Extra reference in a commit message

[Ilya Khlopotov <ii...@ca.ibm.com>]

Hello,

We at Cloudant using our own bug tracker. Quite often we have two tickets
opened for the same problem.
One ticket is in ASF JIRA and another one in our internal bug tracker.
We have some automation, which compiles the issues we worked on in a
current release cycle.
However, it is extremely challenging to get it working right. Since it
needs to consult two independent trackers with no link between them.
We are considering adding a second reference in a commit message.

Does anyone have a strong opinion against it?
Would these unrelated references be too annoying for the community?

Best regards,
ILYA Khlopotov

Re: Extra reference in a commit message

[Joan Touzet <wo...@apache.org>]

No objection from me, if it makes things easier for you - as long as you
are aware that you may get requests from customers saying "Hey, I see
you used my bug tracker ticket id # in a CouchDB commit..." ;)

-Joan

----- Original Message -----
From: "Ilya Khlopotov" <ii...@ca.ibm.com>
To: dev@couchdb.apache.org
Sent: Friday, June 17, 2016 7:07:22 PM
Subject: Extra reference in a commit message


Hello,

We at Cloudant using our own bug tracker. Quite often we have two tickets
opened for the same problem.
One ticket is in ASF JIRA and another one in our internal bug tracker.
We have some automation, which compiles the issues we worked on in a
current release cycle.
However, it is extremely challenging to get it working right. Since it
needs to consult two independent trackers with no link between them.
We are considering adding a second reference in a commit message.

Does anyone have a strong opinion against it?
Would these unrelated references be too annoying for the community?

Best regards,
ILYA Khlopotov

Re: Extra reference in a commit message

[Alexander Shorin <kx...@gmail.com>]

On Sun, Jun 19, 2016 at 5:24 PM, Jan Lehnardt <ja...@apache.org> wrote:
>> On 18 Jun 2016, at 01:07, Ilya Khlopotov <ii...@ca.ibm.com> wrote:
>>
>>
>> Hello,
>>
>> We at Cloudant using our own bug tracker. Quite often we have two tickets
>> opened for the same problem.
>> One ticket is in ASF JIRA and another one in our internal bug tracker.
>> We have some automation, which compiles the issues we worked on in a
>> current release cycle.
>> However, it is extremely challenging to get it working right. Since it
>> needs to consult two independent trackers with no link between them.
>> We are considering adding a second reference in a commit message.
>>
>> Does anyone have a strong opinion against it?
>> Would these unrelated references be too annoying for the community?
>
> I’m definitely not a fan of this. Mostly because the reference is not
> transparent and will confuse new contributors coming in. And secondly that
> it is very easy to miss any useful discussion leading up to a patch, and
> the onus on Cloudant folks to copy and paste everything relevant into
> CouchDB JIRA. I’d prefer to “default to open”
>
> Is there an option to use ASF JIRA for the more substantial / commit-level
> issue tracking even inside Cloudant?
>
> I realise you won’t be able to move the Cloudant private bits over and that
> there are other complications with this idea, but maybe it makes sense to
> keep the private Cloudant bits in the Cloudant issue tracker and the open
> CouchDB bits in CouchDB JIRA?

+1

--
,,,^..^,,,

Re: Extra reference in a commit message

[Paul Davis <pa...@gmail.com>]

This immediately struck me as not a good idea and I've been mulling it
over to try and figure out why. But I think Jan has nailed it here.
Anything that any private company needs changed in CouchDB needs to
move that through normal ASF procedures. Having conversations on
private bug trackers I think leads too easily to private conversations
off the dev list about development which we all know is bad for the
community.

Anyone that needs to track work on private issue trackers should
figure out how to make that private tracker line up to JIRA issues and
not vice versa. That said I wouldn't be opposed to looking into JIRA
to see if there are integration points that we could configure for
automated mirroring for anyone that wants it.

But yeah, what Jan said. Public bits are public and leaking private
bits into that pool is going to be at least confusing and very well
could end up being a whole lot worse.

