[Gridflow-cvs] [svn] commit: r6596 - in /trunk: ChangeLog TODO
svn-gridflow at artengine.ca
svn-gridflow at artengine.ca
Wed Feb 9 15:47:15 EST 2011
Date: Wed Feb 9 15:47:14 2011
New Revision: 6596
--- trunk/ChangeLog (original)
+++ trunk/ChangeLog Wed Feb 9 15:47:14 2011
@@ -1,6 +1,14 @@
/* $Id$ */
-version 9.13 (2011.02...):
+version 9.14 (2011.02...):
+ * [#convolve] runs somewhat faster (when using default op/fold/seed)
+ * [#to~]: add second argument for buffer size (in samples)
+ * [#to~]: report "buffer full" at most once per incoming grid.
+ * [#from]: fix crash upon deleting object
+ * [#from]: add second argument for buffer size (in samples)
+version 9.13 (2011.02.08):
* added classes [azerty_piano], [hpgl_to_lines], [hpgl_objectbox],
[hpgl_messagebox], [hpgl_from_patch], [#draw_lattice]
@@ -20,8 +28,7 @@
* former numop "atan" renamed to "atan2cd"
* new numops: asin acos atan asinh acosh atanh
* numops log and C.log aliased to ln and C.ln
- * [#inner]: add seedless mode
- * [#convolve]: add seedless mode
+ * [#inner] and [#convolve]: add seedless mode (without typecheck annoyance)
* [unix_time]: fixed the month number bug
* [#fold]: much faster in the case folding over a dim that has only 1 index
* some more work on the Win32 port
--- trunk/TODO (original)
+++ trunk/TODO Wed Feb 9 15:47:14 2011
@@ -1,13 +1,17 @@
-necessary for 9.13 :
[ ] do domain name transfer
+[ ] [gf/error] can't take a list. (see [#is])
+[ ] mmx not properly detected anymore (bug in ruby ?)
+[ ] [#rzero_space]
+[ ] rename [#lop_space] to [#rpole_space] ?
+[ ] [#convolve]: remove "a" ?
[ ] fix grid attribute setters
[ ] list-attrs
[ ] multi-valued attributes in source_filter
[ ] reform GFDP so that pd43 and l2ork become debuggable.
+[ ] in [#convolve]'s muladd, use a+b*(-1) = a-b
+[ ] note-help : something wrong with speedlim ? ask patrice (see courriel 7 nov 2010)
[ ] memalign still doesn't exist on OSX and Windows ?
[ ] reverify gem.cxx on OSX Intel-vs-PPC. Did I see a problem with 3 channels again ?
[ ] "seedless" should be really skipping over the seed. but this didn't improve speed for [#fold].
@@ -17,6 +21,8 @@
[ ] #background_model plante
[ ] rajouter les nouvelles patchs de Roman
[ ] option to disable input when in edit-mode, for [#see] and [gf/mouse_spy]
+[ ] [#from~]: pick your own blocksize (default = use same as pd)
+[ ] fix warnings in config.log on OSX
[ ] [#expr] local
[ ] more [#expr] : int rint imodf modf drem fmod size sum Sum
[ ] [fps] doesn't complain about invalid mode name
@@ -72,7 +78,6 @@
[ ] [#in]/[#out]: what happened to frame number in right outlet?
[ ] [doc_cc] [doc_ii] [doc_oo] where $1 is a symbol don't reposition themselves properly.
[ ] go back in all help files and fill in whatever is still missing
-[ ] [#draw_polygon] : some more weirdnesses in rendering. play a bit with #draw_polygon-test2.pd to see.
[ ] [#draw_polygon] : winding-rule, et multi-rule (ce dernier utile pour dessiner plusieurs polygones Ã la fois)
[ ] [#draw_polygon] : multi-rule optimisÃ© (a besoin d'une rÃ¨gle de distributivitÃ© du type : a+b+b = a+2b
[ ] put_at is lacking some checks for number-of-dims, causing segfault.
@@ -144,8 +149,7 @@
[ ] merge << and pd_oprint please
[ ] infinite loop in shoot.pd
[ ] change numop to numop2 all over the documentation
-[ ] renommer t_symbol * Ã symbol (?)
-[ ] renommer t_atom2 * Ã atom (?)
+[ ] renommer t_symbol * Ã symbol ou Symbol, et t_atom2 Ã atom ou Atom ? (peux pas renommer t_atom2 *)
[ ] #fold and #scan begin without seed but can't send "seed none" to [#fold +]
[ ] error: [# -] inlet 0 method grid: same type please (in has float32\; r has int32)
pourrait devenir :
@@ -157,20 +161,12 @@
error: [#for 0 0] inlet 0 method set: can't convert to grid: \\\\a(A_GRIDOUT 0x90ca8e0)
[ ] intÃ©grer -lib segfault
[ ] [lb] ?
