I've been looking at the other comments on this article, and there is a couple of points that I haven't seen address by others that I think are pertinent. 1) W

I've been looking at the other comments on this article, and there is a couple of points that I haven't seen address by others that I think are pertinent. 1) What was it that Starlink wanted to beta test with customers? It might very well be that they shipped hardware with know environmental weaknesses, so not to hold up other parts of the project. While taking a hit on the reliability in edge cases. 2) We know that there are some dishes out there that have heat issues, but we don't know how many dish hardware iterations they have done. The failing dishes could very well be early prototypes that have know issues. Which are fixed in later revisions. Generally speaking I don't see these reports as worrying. I've just come of 10+ years of working on SW on (indoor), embedded hardware. Having alpha and beta hardware with issues are common, you need some hardware to run your test and SW development on. Even if it is fundamentally flawed and you have to throw it away because of those weakness later. For those that only have experience of building software. What you are seeing here, is a view into the sausage factory, and just because it feels messy and doesn't line up with your expectations. Don't mean that they are doing it wrong.