Paul

On Sun, Jun 19, 2016 at 9:24 AM, Jan Lehnardt <ja...@apache.org> wrote:
>
>> On 18 Jun 2016, at 01:07, Ilya Khlopotov <ii...@ca.ibm.com> wrote:
>>
>>
>> Hello,
>>
>> We at Cloudant using our own bug tracker. Quite often we have two tickets
>> opened for the same problem.
>> One ticket is in ASF JIRA and another one in our internal bug tracker.
>> We have some automation, which compiles the issues we worked on in a
>> current release cycle.
>> However, it is extremely challenging to get it working right. Since it
>> needs to consult two independent trackers with no link between them.
>> We are considering adding a second reference in a commit message.
>>
>> Does anyone have a strong opinion against it?
>> Would these unrelated references be too annoying for the community?
>
> I’m definitely not a fan of this. Mostly because the reference is not
> transparent and will confuse new contributors coming in. And secondly that
> it is very easy to miss any useful discussion leading up to a patch, and
> the onus on Cloudant folks to copy and paste everything relevant into
> CouchDB JIRA. I’d prefer to “default to open”
>
> Is there an option to use ASF JIRA for the more substantial / commit-level
> issue tracking even inside Cloudant?
>
> I realise you won’t be able to move the Cloudant private bits over and that
> there are other complications with this idea, but maybe it makes sense to
> keep the private Cloudant bits in the Cloudant issue tracker and the open
> CouchDB bits in CouchDB JIRA?
>
> Best
> Jan
> --
>
>
>

Re: Extra reference in a commit message

[Jan Lehnardt <ja...@apache.org>]

> On 18 Jun 2016, at 01:07, Ilya Khlopotov <ii...@ca.ibm.com> wrote:
> 
> 
> Hello,
> 
> We at Cloudant using our own bug tracker. Quite often we have two tickets
> opened for the same problem.
> One ticket is in ASF JIRA and another one in our internal bug tracker.
> We have some automation, which compiles the issues we worked on in a
> current release cycle.
> However, it is extremely challenging to get it working right. Since it
> needs to consult two independent trackers with no link between them.
> We are considering adding a second reference in a commit message.
> 
> Does anyone have a strong opinion against it?
> Would these unrelated references be too annoying for the community?

I’m definitely not a fan of this. Mostly because the reference is not
transparent and will confuse new contributors coming in. And secondly that
it is very easy to miss any useful discussion leading up to a patch, and
the onus on Cloudant folks to copy and paste everything relevant into
CouchDB JIRA. I’d prefer to “default to open”

Is there an option to use ASF JIRA for the more substantial / commit-level
issue tracking even inside Cloudant?

I realise you won’t be able to move the Cloudant private bits over and that
there are other complications with this idea, but maybe it makes sense to
keep the private Cloudant bits in the Cloudant issue tracker and the open
CouchDB bits in CouchDB JIRA?

Best
Jan
--

Re: Extra reference in a commit message

[Andy Wenk <an...@apache.org>]

Hey Ilya,

I don’t think this is a problem at all. Naturally the commit messages are very short anyway so extending it a bit is not a deal ;-)

All the best

Andy

--
Andy Wenk
RockIt!

Hamburg / Germany

GPG public key: https://pgp.mit.edu/pks/lookup?op=get&search=0x4F1D0C59BC90917D

> On 18 Jun 2016, at 01:07, Ilya Khlopotov <ii...@ca.ibm.com> wrote:
> 
> 
> Hello,
> 
> We at Cloudant using our own bug tracker. Quite often we have two tickets
> opened for the same problem.
> One ticket is in ASF JIRA and another one in our internal bug tracker.
> We have some automation, which compiles the issues we worked on in a
> current release cycle.
> However, it is extremely challenging to get it working right. Since it
> needs to consult two independent trackers with no link between them.
> We are considering adding a second reference in a commit message.
> 
> Does anyone have a strong opinion against it?
> Would these unrelated references be too annoying for the community?
> 
> Best regards,
> ILYA Khlopotov