Why the risk was judged small
- Scale. One reply pointed at the sheer number of repositories in existence and how few are ever opened by anyone who did not write them, let alone by someone wanting to pass them off.
- Attention. Another put it as something you learn with age — people think about you far less than you expect, and we are our own harshest critics.
- The people you are worried about were described as unlikely to get far on stolen work anyway.
The reply from someone it happened to
One commenter had been copied several times: the whole project taken rather than forked, the credit changed to someone else's name, nothing else altered, and republished under their account. Their account of the outcome is the useful part — none of the copies gained traction. They blocked the accounts they found out about, concluded there was little more worth doing, and carried on sharing openly.
The cost of keeping it closed
The clearest argument against privacy was about who is actually looking. If the repositories exist to be seen by someone deciding whether to hire you, then an access request sits directly in that person's way, and most will not bother. A demo that links to readable code is a low-friction path; a demo that links to a permission screen is a dead end.
What to do with the fear itself
- Two replies pushed past the practicalities to the feeling, one suggesting the anger at an imagined thief is mostly ego, and that finding its root loosens it.
- Another pointed to the argument that you are not your code — a separation they noted is easier to see in employment, where you do not own what you write anyway.
- A different tack: make the code part of how you communicate, by writing or speaking about what it does. Then the ideas are attached to you, and someone holding only the files has very little.
- If the repositories are really there as a portfolio, one reply's answer was to build a portfolio instead, and let it be the thing on show.
One point the thread raised and left open
A reply argued the realistic route by which code now gets taken is training data rather than a classmate, without attribution and without anyone knowing whose it was. It was offered as a reason the calculation has changed, not as a reason to keep things private — the same reply still concluded in favour of publishing. Nobody in the thread tested the claim.