Alexandre Janniaux authored 1 year ago and Steve Lhomme committed 1 year ago

Current code can lead to the following error:

    invalid cast from type 'std::nullptr_t' to type 'const volatile info_t*'

Following documentation[^1] from cppreference.com:

    9) The null pointer value of any pointer type can be converted to any
       other pointer type, resulting in the null pointer value of that type.
       Note that the null pointer constant nullptr or any other value of
       type std::nullptr_t cannot be converted to a pointer with
       reinterpret_cast: implicit conversion or static_cast should be used
       for this purpose.

[^1]: https://en.cppreference.com/w/cpp/language/reinterpret_cast

Fix #28526

It works in VLC where there is no such check. So it should work in contribs as
well.

2e1b1609 claimed it was not supported.

Fatih Uzunoğlu authored 1 year ago and Steve Lhomme committed 1 year ago

Patch old MinGW versions to have
full d3d12 support. MinGW 8.0.0 has
partial support (d3d12sdklayers.h
missing).

Fatih Uzunoğlu authored 1 year ago and Steve Lhomme committed 1 year ago

DirectX 12 sdk provided by MinGW 8.0.1 is required for Qt 6,
and patching MinGW 5.0 to include d3d12 sdk requires sheer
amount of effort.

There doesn't seem to be any API changes: https://www.gnutls.org/manual/html_node/Upgrading-from-previous-versions.html

Many of our patches have been upstreamed.

Thomas Guillem authored 1 year ago and Steve Lhomme committed 1 year ago

Related to #28524

This will avoid the event callback being called from an input_ function.

Jeffrey Knockel authored 1 year ago and Steve Lhomme committed 1 year ago

We should not assume that the depth is the bits per pixel.  The RFB spec
says that bpp must be either 8, 16, or 32 and that depth must be <= to
bpp.

Jeffrey Knockel authored 1 year ago and Steve Lhomme committed 1 year ago

libvnc does not support decoding 8 bpp + tight encoding, so upgrade the
requested color depth to 16 to prevent libvnc returning an error when we
try to decode.

Jeffrey Knockel authored 1 year ago and Steve Lhomme committed 1 year ago

In the unlikely event that the server's native pixel format is not true
color, this flag will be received as zero, and so we cannot assume that
it has already been set.  Since we only support true color pixel
formats, we therefore need to ensure that the flag is set ourselves.

Jeffrey Knockel authored 1 year ago and Steve Lhomme committed 1 year ago

Outside of keeping bits per pixel consistent, we never made any attempt
to match the server's native pixel format in the first place, always
requesting our own hardcoded format for any color depth, so we might as
well also always ask the server for LE byte order (regardless of client
endianness) and code around that.  This change has been tested with LE
and BE clients.  Previous to this commit colors were wrong in the case
of (e.g.) LE <-> LE clients and servers.

Alexandre Janniaux authored 1 year ago and Steve Lhomme committed 1 year ago

Rewrite the first patch on top of v1.3.1 and simplify it since we always
need the same name for the static zlib target.

Remove the second patch since there is now a ZLIB_BUILD_EXAMPLES cmake
switch to disable the examples.

Claudio Cambra authored 1 year ago and Steve Lhomme committed 1 year ago

Signed-off-by: Claudio Cambra <developer@claudiocambra.com>

Claudio Cambra authored 1 year ago and Steve Lhomme committed 1 year ago

Signed-off-by: Claudio Cambra <developer@claudiocambra.com>

Vikram Kangotra authored 1 year ago and Steve Lhomme committed 1 year ago

Changelog: https://github.com/taglib/taglib/releases/tag/v2.0

`VLC_PATCHED_TAGLIB_ID3V2_READSTYLE` removed
see: https://github.com/taglib/taglib/commit/c13a42021a78038d5a72c75645abf55261ed833e

Rather than the attached format used for cropping.

Rather than the attached format used for cropping.

Ultimately we should assert that the incoming format is clean.

Rather than the attached format used for cropping.

Rather than the attached format used for cropping.

The value was 0 for text regions, no need to force it again.

Rather than the attached format used for cropping.

Rather than the attached format used for cropping.

Rather than the attached format used for cropping.

Rather than the attached format used for cropping.

Rather than the attached format used for cropping.

We already use the picture pitch.

We map the picture to a texture that is rendered  using the buffer dimensions.
The other format is mostly used for positioning.

We map the picture to a texture that is rendered  using the buffer dimensions.
The other format is mostly used for positioning.