On Sunday, February 19, 2017 at 3:41:13 AM UTC-5, Marwan abdel moneim wrote: > > i wanted to do it without a Mutex > but there still something not clear to me, but i don't know what it is > and i don't understand what "trigger" synchronization means > > i will keep reading, and go back to it later and see it i get it > > thanks for your help > > To correctly do any sort of write access to the same location from multiple goroutines *will *require synchronization. Without it, your code will never be correct. It may work some of the time, or even all of the time on some platform with some version of the compiler. But it will never be guaranteed to always work, unless you use synchronization. If you really don't want to use a sync.Mutex <https://golang.org/pkg/sync/#Mutex>, then you should look at sync/atomic <https://golang.org/pkg/sync/atomic>. But be forewarned, doing anything but the most trivial synchronization with atomics is hard o get right.
On a more general note, I think there is sometimes an over eagerness to attempt "lock free". If you are trying to solve a real problem, I suggest thinking about channel based solutions. If they are not appropriate, just use sync.Mutex, or one of the other synchronization types. Only try to optimize to "lock free" (atomics) if you find there is an actual problem with lock contention under real conditions. -- 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.
