There's a familiar little jolt that comes with opening your feed reader in the morning and seeing the name of a tool you depend on next to the word "acquired". This January it happened again, and I had the same sequence of feelings I always do: a flash of "good for them", a slower wave of "right, so what does this mean for me", and then the dull administrative dread of working out whether I now need an exit plan.
I'm being deliberately a bit vague about which tool, partly because by the time you read this the specifics will have shifted, and partly because the pattern matters more than the particulars. We've had a steady drumbeat of these over the last year or two. A self-hosting darling gets bought. A beloved CLI's company is "joining" a bigger one. A licence quietly changes from one three-letter acronym to another. The press release always uses the word "exciting". It is, for someone. Rarely for the person who just wanted the thing to keep working.
the press release is not for you
The first thing I've learned to do is ignore the announcement's tone entirely. "Nothing will change" and "we remain committed to the open-source community" are not statements of fact, they're statements of current intent, and intent is the cheapest thing in the world to revise. I don't think the people writing them are lying. I think they genuinely mean it on the day. But a year of board meetings and revenue targets does things to commitments that no amount of sincerity survives.
So instead of reading the tea leaves of the blog post, I look at structural facts. Who holds the copyright? Is there a CLA that lets the new owner relicense unilaterally? Is the project actually open under an OSI-approved licence, or is it one of the source-available arrangements that look open until you read clause four? Those things are checkable, and they tell you far more than the word "exciting" ever will.
the question is always: how trapped am I
The practical version of all this is a single question. If this tool went bad tomorrow, by which I mean rug-pulled the licence, gated the good features behind a cloud subscription, or simply stopped shipping fixes, how painful would leaving be?
For some things the answer is "trivial". A CLI that reads a standard file format and writes another standard file format owes me nothing and I owe it nothing. If it turns sour I swap it out over a coffee. For other things the answer is "ruinous", because the tool has quietly become load-bearing: it holds my data in a shape only it understands, or it's wired into a dozen automations, or my muscle memory has fused to it over five years. Those are the acquisitions that actually keep me up at night.
The honest reckoning is uncomfortable. The tools I'd most hate to lose are exactly the ones I've invested the most in, which is to say the ones I'm least able to abandon, which is to say the ones a new owner has the most leverage over. That's not paranoia, it's just the economics of the thing. You don't get acquired for a tool nobody depends on.
what i actually do about it
A few things, none of them heroic.
I keep my data in formats I could read without the tool. If a service exports to JSON or plain Markdown or a documented SQLite schema, I'm relaxed. If the only way out is a proprietary archive that imports nowhere, that's a red flag I should have spotted before I committed.
I take an export now, while everything is calm and friendly, rather than in a panic six months into a degrading relationship. The best time to test your escape hatch is when you don't need it. A backup you've never restored is a rumour.
And I try to be honest about the difference between mild annoyance and genuine harm. Not every acquisition is a betrayal. Plenty of small projects get bought precisely so they can hire people and actually fix the bug I've been moaning about for two years. Sometimes the new owner is the best thing that ever happened to a tool I love. I've seen it go that way often enough to not assume the worst on day one.
But I do start the clock. The moment a dependable thing changes hands, it moves, in my head, from "trusted" to "on probation". Not gone, not replaced, just watched a little more closely. I'll give it six months to show me whether the "nothing will change" was real or rhetorical. Usually the answer reveals itself in the small things: the release cadence, whether the changelog still mentions community contributors, whether the docs start growing a "contact sales" button where the install instructions used to be.
There's no grand conclusion here, just the boring discipline of staying portable. The tools come and go, get bought and sold and renamed and discontinued. The only durable strategy I've found is to love them a little less tightly than they'd like, and to always, always know the way out.