netfilter-devel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Phil Sutter <phil@nwl.cc>
To: Pablo Neira Ayuso <pablo@netfilter.org>
Cc: Florian Westphal <fw@strlen.de>,
	Arturo Borrero Gonzalez <arturo@debian.org>,
	Jeremy Sowden <jeremy@azazel.net>,
	netfilter-devel@vger.kernel.org
Subject: Re: [RFC] nftables 1.0.6 -stable backports
Date: Tue, 10 Oct 2023 15:30:47 +0200	[thread overview]
Message-ID: <ZSVSB6uHXnLMm3L7@orbyte.nwl.cc> (raw)
In-Reply-To: <ZSUpeo2ozoPapyzg@calendula>

On Tue, Oct 10, 2023 at 12:37:46PM +0200, Pablo Neira Ayuso wrote:
> On Tue, Oct 10, 2023 at 10:39:15AM +0200, Phil Sutter wrote:
> > On Mon, Oct 09, 2023 at 01:41:33PM +0200, Pablo Neira Ayuso wrote:
> > > On Mon, Oct 09, 2023 at 01:15:43PM +0200, Florian Westphal wrote:
> > > > Arturo Borrero Gonzalez <arturo@debian.org> wrote:
> > > > > On 10/9/23 12:44, Pablo Neira Ayuso wrote:
> > > > > > - Another possibility is to make a nftables 1.0.6.1 or 1.0.6a -stable
> > > > > > release from netfilter.org. netfilter.org did not follow this procedure
> > > > > > very often (a few cases in the past in iptables IIRC).
> > > > > 
> > > > > Given the amount of patches, this would be the preferred method from the
> > > > > Debian point of view.
> > > > > 
> > > > > 1.0.6.1 as version should be fine.
> > > 
> > > Only one thing: I just wonder if this new 4 numbers scheme might
> > > create confusion, as there will be release with 3 numbers and -stable
> > > releases with 4 numbers.
> > 
> > An upcoming 1.0.9 might be a good chance to switch upstream numbering
> > scheme: Depending on whether it is deemed acceptable to reorder patches
> > in public git history, one could make 1.0.9 contain only the fixes since
> > 1.0.8 and release a 1.1.0 containing what remains. And from then on
> > collect just fixes to 1.1.0 into 1.1.N and new features into 1.2.0.
> >
> > Assuming that downstream does its own "stable releases" already,
> > skipping a 1.0.6.1 or 0.9.8.1 should be OK. Was a 0.9.10, being
> > 0.9-stable, acceptable or are there too many new features between 0.9.8
> > and 0.9.9?
> 
> I made a bit of digging in the history, and we already pulled the 4
> digits handle in the past for iptables.
> 
> https://www.netfilter.org/projects/iptables/files/changes-iptables-1.4.19.1.txt

Appending another "dot digit" is not uncommon in other projects, so I
guess most parsers should get it right.

> As for 0.9.10, it would skip 0.9.9:

I did not mean for 0.9.10 to be 0.9.8 + fixes, but 0.9.9 + only fixes.
So not skip, but include.

> $ git log --oneline v0.9.8..v0.9.9 | wc -l
> 150

Skimming the list, I think there's not too much in there which is not a
fix. While there are only 20 commits having a Fixes: tag, there's the
parser keyword scoping and some cache rework also. In other words,
requiring downstream to update to 0.9.9 from 0.9.8 in order to benefit
from upstream's blessed 0.9-stable release might be acceptable.

> We can start with a few -stable branches, namely 0.9.8.y and 1.0.6.y
> as it has been suggested, I am going to push patches to the branches,
> I will keep you posted.

Fine with me, too.

Cheers, Phil

  reply	other threads:[~2023-10-10 13:30 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-10-09 10:44 [RFC] nftables 1.0.6 -stable backports Pablo Neira Ayuso
2023-10-09 10:50 ` Florian Westphal
2023-10-09 11:05 ` Arturo Borrero Gonzalez
2023-10-09 11:15   ` Florian Westphal
2023-10-09 11:41     ` Pablo Neira Ayuso
2023-10-10  8:39       ` Phil Sutter
2023-10-10 10:37         ` Pablo Neira Ayuso
2023-10-10 13:30           ` Phil Sutter [this message]
2023-10-10 11:53         ` Jan Engelhardt
2023-10-10 15:24         ` Arturo Borrero Gonzalez
2023-10-09 11:44   ` Jeremy Sowden
2023-10-09 11:36 ` [RFC] nftables 0.9.8 " Pablo Neira Ayuso
2023-10-09 11:50   ` Jeremy Sowden
2023-10-10  8:54   ` Pablo Neira Ayuso
2023-10-10 20:08     ` Jeremy Sowden
2023-10-10 22:21       ` Pablo Neira Ayuso
2023-10-11  9:46         ` Jeremy Sowden
2023-10-11 10:01           ` Pablo Neira Ayuso
2024-02-17 20:11   ` Jeremy Sowden
2024-02-18 13:56     ` Jeremy Sowden
2023-10-11  8:01 ` [RFC] nftables 1.0.6 " Pablo Neira Ayuso
2023-10-11 15:25   ` Phil Sutter
2023-10-11 15:49     ` Pablo Neira Ayuso
2023-10-19 14:27       ` Pablo Neira Ayuso
2023-10-19 15:23         ` Phil Sutter
2023-11-02 11:34           ` Pablo Neira Ayuso
2023-11-02 12:27             ` Phil Sutter
2023-11-02 21:23               ` Pablo Neira Ayuso
2023-11-03  9:59                 ` Phil Sutter
2023-11-03 10:44                   ` Phil Sutter
2023-11-03 10:49                     ` Pablo Neira Ayuso
2023-11-03 10:56                       ` Phil Sutter
2023-11-03 11:29                         ` Pablo Neira Ayuso
2023-11-03 12:11                           ` Pablo Neira Ayuso
2023-11-03 15:02                           ` Phil Sutter
2023-11-03 15:15                             ` Pablo Neira Ayuso

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=ZSVSB6uHXnLMm3L7@orbyte.nwl.cc \
    --to=phil@nwl.cc \
    --cc=arturo@debian.org \
    --cc=fw@strlen.de \
    --cc=jeremy@azazel.net \
    --cc=netfilter-devel@vger.kernel.org \
    --cc=pablo@netfilter.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).