You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@aurora.apache.org by Kevin Sweeney <ke...@apache.org> on 2013/12/14 03:20:59 UTC
Tagging process
Hey all,
I think figuring out the tagging process sooner than later is in everyone's
best interest so that we can get out of cherry-pick limbo. I'm out next
week, but if anyone wants to take a stab at it (using
https://reviews.apache.org/r/16265/ as a starting point or just throw it
away, please feel free). My thinking:
.auroraversion on master only ever contains -SNAPSHOT versions. This makes
it harder for someone to reset and accidentally creating bogus non-SNAPSHOT
artifacts.
At each tag there are 2 commits. One on an anonymous branch that does
s/-SNAPSHOT// and one on master that increments the MINOR portion of the
version. So if master is 0.2.0-SNAPSHOT this new branch will have 0.2.0.
For that commit we create an annotated (preferably PGP-signed) tag, 0.2.0.
The other commit on master changes .auroraversion to 0.3.0-SNAPSHOT.
Once everything is verified we'll push 2 things - the new master and the
tag (so origin will not have a name for the branch the tag was created on).
What does everyone think of this process?
--
Kevin Sweeney
@kts
Re: Tagging process
Posted by Maxim Khutornenko <ma...@twitter.com>.
I presume hotfixing would be a completely manual process where we would branch out from a tagged release?
Thanks,
Maxim
On Dec 13, 2013, at 6:20 PM, Kevin Sweeney <ke...@apache.org> wrote:
> Hey all,
>
> I think figuring out the tagging process sooner than later is in everyone's best interest so that we can get out of cherry-pick limbo. I'm out next week, but if anyone wants to take a stab at it (using https://reviews.apache.org/r/16265/ as a starting point or just throw it away, please feel free). My thinking:
>
> .auroraversion on master only ever contains -SNAPSHOT versions. This makes it harder for someone to reset and accidentally creating bogus non-SNAPSHOT artifacts.
>
> At each tag there are 2 commits. One on an anonymous branch that does s/-SNAPSHOT// and one on master that increments the MINOR portion of the version. So if master is 0.2.0-SNAPSHOT this new branch will have 0.2.0. For that commit we create an annotated (preferably PGP-signed) tag, 0.2.0. The other commit on master changes .auroraversion to 0.3.0-SNAPSHOT.
>
> Once everything is verified we'll push 2 things - the new master and the tag (so origin will not have a name for the branch the tag was created on).
>
> What does everyone think of this process?
> --
> Kevin Sweeney
> @kts
Re: Tagging process
Posted by Zameer Manji <zm...@gmail.com>.
This process sounds good to me. So long as the process is automated I think
we will be fine. Will the person who creates the release ("release czar"?)
get to choose what lands on master before the release is cut?
Also does this mean we need to keep the CHANGELOG updated as we commit new
features so the tagged release has a proper changelog?
On Fri, Dec 13, 2013 at 6:20 PM, Kevin Sweeney <ke...@apache.org> wrote:
> Hey all,
>
> I think figuring out the tagging process sooner than later is in everyone's
> best interest so that we can get out of cherry-pick limbo. I'm out next
> week, but if anyone wants to take a stab at it (using
> https://reviews.apache.org/r/16265/ as a starting point or just throw it
> away, please feel free). My thinking:
>
> .auroraversion on master only ever contains -SNAPSHOT versions. This makes
> it harder for someone to reset and accidentally creating bogus non-SNAPSHOT
> artifacts.
>
> At each tag there are 2 commits. One on an anonymous branch that does
> s/-SNAPSHOT// and one on master that increments the MINOR portion of the
> version. So if master is 0.2.0-SNAPSHOT this new branch will have 0.2.0.
> For that commit we create an annotated (preferably PGP-signed) tag, 0.2.0.
> The other commit on master changes .auroraversion to 0.3.0-SNAPSHOT.
>
> Once everything is verified we'll push 2 things - the new master and the
> tag (so origin will not have a name for the branch the tag was created on).
>
> What does everyone think of this process?
> --
> Kevin Sweeney
> @kts
>
--
Zameer Manji
Re: Tagging process
Posted by Brian Wickman <wi...@twitter.com>.
+1
On Fri, Dec 13, 2013 at 6:26 PM, Bill Farner <bi...@twitter.com> wrote:
> This sounds good to me.
>
> -=Bill
>
>
> On Fri, Dec 13, 2013 at 6:20 PM, Kevin Sweeney <ke...@apache.org> wrote:
>
>> Hey all,
>>
>> I think figuring out the tagging process sooner than later is in
>> everyone's best interest so that we can get out of cherry-pick limbo. I'm
>> out next week, but if anyone wants to take a stab at it (using
>> https://reviews.apache.org/r/16265/ as a starting point or just throw it
>> away, please feel free). My thinking:
>>
>> .auroraversion on master only ever contains -SNAPSHOT versions. This
>> makes it harder for someone to reset and accidentally creating bogus
>> non-SNAPSHOT artifacts.
>>
>> At each tag there are 2 commits. One on an anonymous branch that does
>> s/-SNAPSHOT// and one on master that increments the MINOR portion of the
>> version. So if master is 0.2.0-SNAPSHOT this new branch will have 0.2.0.
>> For that commit we create an annotated (preferably PGP-signed) tag, 0.2.0.
>> The other commit on master changes .auroraversion to 0.3.0-SNAPSHOT.
>>
>> Once everything is verified we'll push 2 things - the new master and the
>> tag (so origin will not have a name for the branch the tag was created on).
>>
>> What does everyone think of this process?
>> --
>> Kevin Sweeney
>> @kts
>>
>
>
Re: Tagging process
Posted by Bill Farner <bi...@twitter.com>.
This sounds good to me.
-=Bill
On Fri, Dec 13, 2013 at 6:20 PM, Kevin Sweeney <ke...@apache.org> wrote:
> Hey all,
>
> I think figuring out the tagging process sooner than later is in
> everyone's best interest so that we can get out of cherry-pick limbo. I'm
> out next week, but if anyone wants to take a stab at it (using
> https://reviews.apache.org/r/16265/ as a starting point or just throw it
> away, please feel free). My thinking:
>
> .auroraversion on master only ever contains -SNAPSHOT versions. This makes
> it harder for someone to reset and accidentally creating bogus non-SNAPSHOT
> artifacts.
>
> At each tag there are 2 commits. One on an anonymous branch that does
> s/-SNAPSHOT// and one on master that increments the MINOR portion of the
> version. So if master is 0.2.0-SNAPSHOT this new branch will have 0.2.0.
> For that commit we create an annotated (preferably PGP-signed) tag, 0.2.0.
> The other commit on master changes .auroraversion to 0.3.0-SNAPSHOT.
>
> Once everything is verified we'll push 2 things - the new master and the
> tag (so origin will not have a name for the branch the tag was created on).
>
> What does everyone think of this process?
> --
> Kevin Sweeney
> @kts
>