-[ ] cellular_1d.pd: should get a variable width
[ ] finish adding PD_BLOBS A_BLOB class_addblob support (have to check for vanilla at runtime, etc.)
-[ ] while running #to~-help.pd :
- Invalid read of size 4
- upsampling_perform_linear (d_resample.c:92)
- Address 0x51fa2f0 is 0 bytes after a block of size 128 alloc'd
- calloc (vg_replace_malloc.c:418)
- getbytes (m_memory.c:24)
[ ] think about link between multidimensional [#for] and future [#decode] (or is it [#encode] ?)
[ ] allow direct passing of a grid as an atom of type Grid instead of just GridOut.
[ ] in case of some camera problems, [#camera] says OFF even when it's not closed.
[ ] regular dc1394 options
[ ] irregular dc1394 options: white_shading, white_balance; trigger_mode? trigger, trigger_delay
-[ ] [#in unicap] and/or [#in videojack] ?
[ ] test accuracy of RGB2Y etc
[ ] [#camera] [#cast] setting ? or not ?
[ ] fix SDL modifiers
@@ -187,13 +183,6 @@
to make a CHDH-like animations using [gf/gl].
[ ] see [comment] from cyclone
[ ] Dim::T = intptr_t
-[ ] creating a [pd~] while -lib Gem -lib gridflow are loaded. (it looked like related to GF, but in the end, probably not)
- Program received signal SIGSEGV, Segmentation fault.
- with gdb: canvas_getenv (x=0x8188868, name=0x819f558 "pd~", ext=0x8126d74 ".l_i386",
- dirresult=0xbfffb634 "", nameresult=0xbfffb630, size=4096, bin=1) at g_canvas.c:160
- with valgrind: Invalid read of size 4 at canvas_open (g_canvas.c:160) by 0x64705F76: ???
- Address 0x9c is not stack'd, malloc'd or (recently) free'd
- with no further info (2008.08.17)
[ ] canvasdelete
[ ] DEF_OP(clipunder, a<0?0:a>=b?b-1, 0, false, false)
[ ] DEF_OP(pingpong, ...)
@@ -213,6 +202,9 @@
[ ] [gf/numop_info]
[ ] algebraÃ¯c properties of some numop2s are not written down (0,false,false instead)
[ ] heat.pd still sucks (?)
+[ ] cellular_1d.pd: should get a variable width
+[ ] [#in unicap] and/or [#in videojack] ?
+[ ] [#draw_polygon] : some more weirdnesses in rendering. play a bit with #draw_polygon-test2.pd to see.
[ ] cvOpticalFlow
@@ -240,8 +232,8 @@
[ ] make example on separable convkernels
[ ] add concepts of hot inlet and conf inlet at the code level (?)
[ ] Add a video latency patch (in examples)
-[ ] Numop: warn on misalignment
-[ ] remove GridInlet* ?
+[ ] check whether we could benefit more from alignment of data (currently not much)
+[ ] rearrange FObject,GridInlet,GridOut structures in a different manner ?
[ ] [#expr] grid support
[ ] [#line]? [#pipe]?
[ ] rewrite [#rgb_to_rgba] [#rgba_to_rgb] in C++ if I'm not going to rework [#inner] soon enough.
@@ -284,9 +276,8 @@
[ ] accounting: grids packets numbers bytes copybytes
[ ] memory profiling
[ ] add a "query" class for debugging other objects? (would be able to extract extra info from each object/class)
-[ ] add a GridFlow::LaterException or something (which tells to retry later with the same packet because another
- grid reception is not completed yet.)
-[ ] SSE, liboil
+[ ] add a way to handle missing data so that we don't have to ever use [#t].
+[ ] MMX/SSE in C (remove Asm/Ruby code)
[ ] find a good profiler to use with GridFlow; did I look at http://oprofile.sourceforge.net/about/ ?
what about: qprofile, amd profiler, opreport -cl --demangle=smart
[ ] read http://en.wikipedia.org/wiki/Tacit_programming
@@ -389,7 +380,7 @@
How to release GridFlow :
svn copy svn+ssh://gridflow.ca/home/svn/gridflow/trunk svn+ssh://gridflow.ca/home/svn/gridflow/tags/$tag/
svn export svn+ssh://gridflow.ca/home/svn/gridflow/tags/$tag/ $dir
chmod -R go=u-w $dir && tar cfzvv $dir.tar.gz $dir && rm -rf $dir
More information about the Gridflow-cvs