[PC-BSD Testing] PC-BSD 8.2-RC1 Installation issues

Kris Moore kris at pcbsd.org
Wed Jan 12 13:43:48 PST 2011

On 01/12/2011 16:38, LinuxBSDos.com wrote:
>> On 01/12/2011 16:15, LinuxBSDos.com wrote:
>>>> Should we enforce password lengths/"complexity"? I'm a fan of the UNIX
>>>> model: Be stupid and expect stupid.
>>> I think it makes plenty of sense to enforce password length. Complexity,
>>> I'm not too worried about, but length, yes (don't let you mind stray
>>> from
>>> the topic ;) )!
>>> We should not make it easy for users, especially inexperienced ones, to
>>> install loosely secured systems.
>>> I have a system I installed yesterday where the root passwd is "r" and
>>> the
>>> user password is "s." That's just as bad as auto-login.
>>> I'm just happy that Kris has cleaned that bit up. That said, I think a
>>> 4-character minimum for user password is to0 low. Six to 8 should be
>>> better.
>> Well, I made the user/root passwords a bit less strict, since a common
>> one for virtual-machines is "pcbsd" :)
>>>> IIRC, SSH is disabled in PC-BSD so there shouldn't be too much worry if
>>> a > user prefers 'password'.
>>> I don't think that running a network service is the sole reason for
>>> setting strong passwords. What about protecting physical access?
>> In the case of disk-encryption, yes you want a better password. But when
>> it comes to physical access, just reboot and go to single user,
>> insta-root. (Why i use encryption on my disks now)
> But there is a way to protect single-user mode, right?
> --
> Fini Decima
> http://LinuxBSDos.com

True, as Lars pointed out that is a good way to stop the quick-single
user mode trick.

I just assume if somebody has physical access you are already screwed,
unless you've encrypted. Otherwise you can always boot from a live CD,
mount the disk and mess with stuff or do the single-user trick assuming
they didn't do the console change :)

Kris Moore
PC-BSD Software

More information about the Testing mailing list