On the value of maintenance

13 Mar 2019

Originally posted at https://tech.labs.oliverwyman.com/blog/2019/03/13/on-the-value-of-maintenance/

Most of our posts here talk about new software, but there’s also a lot of value in the longer-term maintenance of code and given I’ve done some of that recently, I thought it worth revisiting various earlier projects in that light. None of these are really large enough to have their own post, but collectively they’re worth noting.

• Potboiler continues to get ongoing work from me (and maybe eventually a real app using it), but of particular note recently is that it now has both cross-event dependencies and integration tests (using the serialisation work I talked about before)
• Wharf got some package upgrades (via Pyup.io that I talked about before noting some security issues), but that then broke it because it was fine for existing deploys but not for new ones. I added in CI support using Travis and then fixed a few other cases. It’s got a fair few users (judging from the level of responses to those issues), so it’s been good to get it properly tested.
• Sked got support for adding/filtering Whosoff calendars
• Panegyric got a new release fixing various smaller issues in it

Also, in the general theme of maintenance, this isn’t just useful on your own projects. An area of fairly easy contribution to open source work (as I occasionally also talk about) is keeping an eye out for build failures when visiting the project pages of any libraries you use. Those projects are usually happy to receive fixes, as I’ve done recently for both failure and error-chain while doing some of the Potboiler work above.

Previously: Serialising Rust tests Next: AWS Lambda and Actix: easy conversion of small web apps into serverless