From nobody Wed Jun 22 17:22:04 2022 X-Original-To: freebsd-arch@mlmmj.nyi.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mlmmj.nyi.freebsd.org (Postfix) with ESMTP id 67BE286775D for ; Wed, 22 Jun 2022 17:22:09 +0000 (UTC) (envelope-from grembo@freebsd.org) Received: from mail.evolve.de (mail.evolve.de [213.239.217.29]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA512 client-signature RSA-PSS (4096 bits) client-digest SHA256) (Client CN "mail.evolve.de", Issuer "R3" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id 4LSqv40Zcsz3DwF; Wed, 22 Jun 2022 17:22:08 +0000 (UTC) (envelope-from grembo@freebsd.org) Received: by mail.evolve.de (OpenSMTPD) with ESMTP id 5241f6f1; Wed, 22 Jun 2022 17:22:05 +0000 (UTC) Received: by mail.evolve.de (OpenSMTPD) with ESMTPSA id 9db16b06 (TLSv1.3:AEAD-AES256-GCM-SHA384:256:NO); Wed, 22 Jun 2022 17:22:05 +0000 (UTC) Content-Type: multipart/alternative; boundary=Apple-Mail-F8A4AB9A-15A8-449D-BA70-5050D931D38D Content-Transfer-Encoding: 7bit List-Id: Discussion related to FreeBSD architecture List-Archive: https://lists.freebsd.org/archives/freebsd-arch List-Help: List-Post: List-Subscribe: List-Unsubscribe: Sender: owner-freebsd-arch@freebsd.org Mime-Version: 1.0 (1.0) Subject: Re: Updating reboot's default From: Michael Gmelin In-Reply-To: Date: Wed, 22 Jun 2022 19:22:04 +0200 Cc: Greg 'groggy' Lehey , Warner Losh , freebsd-arch@freebsd.org Message-Id: <59503878-4DF2-49B7-AF40-A3AA506889DA@freebsd.org> References: To: Warner Losh X-Mailer: iPhone Mail (19E258) X-Rspamd-Queue-Id: 4LSqv40Zcsz3DwF X-Spamd-Bar: ++ Authentication-Results: mx1.freebsd.org; dkim=none; dmarc=none; spf=softfail (mx1.freebsd.org: 213.239.217.29 is neither permitted nor denied by domain of grembo@freebsd.org) smtp.mailfrom=grembo@freebsd.org X-Spamd-Result: default: False [2.70 / 15.00]; ARC_NA(0.00)[]; RCVD_VIA_SMTP_AUTH(0.00)[]; FREEFALL_USER(0.00)[grembo]; FROM_HAS_DN(0.00)[]; RCPT_COUNT_THREE(0.00)[4]; MV_CASE(0.50)[]; NEURAL_SPAM_SHORT(0.61)[0.605]; MIME_GOOD(-0.10)[multipart/alternative,text/plain]; DMARC_NA(0.00)[freebsd.org]; R_SPF_SOFTFAIL(0.00)[~all:c]; NEURAL_SPAM_MEDIUM(1.00)[0.998]; RCVD_COUNT_THREE(0.00)[3]; TO_DN_SOME(0.00)[]; TO_MATCH_ENVRCPT_SOME(0.00)[]; NEURAL_SPAM_LONG(0.70)[0.697]; MLMMJ_DEST(0.00)[freebsd-arch]; FROM_EQ_ENVFROM(0.00)[]; R_DKIM_NA(0.00)[]; MIME_TRACE(0.00)[0:+,1:+,2:~]; ASN(0.00)[asn:24940, ipnet:213.239.192.0/18, country:DE]; RCVD_TLS_ALL(0.00)[]; MID_RHS_MATCH_FROM(0.00)[] X-ThisMailContainsUnwantedMimeParts: N --Apple-Mail-F8A4AB9A-15A8-449D-BA70-5050D931D38D Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable > On 22. Jun 2022, at 19:12, Warner Losh wrote: >=20 > =EF=BB=BF >=20 >=20 >> On Wed, Jun 22, 2022 at 1:03 AM Michael Gmelin wrote= : >>=20 >>=20 >>>> On 22. Jun 2022, at 04:03, Warner Losh wrote: >>>>=20 >>> =EF=BB=BF >>>=20 >>>=20 >>>> On Tue, Jun 21, 2022, 6:35 PM Greg 'groggy' Lehey wr= ote: >>>> On Tuesday, 21 June 2022 at 8:01:58 -0600, Warner Losh wrote: >>>> > 15 or 20 years ago, we talked about changing the default for reboot f= rom >>>> > 'right now' to being safe shutdown. There were arguments made against= it >>>> > due to tiny appliances and such. >>>> > >>>> > Time has past, and this oddity has persisted. It's time to revisit th= at >>>> > decision. >>>> > >>>> > I'd propose that we keep 'fastboot' and 'fasthalt' having the immedia= te >>>> > behavior. However, the 'reboot' command will switch from '-q' behavio= r to >>>> > '-r' behavior. >>>>=20 >>>> Somehow I hear this echo "If it ain't broke, don't fix it". My >>>> understanding has