[Sami_Lehtinen]

I got really distracted about totally inaccurate audio and tape functions. Haven't you ever used those devices? Timing, sounds, tape reel operation, highly in-accurate. Sorry to complain, idea is nice. But execution is really non authentic. You should have written authentic simulation code for audio and tape reels. Now it's just ehh, artistic stuff, just like in movies, which usually means bad. Also tape storage usually used faster seeks, slower reads and some pauses in between.

I also deeply hated Brother electronic typewriters, because I could always write faster than those typewriters were able to extract data from buffer. There was always an risk that you filled the buffer up and then had a overflow. Then you had to wait until the buffer was clear so you were able to see what was lost before continuing. Really enraging. IBM Selectic typewriters were much better and a lot faster. Anyway, main point is that print timing differs between letters based on what's being printed.

Of course you were able to connect both of these to computer and use as a printer.

Good: https://en.wikipedia.org/wiki/IBM_Selectric_typewriter Bad: http://electricinnards.blogspot.fi/2013/04/brother-ax-250-el...

This should be interesting at least to tech hackers. Old but good stuff. https://en.wikipedia.org/wiki/File:Selectric.ogv

[masswerk]

The sounds are actually extracted from footage showing this kind of equipment at work. The tape drive operations are controlled by random (based on what assumptions on the index and block locations should we do this?), but should usually feature some pauses and stops. The action of the tape tensioners is highly exaggerated in order to provide a substitute for the vacuum columns that can't be shown in this layout (providing a bit of a visual explanation for the asynchronous action of the reels).

Please understand that this is not meant to be a full emulation (like, say, Hercules compiled via Emscripten). Otherwise you would have to re-implement Google on top of zOS ... ;-)