- github.com/evanphx/json-patch: [v4.2.0+incompatible → e83c0a1](https://github.com/evanphx/json-patch/compare/v4.2.0...e83c0a1)
- github.com/fsnotify/fsnotify: [v1.4.7 → v1.4.9](https://github.com/fsnotify/fsnotify/compare/v1.4.7...v1.4.9)
- github.com/go-kit/kit: [v0.8.0 → v0.9.0](https://github.com/go-kit/kit/compare/v0.8.0...v0.9.0)

		{"(*Reader).Size", Method, 5},
		{"(*Reader).UnreadByte", Method, 0},
		{"(*Reader).UnreadRune", Method, 0},
		{"(*Reader).WriteTo", Method, 1},
		{"Buffer", Type, 0},
		{"Clone", Func, 20},
		{"Compare", Func, 0},
		{"Contains", Func, 0},
		{"ContainsAny", Func, 7},
		{"ContainsFunc", Func, 21},
		{"ContainsRune", Func, 7},
		{"Count", Func, 0},
		{"Cut", Func, 18},
		{"CutPrefix", Func, 20},

Colours are the yellow and orange. These affect the Senses more strongly
than all the rest together, and next to these in strength are the red
and green. The blue compared with these is a faint and dark Colour, and
the indigo and violet are much darker and fainter, so that these
compared with the stronger Colours are little to be regarded. The Images
of Objects are therefore to be placed, not in the Focus of the mean

	// match: (Load <t1> p1 (Store {t2} p2 x _))
	// cond: isSamePtr(p1, p2) && t1.Compare(x.Type) == types.CMPeq && t1.Size() == t2.Size()
	// result: x
	for {
		t1 := v.Type
		p1 := v_0
		if v_1.Op != OpStore {
			break
		}
		t2 := auxToType(v_1.Aux)
		x := v_1.Args[1]
		p2 := v_1.Args[0]
		if !(isSamePtr(p1, p2) && t1.Compare(x.Type) == types.CMPeq && t1.Size() == t2.Size()) {
			break
		}

      should run synchronously, as opposed to be parallelizable when
      `options.experimental_optimization.map_parallelization=True`. This saves
      memory compared to setting `num_parallel_calls=1`.
* `tf.lite`
    * `Dequantize` op supports `TensorType_INT4`.
        * This change includes per-channel dequantization.

## Keras

Given quantized `lhs` and quantized `rhs`, performs quantized add on `lhs` and `rhs` to make quantized `output`.

`UniformQuantizedAdd` follows Numpy broadcasting rules.
The two input array shapes are compared element-wise.
Starting with the trailing dimensions, the two dimensions either have to be equal or one of them needs to be 1.

`lhs` and `rhs` must be quantized Tensor, where data value is quantized using the formula:

2F4F          ; mapped                 ; 6BCB          # 3.0  KANGXI RADICAL DO NOT
2F50          ; mapped                 ; 6BD4          # 3.0  KANGXI RADICAL COMPARE
2F51          ; mapped                 ; 6BDB          # 3.0  KANGXI RADICAL FUR
2F52          ; mapped                 ; 6C0F          # 3.0  KANGXI RADICAL CLAN

pkg time, const DateTime = "2006-01-02 15:04:05" #52746
pkg time, const DateTime ideal-string #52746
pkg time, const TimeOnly = "15:04:05" #52746
pkg time, const TimeOnly ideal-string #52746
pkg time, method (Time) Compare(Time) int #50770
pkg unicode/utf16, func AppendRune([]uint16, int32) []uint16 #51896
# freebsd riscv64 port
pkg log/syslog (freebsd-riscv64), const LOG_ALERT = 1 #53466