Our releases look to flip-flop between keeping or discarding the date
and title of the referenced qemu-trad commit. I think with the hash
replaced by a tag, the commit's date and title would better also be
purged.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
---
I've been noticing this inconsistency repeatedly because I simply re-use
version update patches for the stable trees - the context mismatch
results in a such updated patch to then not apply, due to said flip-
flopping. For 4.15 I'm inclined to drop the two offending lines perhaps
while updating the version to 4.15.1-pre.

--- a/docs/process/release-technician-checklist.txt
+++ b/docs/process/release-technician-checklist.txt
@@ -53,6 +53,7 @@
 #   QEMU_UPSTREAM_REVISION,
 #   QEMU_TRADITIONAL_REVISION
 #   MINIOS_UPSTREAM_REVISION
+#     (drop any references to the specific commits, e.g. date or title)
 * change SUPPORT.md heading version number; -unstable or -rc tag
 *     (empty in stable branches after .0 release).
 * SUPPORT.md: insert correct version number in release-notes link

On 07.04.2021 11:56, Jan Beulich wrote:
> Our releases look to flip-flop between keeping or discarding the date
> and title of the referenced qemu-trad commit. I think with the hash
> replaced by a tag, the commit's date and title would better also be
> purged.
> 
> Signed-off-by: Jan Beulich <jbeulich@suse.com>
> ---
> I've been noticing this inconsistency repeatedly because I simply re-use
> version update patches for the stable trees - the context mismatch
> results in a such updated patch to then not apply, due to said flip-
> flopping. For 4.15 I'm inclined to drop the two offending lines perhaps
> while updating the version to 4.15.1-pre.

Ian (and others), any thoughts here either way?

Thanks, Jan

> --- a/docs/process/release-technician-checklist.txt
> +++ b/docs/process/release-technician-checklist.txt
> @@ -53,6 +53,7 @@
>  #   QEMU_UPSTREAM_REVISION,
>  #   QEMU_TRADITIONAL_REVISION
>  #   MINIOS_UPSTREAM_REVISION
> +#     (drop any references to the specific commits, e.g. date or title)
>  * change SUPPORT.md heading version number; -unstable or -rc tag
>  *     (empty in stable branches after .0 release).
>  * SUPPORT.md: insert correct version number in release-notes link
> 

On 27.04.2021 18:01, Jan Beulich wrote:
> On 07.04.2021 11:56, Jan Beulich wrote:
>> Our releases look to flip-flop between keeping or discarding the date
>> and title of the referenced qemu-trad commit. I think with the hash
>> replaced by a tag, the commit's date and title would better also be
>> purged.
>>
>> Signed-off-by: Jan Beulich <jbeulich@suse.com>
>> ---
>> I've been noticing this inconsistency repeatedly because I simply re-use
>> version update patches for the stable trees - the context mismatch
>> results in a such updated patch to then not apply, due to said flip-
>> flopping. For 4.15 I'm inclined to drop the two offending lines perhaps
>> while updating the version to 4.15.1-pre.
> 
> Ian (and others), any thoughts here either way?

If I don't hear back by Wednesday, I guess I'll take the lazy consensus
route and put this in.

Jan

Jan Beulich writes ("Re: Ping: [PATCH] docs: release-technician-checklist: update to leaf tree version pinning"):
> On 27.04.2021 18:01, Jan Beulich wrote:
> > On 07.04.2021 11:56, Jan Beulich wrote:
> >> Our releases look to flip-flop between keeping or discarding the date
> >> and title of the referenced qemu-trad commit. I think with the hash
> >> replaced by a tag, the commit's date and title would better also be
> >> purged.
> >>
> >> Signed-off-by: Jan Beulich <jbeulich@suse.com>
> >> ---
> >> I've been noticing this inconsistency repeatedly because I simply re-use
> >> version update patches for the stable trees - the context mismatch
> >> results in a such updated patch to then not apply, due to said flip-
> >> flopping. For 4.15 I'm inclined to drop the two offending lines perhaps
> >> while updating the version to 4.15.1-pre.
> > 
> > Ian (and others), any thoughts here either way?
> 
> If I don't hear back by Wednesday, I guess I'll take the lazy consensus
> route and put this in.

That would of course be correct of you :-).

But I should explicitly:

Acked-by: Ian Jackson <iwj@xenproject.org>

Thanks,
Ian.