Vampire Survivors, Chicory, Hard Drive Failures and Google's approach to Enshittification
Sunday computing is Funday computing.
Welcome to the Sunday before the election in the United States. Before you ask, I indeed cannot believe it is as close as it is. If you want to get a vibe on how I’d vote, maybe go read Nilay Patel’s article on The Verge “A vote for Donald Trump is a vote for school shootings and measles”
As therapy, I’ve been tidying up the various data systems in my life while running an Idle Vampire Survivors in Gold Farming mode.
I assure you, that is totally a real sentence/ Let’s break it down:
Vampire Survivors is a “Bullet Hell” game that’s just , well, perfect. It runs on PC/Steam Deck1 and you earn gold in the game that you can spend on new characters, features and the rest.
This isn’t a loot box/in app purchase scam or anything, it’s just play and advance your options as you get better at it. You can get new maps by beating old maps in certain ways, you can improve your characters chances by evolving weapons and stats, etc.. it’s … very silly, fun and appeals to the completionist in me.
While playing, you can get the game into a state where you’re so buffed (improved) that nothing can kill you and you just sit still and collect gold, etc. So Friday night, I started a run on Vampire Survivors to accumulate gold over the weekend. I had an exciting excursion to Seattle and Duwal, Washington over Saturday.
You see, Duwal is where Local Roots has their farm store. You’ll be interested, surely, in that Local Roots are the Italian chicory kings of Washington state. They even plant and harvest Puntarelle, an extremely hard to find chicory that is just terrific with an anchovy dressing. If you find yourself visiting Microsoft in Redmond, it’s really only a half hour from there to visit their farm store. So do that!
Another reason I run Vampire Survivors, is it makes it nigh impossible to use my computer for anything but my email, my to-do list and writing down my pithy thoughts. Sitting in my email are two that have demanded my attention.
The first was one I started getting on Thursday about a degraded drive in my NAS2, a 12tb Exos drive that was throwing errors. I kicked off a deeper scan and it’s now degraded so much that it’s not part of the storage pool anymore. I have a new drive arriving tomorrow that should fix the problem.
For all the talk of hard drives and unreliability, I don’t think I’ve had a spinning platter drive actually fail3. I had one NAS running for 10 years non stop and I’d imagine if I plugged it in today it would start up like nothing had happened. The MTBF4 of these drives is something like 2.5 millions hours and this one fails at a mere 18,000 hours? That’s nuts , right? It must have been dropped during manufacturing or shipping, I figure.
This isn’t a warning about Seagate drives, mind you. The late, great, Luiz Barroso co-wrote a paper “Failure Trends in a Large Disk Drive Population“. Luiz and I worked for/around Alfred Spector’s research organization at the time. We talked ahead of releasing the paper about whether we should or should not mention manufacturers directly.
He explained that, depending on the year, some drives would be good, some would be stinkers. And the the next year, season, acquirer, or whatever, it would swing the other way. Is it Hitachi? Seagate? Western Digital? Fujitsu? It basically didn’t matter. You could count on it being one of them, probably. We would build it into the contracts and make it so if drives fail before they should, then we would have recourse/credit for new drives. I’m pretty sure I’m personally hosed and out the money for the drive, but if another fails maybe I’ll pursue it with Seagate.
The second email was a note from Google’s search console that my personal site was having trouble and throwing 404s. But… uh… it’s hosted on Google’s own Sites product. So how is that so? While working at Google, Sites was one of those projects that was shuffled around quite a bit. I guess the initial team had their promotions sewn up and moved on to other projects. I’m sort of waiting for it to get shutdown, and like the aforementioned hard drive it’s likely in some degraded state inside Google.
There’s a pretty terrible management pattern at Google. My friend Cory calls it enshittification but it’s sort of different, and goes something like this:
A project launches to much fanfare! Or little fanfare and great Adoration!
Promotions and bonuses all around!
The long, hard work of growing and evolving, monitoring and valuing the project ensues.
Some Engineers and all the Product5 people go to the next shiny product.
The product languishes or declines in quality/reliability, engineering, marketing and budget for managing the user base are moved to shiny new projects.
The product declines in use, revenue or importance.
Someone notices this either because it is somehow competitive with the thing they’re launching6 (reader) or because they get some upside from cutting stuff?
Higher ups say, wow, look at how unimportant this project is, it’s languishing. Who is using it? Any key cloud customers? Top line advertisers? Can we soak any of them for more revenue? If not, should we shut it down? If yes, goto 11. Should we merge it, poorly, into another product?
They don’t shut it down for whatever reason. Instead, the tendency is to stash it somewhere where “Costs are lower” for maintenance or migration to a new product that shinier. Then colleagues in India7 are tasked with figuring out how to take a legacy codebase and get or keep it running on a constantly moving underlying platform.
Goto 8
Shut it down, often with little care for the users, then collect their reward for doing so and then skedaddle before there’s any blowback on them personally. Spin it out? Sell it to another firm8? (as you saw with Domains to Squarespace, Tumblr to Yahoo to Wordpress, etc… ) There are no wrong answers in enshittifcation town!
So where is Sites in that pattern? When I left google in 2023, it was sitting firmly in 9. Given I’m getting warnings from the Search Console about a Sites page, I’d imagine it has graduated to line 11. All but dead for the scheduling. I’d expect an announcement sometime in the spring. Yeah, right before the earnings call when they have the Christmas season ad-sales to juice the numbers. Additionally, they can show how thrifty they’re being by killing a bunch of projects and laying off the people who work on them. Unless they find a way to sprinkle AI on it… maybe that’ll help?9
So if you’re planning your 2025, and you use Sites, I’d consider what new host you’ll put your stuff on.
Yes, on consoles too. Everything, basically.
Network Attached Storage.
Or since those original hard drives in the 80s. I think I had a MFM 10mb drive go on an IBM PC, but it was under warranty. Damn those things were expensive.
Mean time before failure. Neat acronym!
Or they quit in disgust.
Google shut down reader and other social networking stuff to put it ‘all in on google+’ donchaknow….
Please note, I’m not coming down on these folks, imagine the difficulty of being flung neglected projects and being told to ‘make them work’ with some other , unrelated, project in your portfolio.
As you saw with Google Domains move to Squarespace, Tumblr to Yahoo to Wordpress, etc…
It won’t, but maybe it’ll keep it going for another year?