diff --git a/CHANGELOG.html b/CHANGELOG.html index 860f5f40..1baab309 100644 --- a/CHANGELOG.html +++ b/CHANGELOG.html @@ -87,7 +87,7 @@
Enable or disable SIMD and the run-time compiler.
+Enable or disable SIMD.
This can give a nice speed-up, but can also be unstable on -some systems or with some versions of the run-time compiler.
-Enable or disable SIMD and the run-time compiler.
+Enable or disable SIMD.
Bool indicating if SIMD and the run-time -compiler should be turned on.
+Bool indicating if SIMD should be turned on.
This can give a nice speed-up, but can also be unstable on -some systems or with some versions of the run-time compiler.
-Get the major, minor or patch version number of the libvips library.
diff --git a/api/NetVips.SourceCustom.ReadDelegate.html b/api/NetVips.SourceCustom.ReadDelegate.html index 8e168a28..b849080b 100644 --- a/api/NetVips.SourceCustom.ReadDelegate.html +++ b/api/NetVips.SourceCustom.ReadDelegate.html @@ -123,10 +123,10 @@The interface is exactly as Read(byte[], int, int). The handler is given a number
-of bytes to fetch, and should return a bytes-like object containing up
-to that number of bytes. If there is no more data available, it should
-return 0
.
The interface is exactly as Read(byte[], int, int).
+The handler is given a number of bytes to fetch, and should return a
+bytes-like object containing up to that number of bytes. If there is
+no more data available, it should return 0
.
The interface is the same as Write(byte[], int, int), so the handler is -given a bytes-like object to write. However, the handler MUST return the number -of bytes written.
+The interface is the same as Write(byte[], int, int). +The handler is given a bytes-like object to write. However, the handler MUST +return the number of bytes written.
NetVips is fast and needs little memory. The NetVips.Benchmarks project tests NetVips against Magick.NET and ImageSharp. - NetVips is around 23 times faster than Magick.NET and 3 times faster than ImageSharp.
+ NetVips is around 20 times faster than Magick.NET and 3 times faster than ImageSharp.