Wormholes

Great. You can tell CCP Masterplan to fix the other bug you posted up earlier this week also.

Your wormhole bughunters are on the case (damn SISI server can't keep anything working right).

*phew* well that's the correct way to deal with the sticky situation!

This is a pretty hardline stance for you Dusettey, I'm so proud <3. Don't get your panties too twisted, it's a pain to get them back to normal.

Troll in progress. Nothing will ever be posted in those threads.

I'll answer this since I fixed the bug (and I also caused it ). Sometime disclosing this sort of thing is good for the soul. Hopefully you'll understand that things on a test server aren't always intentional, and you shouldn't always assume malice when a programmer has simply made a booboo! Making these changes available to the public as early as possible is great for feedback, as long as you understand sometimes you'll only see incomplete parts of a (perhaps-broken) picture.
[I'm intentionally simplifying the following story, just so I don't have to include a host of background material and explanations]

Each solar system has a 'wormhole manager' piece of software that looks after the wormhole states for that solar system. When you jump through a wormhole, the wormhole manager for your departing solar system performs a series of actions, one of which is that it notes down the time of the jump (for use in the polarization timer in case you come back again).

When I added the 'jumping through a WH makes the remote end appear scannable' feature, I added an additional action: Contact the wormhole manager responsible for the destination solar system, and tell it to make remote wormhole X appear scannable. This works as expected if both source and destination wormhole managers are hosted on the same server node, but would fail if they were on different nodes (because I hadn't included the explicit network permissions to allow this action to travel across the wire).

On Singularity, if you were travelling between solarsystems all hosted on the same node, the polarization timers would work as expected. However if you were travelling between solarsystems hosted on different nodes, the failure to complete the actions described above would leave the wormhole manager in a broken state, specifically it would not update the polarization timer. The way that we cluster certain types of solarsystem together on the same node made it look like there was a pattern between which jumps had working timers, and which had broken ones. However this pattern was entirely emergent according to how the load balancer spread solarsystems around the network.

So this was a story about how a wormhole one day tried to send a "Hey, you should make yourself scannable" postcard to another wormhole living in a different town, but when he was walking over to the post office he tripped over and hit his head, giving himself short-term memory loss.

So now I'm really curious, since I'm the nitpicky sort and don't like things not neat an orderly, why it is that you guys don't have some sort of log for those types of errors (aka when actions don't finish or bug out etc) and subsequently get alerts for when that happens?

We do. As soon as the build hit a multi-node test server and cross-node wormhole jumps started happening, the associated errors starting getting logged. However you are a diligent bunch when it comes to testing. I had this thread pointed out to me on the day it happened, and the patterns of reports in this thread were detailed enough to make me realise what was happening before I even got my morning coffee and started checking those logs!