Vincent Woo

@iclems Cursor locations should update better on the current master. Can you check to see if this is still a problem for you?

I'm going to reopen this issue because I see it routinely on firepad head and I don't think its the transaction size thing.

To add - I think I see the base length mismatch a lot more on mobile devices or older browsers. Could it be string encoding issues?

Yes, we catch these exceptions. They don't generally coincide with other exceptions, so I don't think it's a post-crash thing. Is it possible to amend the firepad code to log...

Here's an example callstack (obviously the stuff is munged but you can see the general flow): This one is the client trying to process a remote op, but you also...

We're still consistently seeing these errors. I haven't had time to debug in depth. Is there any chance we can get some time from you to really dig into them?...

User reports on the topic usually go to the effect of "everything is running fine, but then one person can keep typing with no indication that their text is not...

Is there a general change we can make to have the syncing be less brittle? Can we do a, "okay, taking a step back, let's try to resync from scratch"

I'm seeing it across most major browsers these days. I'll uh, look into logging... everything.

That was a good hypothesis, but I checked the errors against language and they seem pretty evenly distributed across our normal language distribution. Proposal: What if the change listener didn't...