Vlad Sudzilouski
Vlad Sudzilouski
@GaryWilber, how would you call that ID then? we need some name for it. :)
Please add breaking.md note about these changes, how users can temporarily disable it, and the fact that they will lose such capability in the future, thus need to fix any...
@milanro, hard to give any advice here, but looking at the error, and diff between the types, it's expected that you will use MapFactory (or SharedMap.getFactory()) instance instead of SharedMap.
Got it! @skylerjokiel, @anthony-murphy, can you please glance where we failed with violation of minor versions not making breaking changes? Thanks!
@agarwal-navin, ops are not kept around, if connection is active, all of them are sent out. If they were lost in translation, then runtime make retry to send them on...
I think you are pointing to something else. Client can be offline with pending changes and switch to read-only mode. In this case all local changes will never make it...
I'm not sure I like this approach - I think you are overcomplicating it. I think it's Ok not to have a reason for disconnected state when container just loads...
@markfields, @vivekmogalapalli, if you really want to go with keeping disconnected state around, then I'd suggest decoupling it out of ConnectionState into separate variable - same as we track clientId....
@daesun-park, is the description of the PR updated based on recent changes? Did I get it right that what you are trying to do is - Allow new types on...