Something I wonder about often: in software engine...
# random
Something I wonder about often: in software engineering, issue trackers are standard good practice since decades. In data engineering (tracking data issues) I very seldom see any (at least in my domain, transport and geo data) and then usually custom built. Do you have a different view? Or an explanation why?
Some of these issues may be process issues, and hence be tracked in software issue trackers of their asset defining software. Others are perhaps very short-living and immediately acted upon after a monitoring alert, mail or slack exchange. But for the rest?
Yeah I've been looking into open source issue tracking for this exact purpose (send slack alert, create issue) but there's a surprising lack of quality. Is it because systems like Jira have penetrated corporations? Are github issues enough for project management?
I started github issue trackers for German open transit datasets to share issues with the responsibles and other data users. Some data users and a few publishers registered as watchers and I think it is better than not having it. Others reject such public trackers as they could be interpreted as public shaming and prevent orgs from publishing their data openly in the first place. However, without buy-in from data publishers, this is just makes visible the demand. Two thirds of all created issues are still open. I‘d love to see an issue tracker for published datasets. JSON-LD metadata should link it, e.g. as discussionUrl. But seems like to few see a benefit for this. Regarding available issue trackers, I looked into Redmine this WE. There exist a bunch of plugins (eg Slack integration, Geo support) for it and it looks very configurable