> They've got their own package manager, own internal ec2 clone, own build system, own Linux distro, own deploy system
This exactly describes the current cadre of internet giants as well, and it does seem nuts, but it's important to realize why this happens. For one thing, these systems were often developed and made to run at scale well before their better known brethren appeared. Sometimes those better known brethren eventually surpass the internal tools. Sometimes they never do.
> The first thing they make you do is port to their tech stack
There's no simple answer to whether this is a good idea. Switching costs are a real thing. Even if what you're switching to is better, that superiority has to outweigh the cost of the transition itself. There's a cost to the parent company switching, and there's a cost to the acquiree switching. There's also a cost of running a bazillion different software stacks on ten kinds of hardware and operating systems, as I know AOL did in those days. Maybe that's part of the reason they stopped being a viable standalone company. The most successful post-acquisition tech merges I've seen have been very deliberate about the costs and benefits of changing each part, not adopting blanket "always" or "never" rules.
This exactly describes the current cadre of internet giants as well, and it does seem nuts, but it's important to realize why this happens. For one thing, these systems were often developed and made to run at scale well before their better known brethren appeared. Sometimes those better known brethren eventually surpass the internal tools. Sometimes they never do.
> The first thing they make you do is port to their tech stack
There's no simple answer to whether this is a good idea. Switching costs are a real thing. Even if what you're switching to is better, that superiority has to outweigh the cost of the transition itself. There's a cost to the parent company switching, and there's a cost to the acquiree switching. There's also a cost of running a bazillion different software stacks on ten kinds of hardware and operating systems, as I know AOL did in those days. Maybe that's part of the reason they stopped being a viable standalone company. The most successful post-acquisition tech merges I've seen have been very deliberate about the costs and benefits of changing each part, not adopting blanket "always" or "never" rules.