When you have a redundant array such as two mirrored disks, you have the question of "which one is wrong?" when data corruption occurs.

The magic of ZFS is that it is able to checksum the data and metadata, so when you have a mismatch it knows which one is correct and is able to repair it from the mirror. AFAIK this is not something which MD is capable of, because it needs cooperation from both the RAID stack and the filesystem itself.

BTRFS has to do the same thing (manage RAID itself) for the same reason.

OK, ignore that - I just checked my scripts and it looks like I'm misremembering things. I do in fact have the compression turned on. (And yes, 256K blocks).

Frankly, I don't know why it was doing that either, but that's what I was seeing. Big increases by switching off hardware compression. As you likely know, trying to compress data that's already compressed will tend to result in the output becoming bigger instead of smaller, so it's possible it was filling up the buffer more quickly and stalling.

Yes. If you need to move a song between two different sequencers, or you're doing this DAWless stuff and want to have a workstation or dedicated sequencer running everything, you're going to be exporting the song as a MIDI file because everything understands that.

Tape drives do this. The quoted capacity of e.g. an LTO cart is generally twice the physical capacity, assuming that the drive controller will be able to achieve a 2:1 compression ratio. Which it generally doesn't. You tend to get faster throughput if you disable hardware compression.

DOSbox for the authentic experience, but 'Chocolate Doom' or similar source ports will be easier to run out-of-the-box. Chocolate tries to resemble the original DOS version as closely as possible on modern systems, but others add OpenGL support and higher resolutions.

The idea of remaking Doom in unity makes me laugh, though - Doom's staying power was because it was easy to create custom levels, and there are many thousands of them out there, some of which were better than the stock levels. Making a remaster that can't play those is kind of missing the point about what made it awesome...

Bit of a catch-all term, that. There were consumer decks which could be pretty crappy, but remember that even long into the CD era, the songs were recorded in the studio on reel-to-reel tape. On a deck that cost as much as a house, and at high speeds on expensive tape, sure... but still reel-to-reel. The first digital recorders were reel-to-reel as well.

More bandwidth is definitely a big thing - MIDI runs at around 31 Kbit/sec and it's fairly easy to swamp it, especially if you're chaining instruments on a single bus.

Increasing the resolution is Really Big Thing. MIDI is 7-bit, which means that if you do something like sweeping a cutoff filter, you only have 127 possible values which gives you very noticeable stepping artifacts (often called 'zipper noise'). Some manufacturers try to interpolate in software. Others bond two controller streams together so that you get 14-bit precision, or send custom NRPM values but since there are a number of incompatible ways of doing this, you have to have a controller keyboard which works the same way as your synthesizer. Setting out an actually standardised way of doing this would be really handy.

Firstly, MIDI is opto-isolated. Without that, you get weird ground-loop effects like the data leaking into the audio, which happens quite a bit when using USB MIDI.

Secondly, MIDI is peer-to-peer whereas USB has a host and a guest. You cannot plug a USB MIDI keyboard into a USB sound module and expect it to work, you have to have a computer somewhere to act as a broker. With MIDI you can take two cables and link three machines together. USB1.1 doesn't work that way, and that's what the USB-DIN adapters are all using. USB-C might be better in that regard, I'm not sure.

You've also got a very large installed base (probably millions of machines) which are using DIN and USB1.1, switching to USB-C isn't going to. They bent over backwards to ensure that MIDI 2.0 is going to work with your $10000 OB-X with Kenton board.

To be fair, Karn Evil Nine is really three tracks welded together. It is certainly good for a coding marathon, though...

I would expect at least some of it was created by bored teenagers. Back when I was a member of that demographic I put far too much effort into taking things aimed at younger kids and twisting them into something that would today be called 'edgy' for the amusement of myself and my friends. Henry's Cat exhorting people to join an Irish terror group, Honey Monster from Sugar Puffs being turned into a mafia don, editing a Mah-Jongg christmas tile set so the three kings were assassinating baby Jesus in the crib and so forth.

It being the early 90s I had no internet access, and the alterations were done crudely with scissors and marker pen (or the Mah-jongg tile editor). Given the kind of content creation software available now, I can easily imagine some of this stuff being created by a 13-year old with pirated software and too much time on their hands.

"Gone into administration" means that a third party has been appointed to oversee the emergency rescue/shutdown of the company. I'm not sure how that's less intuitive than saying "Gone into chapter 7"...

I believe it's a protective layer over the photomask. I remember reading, last time the EUV thing came up, that TSMC were considering running without using one at all, and taking the hit that they'd need to replace the mask frequently. That might have been just for initial runs or something, I don't really know.

True, but you only really need to worry about coating, slitting and polishing. My understanding from audio tape is that you usually buy in the backing from Dupont or someone. It's still not a trivial process, but it's not impossible either.

ATR Magnetics actually did this. They had their own coating machinery made. That's studio recording tape, though, so the tolerances will probably be a lot lower than for ultra-high density digital media on extremely thin backing.