On Saturday, 16 July 2016 05:31:41 UTC-4, Jan Mercl wrote: > > On Sat, Jul 16, 2016 at 10:39 AM T L <tapi...@gmail.com <javascript:>> > wrote: > > > I know this result, but I just need the reason, not the result. > > The method set distinction of T vs *T allows to forbid calling methods > intended to mutate the receiver (ie. receiver is *T) on the wrong receiver > type (T). The later case would mutate the value (not the value that > should be pointed to) stored _in_ the interface, not the value at the > caller's site. Declaring the method receiver as *T forces callers to use > the properly typed thing (*T instead T) they pass via interfaces as the > compiler will reject the incorrect case. >
Here's another way to think about it: the methods of T are the methods appropriate for any *value* of type T. The methods of *T are the methods appropriate for a *variable* of type T (the variable pointed to by the *T pointer). The methods of *T can read the value of the variable, but they can also update it. So you can do more with a variable than a (non-addressable) value, and that's why *T has more methods than T. -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
