1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
|
todo 0.12.x:
best to do an intermediate release for documentation. 0.13 is not
there yet.
* check reference.txt
known bugs:
* running out of memory (using delay lines or loops) will likely crash pd.
short story: don't, use pdp_control's memlimit for limiting pdp's memory usage.
* no known other fatal (crashing) bugs. please let me know if you find one.
pdp/libpdp remark:
i'm working on a version 2 of pdp. until that is finished, pdp proper will
probably be frozen, except for bug fixes.
todo 0.13:
* yves shm code
* make fobs (all c object code usable in packet forth should be wrapped (automaticly??))
* pdp_xwindow event handling is not thread safe (fixed ?)
* fine tune udp protocol
* solve scheduling probs: pd+3dp = ok, pd+pdp+(3dp) = not
(lots of frags -> video slow, few frags -> audio drops.
pdp threading is not implemented everywhere and is a problem for feedback stuff. think about
a synchronous dropping scheme (central point of failure), like for 3dp. most of the threading
problems could be solved with the new threadsafe pd 0.37 -> check this out)
* add source type to source objects (noise,plasma,constant,...) (solve with forth)
* pdp_crop/pad
* seed plasma
* reorganize headers to cut down compile time (done for core stuff, move pd objects to forth)
* fullscreen x command -> add window manager suppression
* where to solve autoconvert? per type, or central in pdp_base? 2 cases:
- template based autoconvert: in pdp_base
- compatibility based autoconvert: i.e. pdp_add (solve in base: move_passive)
* finish high level packet conversion:
- move 8bit images to packet level (iow: cleanup pdp_qt, pdp_v4l, pdp_xv) (solve with forth)
- add bitmap packet support to source modules (optim for pdp_opengl usage)
(solve with sliceforth: pdp image processing should become independent of image encoding: at least
packed 8/16 rgb(a), and all planar rgb/yuv formats should be well supported: sliceforth feeders)
- get rid of pdp_type* methods
* fix mac/linux name differences (use pdp_video, pdp_movie, pdp_window)
* pdp_xv framerate doc + pwc addons
todo 0.14:
* stress test the memory manager. it looks like there's a bug that pops up after running for a while
* enable audio output on bttv cards in pdp_v4l
* fix cache slicer objects + thread problems (fixed in sliceforth)
* derive as much classes as possible from pdp_base class (obsolete: use forth instead)
* displacement warp, substitution warp, refraction warp, timespace warp, wormhole warp
* check the use of varargs for image processing dispatching routine (obsolete: forth)
* more scopes
* 2D affine transforms parametrized by vectors (center, x-axis, y-axis) (np1 + matrix processing objects)
* some abstractions around pdp_cheby (colour shape, ...)
* efficient rescalers: pdp_double pdp_halve (+ efficient arbitrary scaler in sliceforth)
* move float color conversion and float<->fixed point conv code to system
* crop, shift (pad+scroll)
* solve biquad (blur) mmx boundary init prob (bands at edges: solve with proper init?) (sliceforth)
* optimize blur effect (1 pole ? / dependency stalls) (sliceforth)
* jpeg/png packet and streaming support (pdp_netsend / pdp_netreceive : binary compat issues!!)
* find out why resampling code (warping) is so slow (reg->mem->reg->mem stalls?) (solve in sliceforth)
* mmx profiling (done: sliceforth)
* add audio resampling in pdp_qt~
* use pd_error instead of post for errors
* ascii art packet processors
* 3D time space interpolation
* colour keying
* motion tracking
* moebius transforms
* type multiplexing: find a way to use the same name for packet processors. (solved in forth: polymorhpy)
i.e. pdp_add should add images, but also vectors, ascii packets, textures,...
experimental stuff
* modify 2D rotation to be used as arbitrary rotation (givens)
* add 3D rotation
* frame rate limited delay line (burst delay?)
* optimize resampling code (mipmapped packed 16bit format?)
* better type handling
core is in place. todo:
- better (tree) search algorithm (breadth first + cost)
- add more converters
|
