[Osmf-talk] Seeking feedback and interest in the OSMF Engineering Working Group

Steve Coast steve at stevecoast.com
Wed Jul 21 15:23:27 UTC 2021

Metrics are great in the short-term, but in the long-term they get gamed.

The way that good organizations fix this is to change the metric at regular intervals, so that you only ever have short-term metrics.

For example, at Intel there used to be (and may be again) a “tick-tock” approach where every even year they (approximately) focus on faster processors, and every odd year they focus on making smaller and more efficient processors. Quarterly goals or OKRs accomplish a similar thing.



From: john whelan<mailto:jwhelan0112 at gmail.com>
Sent: Wednesday, July 21, 2021 5:19 AM
To: Andy Allan<mailto:gravitystorm at gmail.com>
Cc: OSMF Talk<mailto:osmf-talk at openstreetmap.org>
Subject: Re: [Osmf-talk] Seeking feedback and interest in the OSMF Engineering Working Group

I would agree a metrics driven approach is rarely the best approach.  It's too simple and sometimes things take time to gel.

Cheerio John

On Mon, Jul 19, 2021, 11:44 Andy Allan <gravitystorm at gmail.com<mailto:gravitystorm at gmail.com>> wrote:
On Thu, 15 Jul 2021 at 23:29, Michal Migurski <mike at teczno.com<mailto:mike at teczno.com>> wrote:

> Our software projects so far have not gained contributors and maintainers over time, and in fact the opposite has happened. iD no longer has Bryan or Quincy contributing to its development. The team behind the API & website has been reduced to you and Tom for many years. We are lucky to have had dedicated and skilled people like yourself, but the board has not chosen to prioritize or support a conducive environment for growing the community of engineers. This story has been consistent over the past five years and should be cause for alarm for the majority of community members who prioritized stability in the recent survey.

I feel I should point out that I wasn't even a maintainer 5 years ago!
So perhaps the story is the same, but the actual situation does
change, and so I reiterate my request that you double-check with the
people involved before drawing any more conclusions based on incorrect

Another example is including the JOSM pull request data in your
analysis, despite them not using github for development! That would
skew your analysis somewhat. Again, it's something that could easily
have been caught before you published your charts and targets.

> Instead I am proposing a way for the board to set a goal for the EWG using readily-available public data. We’ll know the working group is successful when we see a movement in the contribution and maintenance patterns on core repos.

Having worked in various metrics-driven organisations, I'm deeply
cautious about the choosing of targets. And I'm still deeply skeptical
of these "contribution and maintenance patterns" that you have chosen.
If the board starts measuring progress on those charts and your stated
goals, then there are two simple ways we can "improve" those outcomes
- simply close community pull requests immediately, and arbitrarily
decline merging maintainer PRs. That's a step backwards for everyone
involved, but would meet those defined goals.

So my hope is that you set those charts and goals aside, and
reconsider your approach.


osmf-talk mailing list
osmf-talk at openstreetmap.org<mailto:osmf-talk at openstreetmap.org>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/osmf-talk/attachments/20210721/0dadbc8d/attachment.htm>

More information about the osmf-talk mailing list