Email or username:

Password:

Forgot your password?
Top-level
Johannes S

@tychosoft @chancerydaily @riley @thomasfuchs even with CVS and SVN I had to alter history from time to time. Sometimes people use the systems in a very wrong way and commit the absolutely wrong stuff. Sometimes legal reasons require removal.

Being able to rewrite however allows me to work with small iterative commits, but present something reviewable.

But there are still systems, like Fossil, which do not like rewrites and are quite interesting.

2 comments
DELETED

@johannes @chancerydaily @riley @thomasfuchs If confidential information was revealed, its probably better to presume its already compromised rather than pretend it didn't happen and hope nobody noticed. Gross errors can always be resolved with a revert commit. At least that could be an argument I would make for having immutable change control; that the abuses mutable histories invite may be worse than the problems it may solve.

Johannes S

@tychosoft @chancerydaily @riley @thomasfuchs it's not only confidential information, but also information I may not distribute (Copyright etc.) also GDPR requires handling personal information. "Somebody may have cloned" is not an argument.

But yes, rewrites in published repositories shouldn't be done lightly.

Go Up