib/fwsnort/fwsnort.save | iptables-restore

Imagine the following:
1. today I install fwsnort and try it
2. later today I uninstall it
3. 2 years later I purge all long-removed packages

This would in 2 years set the iptables rules to what they
were today before I shortly played with fwsnort.


A case could be made for "fwsnort --ipt-flush" in prerm.

Or considering that activating any fwsnort rules is not done
automatically and that the package should not interfere with
what the the admin has done.



Processed: Re: Bug#862485: fwsnort mustn't set iptables rules when purged

2017-05-13 Thread Debian Bug Tracking System
Processing control commands:

> tag -1 + moreinfo
Bug #862485 [fwsnort] fwsnort mustn't set iptables rules when purged
Added tag(s) moreinfo.
> severity -1 important
Bug #862485 [fwsnort] fwsnort mustn't set iptables rules when purged
Severity set to 'important' from 'critical'

-- 
862485: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=862485
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems



Bug#862485: fwsnort mustn't set iptables rules when purged

2017-05-13 Thread Axel Beckert
Control: tag -1 + moreinfo
Control: severity -1 important

Hi Adrian,

Adrian Bunk wrote:
> Severity: critical

I think that's overly exaggerated.

> Tags: security

I also disagree with this tag.

> The #861999 fix adds the following on purging:
>   grep -v FWSNORT /var/lib/fwsnort/fwsnort.save | iptables-restore

Yes. In postrm on purge.

> Imagine the following:
> 1. today I install fwsnort and try it
> 2. later today I uninstall it

You usually purge package if you play with packages which make changes
to your system.

> This would in 2 years set the iptables rules to what they
> were today before I shortly played with fwsnort.

I consider this (i.e. just removing but not purging when wanting to
get rid of a package and all its effects) to be the admin's fault, not
the package's fault.

> A case could be made for "fwsnort --ipt-flush" in prerm.

This would be against the expectation of users that configurations,
settings etc. are removed on purge and not on removal.

> Or considering that activating any fwsnort rules is not done
> automatically and that the package should not interfere with
> what the the admin has done.

I disagree. I expect a package to clean up its changes on purge which
result on common usage. To be more specifically, seeing 11'000
iptables rules left on my system after pruging fwsnort with no chance
to remove them without reinstalling the package or removing 11'000
rules by hand. Not cleaning up these rules is a bug. And cleaning up
is a task for "purge", not for "remove".

Regards, Axel
-- 
 ,''`.  |  Axel Beckert , http://people.debian.org/~abe/
: :' :  |  Debian Developer, ftp.ch.debian.org Admin
`. `'   |  4096R: 2517 B724 C5F6 CA99 5329  6E61 2FF9 CD59 6126 16B5
  `-|  1024D: F067 EA27 26B9 C3FC 1486  202E C09E 1D89 9593 0EDE



Bug#862485: fwsnort mustn't set iptables rules when purged

2017-05-13 Thread Adrian Bunk
Control: severity -1 serious

On Sat, May 13, 2017 at 07:27:27PM +0200, Axel Beckert wrote:
> Control: tag -1 + moreinfo
> Control: severity -1 important
> 
> Hi Adrian,

Hi Axel,

> Adrian Bunk wrote:
> > Severity: critical
> 
> I think that's overly exaggerated.
> 
> > Tags: security
> 
> I also disagree with this tag.

messing up the iptables setup at an unexpected time can have bad 
consequences.

> > The #861999 fix adds the following on purging:
> >   grep -v FWSNORT /var/lib/fwsnort/fwsnort.save | iptables-restore
> 
> Yes. In postrm on purge.
> 
> > Imagine the following:
> > 1. today I install fwsnort and try it
> > 2. later today I uninstall it
> 
> You usually purge package if you play with packages which make changes
> to your system.
> 
> > This would in 2 years set the iptables rules to what they
> > were today before I shortly played with fwsnort.
> 
> I consider this (i.e. just removing but not purging when wanting to
> get rid of a package and all its effects) to be the admin's fault, not
> the package's fault.
> 
> > A case could be made for "fwsnort --ipt-flush" in prerm.
> 
> This would be against the expectation of users that configurations,
> settings etc. are removed on purge and not on removal.

When you remove (not purge) a package containing a webserver, do you 
expect that the webserver is stopped or do you expect that the webserver
is still running after removing the package?

> > Or considering that activating any fwsnort rules is not done
> > automatically and that the package should not interfere with
> > what the the admin has done.
>
> I disagree. I expect a package to clean up its changes on purge which
> result on common usage. To be more specifically, seeing 11'000
> iptables rules left on my system after pruging fwsnort with no chance
> to remove them without reinstalling the package or removing 11'000
> rules by hand. Not cleaning up these rules is a bug. And cleaning up
> is a task for "purge", not for "remove".

