[chuck-users] heads up for 1.2.0.7
mrcarp at comcast.net
mrcarp at comcast.net
Sat Sep 9 20:54:13 EDT 2006
One thing that comes to mind regarding to the new API naming is that it'd be faster to type the deprecated names rather than their new names. I also like how the all lower-case names look in the code.
Matt
-------------- Original message ----------------------
From: Ge Wang <gewang at CS.Princeton.EDU>
> Greetings fellow ChucKists,
>
> Here is a (not so) quick summary/warning about upcoming changes in
> chuck-1.2.0.7, to be released probably next week.
>
> what it don't got:
> - garbage collection is still on a holding pattern and will not be in
> 1.2.0.7
> - the long-overdue string operations, unfortunately, insist on waiting
> for garbage collection, and also will not in 1.2.0.7
> - file i/o api's have been implemented (thanks to Martin Robinson),
> will be included soon but not in 1.2.0.7
> Sorry about that.
>
> Now that we've apologized about what's not in the release, it's time to
> apologize about what is. Due to recent discussions on this list, two
> sets of changes and additions have been made. We want to preview them
> here to get feedback before setting them in concrete.
>
> API deprecations. Our discussion more or less determined that it's
> best to unify the naming scheming now before things get further out of
> hand, particularly for UGen names and such. So the current solution
> deprecates the object names in question and provides more consistent
> naming. By default, when a deprecated name, such as 'sinosc', is
> encountered, a warning will be issued, such as 'line(1): deprecated:
> 'sinosc' --> use: 'SinOsc''. The behavior should be otherwise
> unchanged. In other words the program should still work as before.
> We've also added a --deprecate flag that allows you to stop, warn, or
> ignore upon encountering a deprecated item. The default is 'warn'.
>
> Here are the objects deprecated so far, and their replacements:
>
> - (api) deprecated --> new classes
> --------------------------
> sinosc --> SinOsc
> triosc --> TriOsc
> sqrosc --> SqrOsc
> sawosc --> SawOsc
> pulseosc --> PulseOsc
> phasor --> Phasor
> osc --> Osc
> noise --> Noise
> cnoise --> CNoise
> impulse --> Impulse
> step --> Step
> halfrect --> HalfRect
> fullrect --> FullRect
> gain --> Gain
> zerox --> ZeroX
> delayp --> DelayP
> sndbuf --> SndBuf
> pan2 --> Pan2
> mix2 --> Mix2
> onepole --> OnePole
> onezero --> OneZero
> polezero --> PoleZero
> twopole --> TwoPole
> twozero --> TwoZero
> biquad --> BiQuad
>
> std --> Std
> math --> Math
> machine --> Machine
>
> A few more UGen's. From another discussion on list, also by popular
> demand, it seems convenient to have out-of-the-box filters, in addition
> to general 1st and 2nd order filters (OnePole/Zero, BiQuad, etc.).
> We've examined and plundered from Csound, SC, and PD source code, and
> chuckified a few UGen's. We are going to do this in installments,
> adding various UGens every release. In 1.2.0.7, They include:
>
> LPF : lowpass filter (2nd order butterworth) (also with
> cutoff resonance control)
> HPF : highpass filter (2nd order butterworth) (also with
> cutoff resonance control)
> BPF : bandpass filter (2nd order butterworth)
> BRF : bandreject filter (2nd order butterworth)
> ResonZ : resonant filter (BiQuad with equal-gain zeros)
> FilterBasic : base class to above filters
>
> Of course, all of these can be designed and implemented using existing
> filters, but these provide direct access to cutoff and Q out-of-the-box
> and behave in a familiar um Butterworthy way. (More are on the way)
>
> We'd appreciate your feedback on these. For example:
> 1. api changes are in general, not reasons for group celebration. what
> can be done to further smooth the transition?
> 2. new naming optimal?
> 3. anything else
>
> Thanks!
>
> Best,
> chuck team
>
> _______________________________________________
> chuck-users mailing list
> chuck-users at lists.cs.princeton.edu
> https://lists.cs.princeton.edu/mailman/listinfo/chuck-users
More information about the chuck-users
mailing list