always been that shutdown(8) is the program that >>>> shuts down and maybe reboots the system, while reboot(8) is a quick >>>> and dirty way to reboot the system, along with halt(8) if you don't >>>> want to reboot. >>>>=20 >>>> So why change this? At the very least you'll confuse people who want >>>> to use the old method. My guess is that you have some reason that's >>>> not immediately apparent, but what? >>>=20 >>>=20 >>> Other systems have the behavior I'm advocating. We are the odd duck. Thi= s means we tend to violate POLA here. And there is no good reason to do this= when fastboot is available. Nobody that advocated to keep this difference a= s useful the last time it came up still wants to advocate. Most people find t= he behavior annoying and only a vanishingly small minority of people like it= . In fact, so far nobody has even asked to please not, let alone come up wit= h a good reason to retain this behavior. So, I'm polling arch@ to see if any= one like that shows up. >>>=20 >>=20 >> Well, to be honest, I=E2=80=99m used to the current behavior and would pr= efer to keep it (POLA for existing users). I didn=E2=80=99t answer to advoca= te against the change as >>=20 >> 1. I have no metric to counter your argument that this is a real problem f= or people used to other OSes (neither how many people pick up FreeBSD in gen= eral nor how many are unpleasantly surprised by how `reboot` works) >> 2. I will certainly be able to adapt and get used to the new behavior >> 3. Given the amount of change in the world right now, it=E2=80=99s a =E2=80= =9Cpick your battles=E2=80=9D situation. There is and will be so much to suc= k up, arguing about this with someone who clearly put some thought into it s= eems like a waste of everybody=E2=80=99s time. >=20 > I posted so I could understand other views, so I'd like to ask some questi= ons if I may. >=20 > Is your reliance on the current default due to shell and similar scripts y= ou have? Or is it due to your interactive operations? Interactive operation, my playbooks/automation use clean reboot (shutdown -r= ). > What do you like about the current behavior: How quickly the reboot happen= s? Or you have a lot of running processes you don't want killed or to have a= chance to clean up? Both > What build process do you use to create your FreeBSD images? Images from t= he RE, buildworld, nanobsd, poudriere, etc... >=20 Official release & buildworld & poudriere & packer > The only thought I've put into this is from my perspective, and while it i= s often a good reflection of the larger community, there are times there's a= mismatch, so I'd like to at least understand why you hold these views. Ther= e may be a simple way to accommodate both sides. >=20 For me this is almost solely about muscle memory. If changing it serves the m= any, I can adapt and start using `fastboot` again (which I used for many yea= rs, until I realized that `reboot` did exactly the same) If we were redoing things from scratch, what you propose would make more sen= se than the current behavior for sure. Cheers Michael > Warner > =20 >> Cheers >> Michael >>=20 >>> Warner=20 >>>=20 >>>=20 >>>> And no, I don't really have an axe to grind in this matter. >>>>=20 >>>> Greg >>>> -- >>>> Sent from my desktop computer. >>>> See complete headers for address and phone numbers. >>>> This message is digitally signed. If your Microsoft mail program >>>> reports problems, please read http://lemis.com/broken-MUA.php --Apple-Mail-F8A4AB9A-15A8-449D-BA70-5050D931D38D Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable

