Alexander Gagarin

Jason @jagerman, totally agree with an approach of having `argument_policy` the same way as we already have `return_value_policy`. Is it ever planned to be implemented?

We also had long story of tuning CAF performance on Windows. And yes, sharing scheduler works stable and predictable, but you can aggressively tune stealing to outperform sharing (but probably...

> @uentity Thanks for sharing! I suppose this isn't running on a Desktop, though. This is running on a desktop =) I understand that thread pool is huge, but that's...

> The OS scheduler is going to add some amount of latency, but that shouldn't be in milliseconds range. I just remember a noticeable post that compared a spinlock implementations...

@Neverlord oh, it's interesting! You have switched to `shared_mutex`. When I have a time I will also try to check the results. Also we need to see if numbers changed...

@Neverlord sorry for late response - I had to update my codebase to modern CAF API to be able to compile with your branch. I can confirm that I see...

Same here. I also get a similar error about type `Short`: ``` ./myservice.go:54:56: undefined: String ./myservice.go:54:66: undefined: Short ./myservice.go:64:61: undefined: String ./myservice.go:64:71: undefined: Short ./myservice.go:74:56: undefined: Short ./myservice.go:74:65: undefined: String...

Can you please share the result of modification?

Hi! The feature you suggesting would be a very appreciated and useful thing! I vote for this with both hands as also having troubles with requirement that types must be...

You're right that library isn't actively maintained and it's a pity, because the overall design is not bad and from my point of view Cereal is the best existing replacement...