or maybe you were saying… having a bash-like console inside* renoise, to allow direct access to signal flow?
Re: Scratching
How about as an addition to the idea of being able to setup trigger points for an alternative version of 0900, you could have a command rather like ‘glide to note’ but instead it’s ‘glide to trigger point’.
Lets say the sample has nearly finished playing and you start a ‘glide to trigger point’ to a point which is near the start of the sample.
It could be like this
0Xab : a=target trigger point b=speed
The play head could have a ‘velocity’ (perhaps set in the wave editor, and also automatable with pattern commands?) and depending on the speed set in the glide command, the play head would slow down to a stop, go in reverse, accelerate towards the target point, and slow down to a stop as it approaches the point. The sample would be held still (silence) until the ‘glide to trigger point’ command has ceased at which point the sample continues to play.
At this ‘held still’ point, you could change the ‘velocity’ of the play head to a low setting to make the sound ‘wind up’ to full speed.
A bit fanciful maybe, but I’d definately use this feature.
Hope that all makes sense btw 
hahah! No, it’s not that I want a bash shell inside renoise.
It’s that I feel that smaller, simpler, discrete processes that can be stacked are more nimble and flexible than those which are huge and have a specific purpose.
I used the vi shit as an example of this.
To give an example:
A filter or an action that has one distinct purpose only has as many different sounds as there are combinations of parameters; let’s call this f0. While the sound you may get from a complicated effect is "wacky or ‘what you want’ " you’re still @ f0 sounds.
Let’s say we have 8 basic effects { f0 … f8 } . If we can pipe these effects into one another then we have f8! (where f8 = f8 * f7 * f6 * … f0) possible values (states/sounds) in a chain that has all 8 effects in it. If there are seven effects in the chain, then we have 8 possible effects chains, 6 = 28, 5 = 56, 4 = 70, 3 = 56, 2 = 28, 1 = 8. (This is assuming that each of the effects can be used once in this chain.) So, there are 303 ways to combine 8 effects if each of them can only appear once.
Now, you’re probably saying to yourself - why the hell would I care about that if the one effect does what I want it to do, and I just have to type a value and it goes and sounds how I like it?
The point is flexibility: if the basic effects I mentioned above were all wave manipulations - fine grain manipulations of sound that allowed you to do a particular action on a particular waveform, then you would have 303 different possible “effects” per line. If you had something like a “scratching” program or something that did it for you, you would have one “effect” that you COULD possibly pipe into something else, but because it doesn’t have an output that you may necessarily be able to fully control (or may be slightly erratic since it’s doing the “scratching” for you), then your output may be erratic and you’ll have to sit and tinker to get the sound right.
The wave effects that we have now (arpeggio, pitch down, offset, pitch up, etc.) are great - and I think with a bit more intuition and hard thinking, more building blocks like these can be thought up.
All of this boils down to thinking about music and composition the way I do - I tend to think of tracking as a sort of modern-day sheet music. All of the “effects” in sheet music (triplets, volume changes, time signature, key changes, etc.) are easily identifiable and easily parsed visually. Likewise, they add a great deal of flexibility to the music and aid the composer in getting out what he hears in his head.
While there were some great experiments in the 20th century to move away from this paradigm (George Crumb’s “Black Angels” comes to mind immediately), I don’t see this as what a tracker is trying to do.
I tend to see these kind of broad strokes of effects, pleas for complicated effects to do legwork for easily replicable to move away from making the tracker about an extension of the artist and more of a toy.
hah! i did completely misunderstand you.
1st and formost i agree 100%!
the ability to do so many of the same types of things, many different ways, does set tracking aside from the other modes of composition. (except dsp packages)
i wouldnt like to ever see that go away. it is that freedom that is so appealing, but is so confusing for newcomers, an others trying to explain it.
maybe your idea could be as simple as this example of giving 1 effect # (i like asterisk) the ability to do 16 different but similar functions. such as modulators, oscillator types, etc… that would be intentionally strung along the extra effects columns
by giving them audio control -and- signal control, governed by switches of a predetermined type. so they could be compatible with each other without causing crashes.
okay i just changed my mind.
what i like most about tracking is the implicit sharing an exploration of new ideas. 
i like all the women you get when they find out that you track.
.x