I'm torn. Part of me accepts that we're in a commercial age. But before PSR, plenty of Premier League owners put glory before profit. Owners like Alan Sugar were ridiculed. Feels a little like men's teams got to live on glory and credit card debt for decades but the women's teams won't get that.

I was impressed by James's tracking back today. She always seemed to go deep in matches where she wasn't getting on the ball, but since coming back from injury seems much more active in defence. Not sure if I'm just noticing it more.

It's hard to accept the idea that someone you like and admire might have done something horrible. For younger fans, this could be their 1st experience of it. And it's hard even after a few go arounds. The attacks on Caracas are vile, but a lot of people are just unsure and I won't write off everyone

Wow, Levante were stubborn, almost too much! Spent last few min of the match saying to the TV "Take the win, no don't go down again, take the win..." Broke up play effectively (if tediously) and Barcelona seemed to fall into the trap of forcing things and losing even more rhythm.

100% this. Lent someone an old umbrella. Later realised they'd been too British-edly polite to tell me the handle had decomposed into weird sticky plastic. Don't do that to your guests (like I did).

I'm surprised, but Walsh may be a better fit for Chelsea than for Arsenal. Arsenal maybe need a DM with a little more pace, a little more "destroyer". Barcelona seemed to train the long ball switches out of Walsh. If Chelsea bring that back, happy days for their fast front line + for Lionesses.

Maelstrom seems to heavily rely on its big brother Jepsen and they're both written in Clojure. So I guess I'm learning a little Clojure now.

I figured Maelstrom's broadcast checker allowed a small number of missing values, given the relatively loose consistency of gossip systems.

But the exercise intro claims it checks all values are present on all nodes. Since that isn't the result I'm seeing, started poking around the Maelstrom code.

Working out the latency problem proved surprisingly easy. It was a good old fashioned race condition.

The real rabbit hole might be understanding why Maelstrom ever considered it a valid analysis. A couple of nodes had >=1 missing value.

Bug bounty hunters who haven't been devs sometimes seem naive about remediation. "It's been a week, why isn't it fixed??"

But for anyone who's exploited a discrepancy in state, for example, then kudos because chances are you found an inconsistency that's baked deep into the system.

Curious if Slegers has given them problems (better than expected results, now they don't know what to do)? Or are they struggling to hire for what, from the outside, seems like one of the top job in women's football?

Hmm. Maybe I need to prioritise understanding maelstrom results. The nodes work fine until `--nemesis partition` and then, even though I know what's happening, I understand a fraction of the output.

Working theory is that CIN was formalised as the XNS protocol suite. Wikipedia for XNS mentions the Clearinghouse directory service. Since that's what I'm reading about, I'm fairly confident that was originally part of CIN.

364 was once randomly assigned to me. 1 year - 1 day was easy to remember, and I liked the serendipity, so it stuck.