The author's gripe seems to be that the hiding of dotfiles was unintended, ergo dotfiles are Bad. Whether they were intended is irrelevant; their wide usage vindicates the practice. After all, traction = value. The problem of program state/configuration/metadata storage is adequately met by dotfiles.
There are, no doubt, numerous unintended behaviors of programs. Most of these are simply ignored and certainly not leveraged the way the dot behavior is.
People don't go out of their way to abuse an unintended system behavior; they simply leverage all capabilities of a system ("intended" or not) to meet their needs. Had dotfiles not gained traction, some other solution would have been designed (or "engineered") to meet the needs of program state/configuration/metadata storage.
[Tangent: All of this reminds me of grammar freaks that harp on "correct" usage, completely oblivious to the fact that grammar changes, and "correct" is merely a lightweight pointer to the current norm.]
There are, no doubt, numerous unintended behaviors of programs. Most of these are simply ignored and certainly not leveraged the way the dot behavior is.
People don't go out of their way to abuse an unintended system behavior; they simply leverage all capabilities of a system ("intended" or not) to meet their needs. Had dotfiles not gained traction, some other solution would have been designed (or "engineered") to meet the needs of program state/configuration/metadata storage.
[Tangent: All of this reminds me of grammar freaks that harp on "correct" usage, completely oblivious to the fact that grammar changes, and "correct" is merely a lightweight pointer to the current norm.]