Hi,
There is no right or wrong here. The compiler—and I—just consider best practice to write the open arguments after an agent declaration.
It is always the same trend with SmartEiffel and Liberty Eiffel: be explicit rather than implicit.
It may be slightly overkill for inline agents since it is quite legible as it is; but it is a good check for the non-inline agent declarations as it allows to at least make the number of arguments locally visible, and to ensure that the number of arguments is right.
It could have been worse; the best thing would be to spell out the
type of the arguments too (with curly brackets instead of question marks).
Also note that in all cases you will have to add the arguments part if you have closed arguments; in that case, the remaining open arguments still have to be explicitly written.