I always look at the patch releases to help determine when a good time is to update Unity. Looking at the bug fixes, it just seems strange with what is going on with Unity. Right now Unity is managing 4 different production releases: 5.6, 2017.1, 2017.2, and 2017.3.
Here is some data I put together to show the current state of 2017.
The bug fixes for 2017.2 are decreasing as less reported bugs are active
The bug fixes for 2017.3 are increasing as Unity is working on 2017.3 the resolve the bugs, which is the latest version
But what is going on with 2017.1? The number of fixes are increasing per time, which I would expect the opposite to happen. What is accounting for this?
Your data doesn’t include any fixes in any of the official releases, only the patch releases. Maybe 2017.1 is just getting a lot of backported fixes from 2017.2 and 2017.3 official releases in the 2017.1 patch releases?
Lay them out on the same timescale. Show fixes per month on the y, and month on the x. I think you will find this sort of comparison will give you much more interesting data. Fixes per release probably isn’t super relevant.
2017.1 has an update every month, 2017.3 every week. They probably back-port more fixes in one month, than in one week.
PS: Great idea with tracking the number of fixes. I was actually curious myself for a while, because I imagine Unity adds more issues than fixing bugs. Would be interesting to see your “bug-fix chart” in relation to the number of bugs added to the issue tracker
Glad it works out now. I spend a lot of time working with stats in my day job. This is not an unusual error at all.
I have a hunch that if you keep tracking it for a few more months, you will see there is also a seasonal effect. The Dec-Jan holiday season wreaks havoc on general productivity. Its hard to make much progress when half of your team is taking leave.
