invalid quantization options now result in an error rather than defaults

[imager.git] / TODO

diff --git a/TODO b/TODO
index 933fe2e512298493cb33ae086e398e4b969b48e8..61fb9634f2ec22eab8ddaa93618293ad998ac13d 100644 (file)
--- a/TODO
+++ b/TODO
@@ -6,21 +6,139 @@ Release Plans (subject to change)
 brown-bag bugs may add intermediate releases.  The dates are goals,
 not commitments.
 
-For 0.58 (hopefully):
+For 0.65:
 
-RGB file support (#8666)
+PNG re-work (#29268)
 
-and probably more
+For 0.64: (speculative)
 
-For 0.57:
+thick lines (#28854)
 
-rgb.txt search (and testing) (#26064)
+have alpha channel work treated as coverage for primitives (more
+general combine parameter)
 
-guassian blur radius (#25645)
+add some extra combine modes (hard/soft light, screen)
 
-text rendering FT1 (#11972)
+For 0.63:
 
-text rendering T1 (#11972)
+general alpha channel fixes (#29879, etc) (done)
+
+give 4/2 channel images a background color when saving to pnm (#30074)
+(done)
+
+give 4/2 channel images a background color when saving to BMP (#30075)
+(done)
+
+give 4/2 channel images a background color when saving to JPEG
+(#29876)
+
+
+For 0.62:
+
+TIFF improvements (to be detailed) (#20329) (done)
+ - read/write 16-bit RGB w/ and w/o alpha
+ - read 16-bit CMYK w/ and w/o alpha
+ - read/write 32-bit RGB w/ and w/o alpha
+ - read 32-bit CMYK w/ and w/o alpha
+ - write mono images as bi-level
+ - read bi-level image as mono/paletted
+ - read/write 16/32 grayscale w/ and w/o alpha
+both contig and non-contiguous
+ - possible extra code for handling 8-bit CMYK (#29353)
+
+Makefile.PL --incpath and --libpath should expand ~/ paths (#29484)
+
+scaling alpha channel image issues (done)
+
+PBM images inverted (#30859) (might be fixed in TIFF branch) (was,
+added test, done)
+
+don't load B at compile-time in t82inline.t (#30508) (done)
+
+the hardinvert filter inverts the alpha channels, this probably isn't
+desirable (#30002) (done)
+
+"dh-make-perl --build --cpan Imager" fails on require Makefile.PL
+(#30520) (fixed)
+
+Wrong link in Imager Pod (#30889) (done)
+
+more diagnostics in 5/t57infix.t (fixed the problem)
+
+extract out the scale() scale factor calculations (done)
+
+For 0.61:
+
+regmach.c fixes/tests (#29296) (done)
+
+sample: scaling an animated gif (#27591) (done)
+
+transform2() docs (#29267) (done)
+
+i_gsamp() chan_count validation (#28985) (done)
+
+Imager::Font::Wrap doesn't correctly set savepos (#29771) (done)
+
+testdrive warnings (#29610) (done)
+
+Imager::Files still refers to rgb file type (#29354) (done)
+
+t/t01introvert test failure line 495 (#29413) (done)
+
+Imager::Expr::Infix problem (#29562) (done)
+
+CMYK alpha TIFF problems with old libtiff (#29353) (done)
+
+matrix_transform() should use fabs() instead of abs() range checking
+sz (#29936) (done)
+
+warning compiling datatypes.c on x64 (#30204) (done)
+
+dynfilt/t/t60dyntest.t relies on a file that may not exist (#30203)
+(done)
+
+Imager::Font::Wrap doesn't correctly set savepos (#29771) (done)
+
+For 0.60 (hopefully):
+
+RGB file support (#8666) (done)
+
+convert() doesn't preserve sample size (#28492) (done)
+
+be explicit that scale*() doesn't modify the source (#28570) (done)
+
+repeated text in Imager::Files (#27589) (done)
+
+error message when loading an unsupported font format is confusing (#27571) (done)
+
+enable transparency by default for gif (#27615) (done)
+
+Imager doesn't load on Windows 98 (#27653) (done)
+
+new color counting code (#28142) (done)
+
+sample: scaling an animated gif (#27591)
+
+test getcolorcount() (done)
+
+do a compiler warning clean up run (done)
+
+change the ico/cur reader to apply the mask as an alpha channel by
+default (#29001) (done)
+
+For 0.58:
+
+rgb.txt search (and testing) (#26064) (done)
+
+guassian blur radius (#25645) (done)
+
+text rendering FT1 (#11972) (done)
+
+text rendering T1 (#11972) (done)
+
+For 0.57
+
+BMP security fix only
 
 For 0.56:
 
@@ -288,6 +406,19 @@ New Features:
 
 - arc outlines
 
+- filled/outlined ellipses and elliptical arcs
+
+  Complexities include:
+
+  a) when drawing a pie segment, are the angles specified the real
+  angles (since the scaling changes the angles if you just use sin() and
+  cos() directly) or the angle before scaling?  I can see cases where
+  both are useful.
+
+  b) can you control the angles of the axes of symmetry? (the axes would
+  remain at right angles to each other) Does this change the effect of
+  the d1 and d2 parameters?
+
 - complex lines/curves:
   - thick lines:
     - dotted/dashed lines