On 22. Jun 2022, at 19= :12, Warner Losh <imp@bsdimp.com> wrote:

=EF=BB=BF


On Wed, Jun 22, 2022 at 1:03 AM Michael Gmelin <grembo@freebsd.org> wrote:


On 22. Jun 2022, at 04:03, Warner Losh <imp@bsdimp.com> wrote:

=
=EF=BB=BF
<= br>
On T= ue, Jun 21, 2022, 6:35 PM Greg 'groggy' Lehey <grog@freebsd.org> wrote:
On Tuesday, 21 June 2022 at  8= :01:58 -0600, Warner Losh wrote:
> 15 or 20 years ago, we talked about changing the default for reboot fro= m
> 'right now' to being safe shutdown. There were arguments made against i= t
> due to tiny appliances and such.
>
> Time has past, and this oddity has persisted. It's time to revisit that=
> decision.
>
> I'd propose that we keep 'fastboot' and 'fasthalt' having the immediate=
> behavior. However, the 'reboot' command will switch from '-q' behavior t= o
> '-r' behavior.

Somehow I hear this echo "If it ain't broke, don't fix it".  My
understanding has always been that shutdown(8) is the program that
shuts down and maybe reboots the system, while reboot(8) is a quick
and dirty way to reboot the system, along with halt(8) if you don't
want to reboot.

So why change this?  At the very least you'll confuse people who want to use the old method.  My guess is that you have some reason that's not immediately apparent, but what?

Other systems have the behavior I'm advoca= ting. We are the odd duck. This means we tend to violate POLA here. And ther= e is no good reason to do this when fastboot is available. Nobody that advoc= ated to keep this difference as useful the last time it came up still wants t= o advocate. Most people find the behavior annoying and only a vanishingly sm= all minority of people like it. In fact, so far nobody has even asked to ple= ase not, let alone come up with a good reason to retain this behavior. So, I= 'm polling arch@ to see if anyone like that shows up.


Well, to be honest, I= =E2=80=99m used to the current behavior and would prefer to keep it (POLA fo= r existing users). I didn=E2=80=99t answer to advocate against the change as=

1. I have no metric to counter your argument that t= his is a real problem for people used to other OSes (neither how many people= pick up FreeBSD in general nor how many are unpleasantly surprised by how `= reboot` works)
2. I will certainly be able to adapt and get used t= o the new behavior
3. Given the amount of change in the world righ= t now, it=E2=80=99s a =E2=80=9Cpick your battles=E2=80=9D situation. There i= s and will be so much to suck up, arguing about this with someone who clearl= y put some thought into it seems like a waste of everybody=E2=80=99s time.

I posted so I could understand ot= her views, so I'd like to ask some questions if I may.

<= div>Is your reliance on the current default due to shell and similar scripts= you have? Or is it due to your interactive operations?

Interactive operation, my playbooks/auto= mation use clean reboot (shutdown -r).


What do you like about the current behavior: How quickly the reboot happe= ns? Or you have a lot of running processes you don't want killed or to have a= chance to clean up?

Both

What build process do you use to c= reate your FreeBSD images? Images from the RE, buildworld, nanobsd, poudrier= e, etc...


=
Official release & buildworld & poudriere & packer


=
The only thought I've put into this is from m= y perspective, and while it is often a good reflection of the larger communi= ty, there are times there's a mismatch, so I'd like to at least understand w= hy you hold these views. There may be a simple way to accommodate both s= ides.


For me this is almost solely about muscle memory. If changing it serves the= many, I can adapt and start using `fastboot` again (which I used for many y= ears, until I realized that `reboot` did exactly the same)

If we were redoing things from scratch, what you propose would make m= ore sense than the current behavior for sure.

Cheer= s
Michael


Warner
=
 
Cheers
Michael

Warner 


And no, I don't really have an axe to grind in this matter.

Greg
--
Sent from my desktop computer.
See complete headers for address and phone numbers.
This message is digitally signed.  If your Microsoft mail program
reports problems, please read http://lemis.com/broken-MUA.ph= p
= --Apple-Mail-F8A4AB9A-15A8-449D-BA70-5050D931D38D--