It all comes down to management. There should be one person, or one person per platform e.g. iOS, who decides which features and which bugs are addressed and when. They should also ensure that there is a process for releasing fixes and new versions which includes regression testing. That means lumping changes together to reduce the testing overhead. It looks as if they just release changes at random. Now testing is time consuming, it may well be that they could release a candidate version to a select group of users who agree to help out, before releasing it more widely. And maybe send each one a LingQ baseball cap as a thank you.
Then there is the architecture side, whereby software is written in a manner that reduces complexity, and coupling. The Software As A Service approach is one methodology that tries to decouple code, making it easier to test and maintain, and reducing unexpected regression in other components. I’ve met too many engineers who write unmaintainable code, and are admired by management because they produce fixes quickly. Pity their colleagues who have to clean up their mess. I’m not suggesting this applies to LingQ engineers, they may be the mutts (brilliant), just that there are ways to write code that is more robust and stable, and perhaps there are areas they could improve.
Yes we are entitled to reliability. And to have basic bugs fixed. When I watch a YouTube video, sometimes I can’t even read the current text. I can tolerate some issues, such as some videos being unimportable, but not seeing the current text is serious. And not updating the current line when a user watches a YouTube video is serious, it is such a complete PITA.