David Buchmann

its an error, yes, but in production mode you rather want things to survive than a hard exception. it can log an error, or if it can know we are...

yeah agree, trigger a warning level error sounds like the right thing to do. ignoring the cache in that case could lead to cache stop working at some point down...

when we started with the message factory interfaces (and the client interface, as well) we all agreed that both things would best fit into a PSR as they are meant...

the current situation seems to make no sense. completely removing them seems the consistent approach. i was considering if we should have a configuration option, but the only purpose of...

@krewetka i was referring to "It would be good to have an option to put it to context without writing our own logger." - my suggestion is to create a...

from https://www.php.net/manual/en/function.spl-object-hash.php > This function returns a unique identifier for the object. This id can be used as a hash key for storing objects, or for identifying an object, as...

fixed in #211 i think the actual problem was the same problem that i recently ran into with some other library, that non-strict comparisons of spl hashes can mess up...

what is the problem, can php-spec not use actual message objects instead of mocks? i'd rather stay consistent with the testing tool if that is not adding another complexity

nothing is obvious for me with phpspec :-P should we mix phpunit and phpspec there, or just write some phpunit tests in addition? we should not overcomplicate things, having one...

if MockClient makes things harder, can we use message implementations with a phpspec mocked client? i see no reason not to do that if thats easier.