"Remove an installed package. This removes everything except conffiles"
This the dpkg (and similar in apt) description of what remove does.

A package that is removed but not purged is in the Config-Files states.
This means the old configuration is still present if the package gets
installed again.

Purging is supposed to remove the (at that point already unused)
configuration files of the package.

Purging is not supposed to do any reconfiguration of the system.

>   Regards, Axel

cu
Adrian

-- 

   "Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
   "Only a promise," Lao Er said.
   Pearl S. Buck - Dragon Seed



Processed: Re: Bug#862485: fwsnort mustn't set iptables rules when purged

2017-05-13 Thread Debian Bug Tracking System
Processing control commands:

> severity -1 serious
Bug #862485 [fwsnort] fwsnort mustn't set iptables rules when purged
Severity set to 'serious' from 'important'

-- 
862485: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=862485
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems



Processed: Package is completely broken

2017-05-13 Thread Debian Bug Tracking System
Processing commands for cont...@bugs.debian.org:

> severity 858178 serious
Bug #858178 [uuidcdef] uuidcdef: buffer overflow
Severity set to 'serious' from 'normal'
> thanks
Stopping processing here.

Please contact me if you need assistance.
-- 
858178: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=858178
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems



Bug#862485: fwsnort mustn't set iptables rules when purged

2017-05-13 Thread Axel Beckert
Control: tag -1 + confirmed - moreinfo

Hi Adrian,

Adrian Bunk wrote:
> > > Tags: security
> > 
> > I also disagree with this tag.
> 
> messing up the iptables setup at an unexpected time can have bad 
> consequences.

bad != security

> > > A case could be made for "fwsnort --ipt-flush" in prerm.
> > 
> > This would be against the expectation of users that configurations,
> > settings etc. are removed on purge and not on removal.
> 
> When you remove (not purge) a package containing a webserver, do you 
> expect that the webserver is stopped or do you expect that the webserver
> is still running after removing the package?

Granted. Thanks for that comparison.

> > > Or considering that activating any fwsnort rules is not done
> > > automatically and that the package should not interfere with
> > > what the the admin has done.
> >
> > I disagree. I expect a package to clean up its changes on purge which
> > result on common usage. To be more specifically, seeing 11'000
> > iptables rules left on my system after pruging fwsnort with no chance
> > to remove them without reinstalling the package or removing 11'000
> > rules by hand. Not cleaning up these rules is a bug. And cleaning up
> > is a task for "purge", not for "remove".
> 
> "Remove an installed package. This removes everything except conffiles"
> This the dpkg (and similar in apt) description of what remove does.
> 
> A package that is removed but not purged is in the Config-Files states.
> This means the old configuration is still present if the package gets
> installed again.
> 
> Purging is supposed to remove the (at that point already unused)
> configuration files of the package.

Point taken. Will move that line (or an "fwsnort --ipt-flush") into a
(to be created) prerm and do another QA upload. (Unless you're already
onto it. Feel free to do that.)
 
Regards, Axel
-- 
 ,''`.  |  Axel Beckert , http://people.debian.org/~abe/
: :' :  |  Debian Developer, ftp.ch.debian.org Admin
`. `'   |  4096R: 2517 B724 C5F6 CA99 5329  6E61 2FF9 CD59 6126 16B5
  `-|  1024D: F067 EA27 26B9 C3FC 1486  202E C09E 1D89 9593 0EDE



Processed: Re: Bug#862485: fwsnort mustn't set iptables rules when purged

2017-05-13 Thread Debian Bug Tracking System
Processing control commands:

> tag -1 + confirmed - moreinfo
Bug #862485 [fwsnort] fwsnort mustn't set iptables rules when purged
Added tag(s) confirmed.
Bug #862485 [fwsnort] fwsnort mustn't set iptables rules when purged
Removed tag(s) moreinfo.

-- 
862485: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=862485
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems



wvstreams is marked for autoremoval from testing

2017-05-13 Thread Debian testing autoremoval watch
wvstreams 4.6.1-11 is marked for autoremoval from testing on 2017-06-13

It (build-)depends on packages with these RC bugs:
858178: uuidcdef: buffer overflow



gnome-ppp is marked for autoremoval from testing

2017-05-13 Thread Debian testing autoremoval watch
gnome-ppp 0.3.23-2 is marked for autoremoval from testing on 2017-06-13

It (build-)depends on packages with these RC bugs:
858178: uuidcdef: buffer overflow



xplc is marked for autoremoval from testing

2017-05-13 Thread Debian testing autoremoval watch
xplc 0.3.13-6 is marked for autoremoval from testing on 2017-06-13

It is affected by these RC bugs:
858178: uuidcdef: buffer overflow