A web address that does not exist is no longer tried three times

When the portal calls out to another site — an agent reading a page you asked about, for instance — and the call fails, it tries again. That is right for a network that hiccupped, and wrong for the one failure that will never improve: a hostname that does not exist.

Until now both were treated the same. A URL pointing at a domain that has been retired, or that was mistyped, was attempted three times, and each attempt was recorded as a platform error. Those error records are what the platform opens its own bug reports from — so a dead link in somebody's data produced a bug report about the platform, which had nothing wrong with it.

Such a call now fails once and is recorded for what it is. Everything else keeps retrying exactly as before, including the case that looks similar and is not: a name server that failed to answer is a real hiccup, and those calls are still tried again.

Reconnecting…
The connection to the server was interrupted. Trying to restore it…
Trying again…
The connection could not be restored. Reloading the page…
The server was updated. Reloading the page to pick up the latest version.