Satie had some good instructions:

"In the morning on an empty stomach"

"Hypocritically"

"With a lot of difficulty"

“With conviction and a rigorous sadness”

“With a healthy superiority”

“Don’t eat too much”

“Shake like a leaf”

“Do not cough”

“Go away”

“Like a nightingale with a toothache”

Should apply that to lines of code in code review .

   x.ugly_hack=blah <- “With conviction and a rigorous sadness”

> Midi step sequences would have a hard time representing the "emotion" some composers are looking for.

They all allow different amounts of precision and interpretation of the performance. Staff notation is open to interpretation on the performance level. MIDI is an exact recording (or programming) of a performance.

With staff notation you can mark eighth notes and say "staccato, lag behind the beat", and there's an infinite number of subtly different ways to play it, even within the own composers interpretation. With MIDI, it would represented as "Note on A3, 12 pulses after first quarter note, velocity of 87. Note off A3, 54 pulses after first quarter note, velocity of 0. Note on C3 (etc....)". MIDI is great for computers (or routing signals during live performance) because it's exact.

I don't think that's quite right. Different midi sequencers will play in a different way, based on the decisions of the midi sequencers author.

Not quite sure what you mean by a different way, or sequencers author. Do you mean the musician, or like Roland/Korg/ARP?

MIDI is a pretty exact specification. If a sequencer is changing the timing of anything, the underlying MIDI structures are not the same then. You can have different amounts of swing or different PPQ values, but then I wouldn't consider that "the same MIDI".

You can feed it to different sound generators of course, different drum machines, synths, a laptop, what-have-you, (most sequencers have both builtin together) but that's different from the MIDI itself.

That's not how it works. MIDI is a standard.

Given a specific MIDI file, with specific applied quantization, they should absolutely play the same way -- the only exception is the timing resolution they offer and smallish latency issues (which in practice should be 100% transparent).

Because a machine is playing, so as with any computer code, you have to formalise everything and leave nothing implicit.

So much of the notation is not codified, and is part of the artistry of the musician. Even if all those Italian/French words are included (martele, staccato, largo, et al.), and crescendos and ritards and are marked, one cannot play the tune from sight with the desired emotion without first hearing it or playing it over and over.

>Midi step sequences would have a hard time representing the "emotion" some composers are looking for.

Most classically trained players do too.

Besides we have much more control over step sequences than you probably think, e.g.

http://www.steinberg.net/en/company/technologies/vst_express...

Plus, of course, tons of CC, etc.

Compared to the possibilities for expression in step sequencing, something like a piano, which basically just has velocity and sustain/dumper is not even close.

It can be done, you just need to know the variables to tweak and how.