netdev - Re: Does anyone use Appletalk?

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Wed, 01 Nov 2023 23:33:42 +0100
From: John Paul Adrian Glaubitz <glaubitz@...sik.fu-berlin.de>
To: Dan Williams <dcbw@...hat.com>, Geert Uytterhoeven <geert@...ux-m68k.org>
Cc: linux-m68k <linux-m68k@...ts.linux-m68k.org>, Arnd Bergmann
 <arnd@...db.de>,  Jakub Kicinski <kuba@...nel.org>, netdev
 <netdev@...r.kernel.org>
Subject: Re: Does anyone use Appletalk?

On Wed, 2023-11-01 at 15:27 -0500, Dan Williams wrote:
> But... Time Machine debuted with 10.5 and AppleTalk got removed in
> 10.6; did the actual TimeCapsules ever support AppleTalk, or were they
> always TCP/IP-based?

netatalk has two actively maintained versions, one for AppleTalk (2.2.x
series) and one for TCP/IP (3.x series). Both are still being developed
and supported [1].

> (also TimeMachine-capable Airport Extremes [A1354] are like $15 on
> eBay; that's cheaper than a Raspberry Pi)

I know that commercial entities don't have interest in legacy architectures
and protocols. But Linux isn't a commercial-only project so legacy applications
have a valid use case. Most people in the Linux community don't have a use case
for IBM mainframes, yet they aren't in sending patches to get s390 support removed.

I understand that sometimes old code needs to be dropped when it becomes
a burden which is why I also agreed to drop ia64 support since I have
heard complaints from multiple upstream projects and I also know that a
lot of stuff there is broken with no one willing to fix it.

But I don't understand the removal in this case. What particular burden
does a legacy networking protocol pose if it can be easily disabled at
compile time to reduce the attack surface?

> This patch only removes the Linux-side ipddp driver (eg MacIP) so if
> Time Capsules never supported AppleTalk, this patch is unrelated to
> TimeMachine.
> 
> What this patch *may* break is Linux as a MacIP gateway, allowing
> AppleTalk-only machines to talk TCP/IP to systems. But that's like
> what, the 128/512/Plus and PowerBook Duo/1xx? Everything else had a
> PDS/NuBus slot or onboard Ethernet and could do native
> MacTCP/OpenTransport...

Which is a valid use case for people from the retro-computing community
as can be seen from the netatalk description above. I don't think that
Arnd reached out to the netatalk project and asked whether the code
is still needed, did he?

Adrian

> [1] https://en.wikipedia.org/wiki/Netatalk

-- 
 .''`.  John Paul Adrian Glaubitz
: :' :  Debian Developer
`. `'   Physicist
  `-    GPG: 62FF 8A75 84E0 2956 9546  0006 7426 3B37 F5B5 F913