Re: lpd: the thin edge of the wedge?

On 24.02.2026 at 03:43, Greg 'groggy' Lehey wrote:
> I'm really quite concerned about the plans to remove lpd.  I
> understand that there are security issues with lpd, even if I haven't
> heard any reports of exploits in over a third of a century, but the
> approach seems wrong to me.  If we follow this direction, we can pare
> down FreeBSD to a bare minimum (the kernel and what else?).
> 
> So what should be done?  I'm explicitly copying core@ on this, though
> I assume that you have all been following things.  But this is a basic
> issue for the project, so core@ (and not srcmgr@) should have a
> position on this.
> 
> I understand that des no longer feels interested in maintaining lpd or
> fixing its apparently numerous bugs.  But there are alternatives:
> 
> 1.  Find somebody who*is* interested.  I haven't seen anything on the
>      mailing lists asking for this.  Why not?
> 
> 2.  Take the corresponding code from another BSD, like we have done in
>      the past.  des tells us, without details, that the OpenBSD code
>      has the same bugs.  I've asked numerous times, but nobody has told
>      me whether they have spoken with the OpenBSD project about it.
>      And what about NetBSD or DragonflyBSD?
> 
> 3.  Make it a shared project amongst BSDs, like make(1).
> 
> What seems completely wrong to me is to outsource it to a port,
> especially one that is unmaintained, has a GPL license and has
> conflicts with existing installations.


Dear Greg,

in the past, I truly appreciated and made extensive use of the Vinum 
volume manager. Today I use ZFS instead, but I remain grateful to 
everyone whose work helped make Vinum such a solid and elegant solution 
in its time.

I read your books with great attention and enthusiasm. Times change, 
however, and unfortunately some things inevitably fade into history, 
whether we want them to or not. If we wanted to preserve all remaining 
traces of the old BSD in the FreeBSD operating system in order to 
increase its "museum value" to the level of a living open-air museum, 
then keeping the traditional lpd implementation would indeed make 
perfect sense.

The question is whether such a museum can still be considered complete, 
given that uucp, telnetd, ftpd, and many other daemons have already been 
removed. A surrogate museum like this no longer has real historical 
value, and no software archaeologist would take it seriously. Therefore, 
it is already far too late to continue treating FreeBSD as a living 
software museum.

Joking aside, I did not expect anyone to seriously defend keeping lpd in 
the base system, although in the past I myself defended various base 
components that were being removed as obsolete. I did so because I 
believed that those particular solutions still worked very well in the 
21st century and were often needed immediately after installing the system.

In the case of lpd(1) and the traditional printing toolchain, no one has 
such illusions. For a quarter of a century now, all new deployments have 
been based on the CUPS printing system, which is permissively licensed 
and readily available in the ports.

With best regards,
-- 
Marek Zarychta
FeeBSD user