> This is demonstrably wrong, and I and others have already provided examples, which you've ignored. I don't know what exactly you consider to be an example to

> This is demonstrably wrong, and I and others have already provided examples, which you've ignored. I don't know what exactly you consider to be an example to demonstrate your point, but I must have missed it because I haven't read anything factual on the matter. If you think merely listing a bunch of infrastructure components and deeming the difference between external and internal versions is sufficient evidence, I have already stated that it doesn't convince me. Maybe we have vastly different experiences in the software development we have participated in our careers, but I think the root of our estimation mismatch is you equivocating rewriting the infra glue code with the entirety of the application. I certainly don't see it that way. > Moving from stubby to GRPC is a lot of effort. Moving from Google's internal user ACLing to OAuth or whatever is a lot of effort. Switching the APIs used for Google's internal monitoring infra with the ones that are used by GCP is a lot of effort. Removing dependencies on borg particulars is a lot of effort. Doing all of those things is a lot of a lot of effort and reasonably cost prohibitive. And that's ignoring a lot of other sources of "a lot of effort". To my point, I have explicitly stated this infra glue code will be in need of re-writes. While "a lot" doesn't help quantifying your proposition much, you (or others in this thread) seem to equivocate this to designing and implementing the whole application from scratch, and that is where I disagree. Sure this going to be application specific, but I would estimate the engineering hours spent on replacing the infra glue code to be much less than that went to developing the business logic. I don't deem that to be cost prohibitive when we are talking about a company potentially selling that IP, much less so if it is going to run in their public cloud platform. Besides, I don't know much about Borg, but if an application like an RSS aggregator is taking dependencies on implementation details of their replicated task scheduling and provisioning system, I think they must have been doing it very, very wrong. > I don't know how I can make this any more clear. Your estimation is incorrect. You are misinformed. Keeping repeating your assertions is definitely not going to help making anything clearer. Let your arguments do the talking, let the readers arrive to the conclusion on who is wrong or possibly misinformed. You are bordering on shallow dismissals which HN guidelines frown upon. It is bad form, please don't keep doing this.