[SRU discussion] Renaming the 'Regression Potential' section

[Robie Basak]

I appreciate you bringing this up and proposing improvements!

On Thu, Jul 30, 2020 at 10:36:02AM +0100, Iain Lane wrote:
> I often come across SRU bugs from developers that seem to treat the 
> Regression Potential section as a place to argue why their upload is not 
> risky and should be accepted. Like
> 
>   [ Regression Potential ]
>   Low. This only changes X Y Z.
> 
> I'm not going to bother repeating the arguments for why that's
> wrong...

Agreed. From the other side, I also find this frustrating when I need
this information to weigh up the acceptability of an SRU or whether the
proposed test plan is appropriate or should have something additional
added to it.

> My *straw man* proposal is to rename the section to something like 
> 'Where problems could occur' or something more explicit than what we 
> currently have.
> 
> This is obviously partly/mostly a problem that the expectations haven't 
> gotten through to people, so while I think we should rename, any change 
> will need to be communicated so that people know about it and then 
> enforced by the SRU team for a little while.

On Fri, Jul 31, 2020 at 01:31:36PM +0200, Balint Reczey wrote:
> IMO the section name is quite clear and reflects what's needed.
> Changing the name just puts an extra burden on people filling those
> SRU bugs to change muscle memory...

I don't think any SRU team member will reject or delay an upload on the
basis of the exact name used to describe the information -  provided
that the required information is made available in the bug. So I
wouldn't worry about changing the name on this basis. On the other hand,
if as Iain says a different name helps people provide the necessary
information, then I'm in favour of changing the documented name to use.

In other words, even if we change it, go ahead and keep typing
"[Regression Potential]" if that's what your muscle memory wants to do.
I won't care unless the actual required information is not present.

Since I don't care about the exact name used, changing things as Iain
proposes wouldn't really be a change in SRU process at all. It would
just be a change to a recommendation that helps uploaders provide the
required information to reduce review ping-pong.

>                                ...but if the provided content is not
> enough/wrong please give feedback about it and do it in a timely
> manner. Having filled SRU information in a bug and getting a feedback
> one week later after the the upload is put in the queue is extra
> frustrating.
> There is a vanguard every day for SRU uploads thus please process each
> upload the next business day the latest and give feedback if the
> upload can't be accepted either due to insufficient filling of the SRU
> template or due to problems with the upload itself.

I appreciate the frustration. I can think of some ways to improve this,
but I think that's a matter for a different thread. Separately, note
that as a community team it isn't reasonable to expect the SRU team to
provide an SLA. If as a Canonical employee you'd like SRU team members
who are also Canonical employees to arrange themselves to reduce review
latency, then that's something to take up internally within Canonical. I
don't think complaining about it here is going to be useful.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <https://lists.ubuntu.com/archives/ubuntu-release/attachments/20200731/e64bedf1/attachment.sig>