[gommm]

I think that it would be good to dedicate a few months with no new feature and just focus on fixing all the regression bugs, there's a lot of bugs that are reported for many months to a year seemingly without activity.

I used to like Intellij but got fed up and just stopped using it. Dealing with regression bugs isn't worth it. An IDE should be extremely stable, releasing new features is great but it shouldn't be to the detriment of existing features.

[stefs]

https://www.joelonsoftware.com/2000/08/09/the-joel-test-12-s...

5. Do you fix bugs before writing new code?

[selcuka]

That might indicate a good place to work for a software engineer, but not necessarily a profitable business. If a bug affects 1% of your customers but a feature will potentially bring 2% more customers, the business will obviously prioritise the latter.

Disclaimer: I know that the above is an oversimplification but the point is there has to be a balance between fixing bugs and adding features. You can't wait until all reported bugs have been fixed.

[gommm]

I think that's often the reasoning behind this but I think the oversimplification doesn't help. It's usually harder to attract new customers than to retain them so retention is important and regression bugs affect that. They slowly erode the confidence until the customers who were your major advocates now become strong critics wherever they go.

[selcuka]

> I think the oversimplification doesn't help

True

> It's usually harder to attract new customers

To my defense that difficulty was factored in the 2% figure.

[gommm]

oh yes, I didn't mean to harp on you, what I meant is that I believe a lot of executives make the same calculation, oversimplifying things in the same way which leads to the repeated QA issues both Apple and Microsoft have or to the issues that Jetbrains has. New features are exciting and are seen as ways to get new customers.

Stability and bug fixes is not sexy and I believe underestimated as a way to retain customers but it's critical to customer happiness.

[bathtub365]

> A score of 12 is perfect, 11 is tolerable, but 10 or lower and you’ve got serious problems. The truth is that most software organizations are running with a score of 2 or 3, and they need serious help, because companies like Microsoft run at 12 full-time.

Over the weekend I spent a good chunk of a day fighting with a copy of Windows 11 for ARM preview that has a major bug wherein the Start Menu just straight-up doesn’t work. I’m not sure Microsoft is running at 12 full-time 21 years on from when this article was posted.

[gommm]

I don't think that Microsoft ever ran at 12 full-time given my experience with windows 9x... But then Joel is biased when it comes to Microsoft.

[joe_guy]

It was also 20 years ago when Microsoft didn't constantly release half finished versions of Windows.