This is not what I am referring to. I am stating that with N generics types in a method the developer needs to write NxN case methods. Not feasible.
Sent from my iPhone > On Sep 19, 2018, at 9:25 AM, Wojciech S. Czarnecki <o...@fairbe.org> wrote: > > On Wed, 19 Sep 2018 06:52:09 -0500 > Robert Engels <reng...@ix.netcom.com> wrote: > >> Also, I did read what you wrote. I questioned “what happens when you have N >> different generic types in the call signature. The select type statement >> has N*N cases. “. You did not respond. > > The object code (i.e. ssa then machine) is instantiated **only** when its > **used** with given types set. > > If you really do call f.genpkg.(x) method on 200 different types for f, > each with 200 N typed vars x you will pollute symbol table **at compiling > time** (40000 entries you will not see until you dump the table). > > But resulting object code will have at most as many variants as M > and N have different base types / for-type-cases at most. > > It further can be size-optimized by factoring out type-sized operations > reusing a given code chunk for more than one base/case. > > Hope this helps, > > -- > Wojciech S. Czarnecki > << ^oo^ >> OHIR-RIPE -- 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.
