On 11.11.2018 07:47, Branko Čibej wrote: > Our usage of @ in paths may be > well-defined too, but its interaction with aggressive canonicalization > creates unintentional side effects in the implementation.
A really educational example of this is svn_cl__copy. It goes out of its way to ignore a specific kind of peg-revision parsing error. I'm having a really hard time trying to convince myself that can be the right thing to do. If one particular command has to try to fix parsing errors in order to work correctly, then I'd say that the parser is just wrong. -- Brane
