When checking whether a given username is valid, it being trimmed is now
also a criterion. All names with leading and/or trailing whitespace will
be considered invalid from now on.
A username will now always be trimmed before it gets assigned to a user.
That means that names with leading and/or trailing whitespace are now no
longer possible.
The functions snd_card_get_name and snd_card_get_longname allocate a
String and return 0 if success otherwise the negative error code.
There is a need to check the return value before using the allocated String.
This commit fixes this by checking the return value and exiting if the
return value is not zero.
Ref:alsa-project.org/alsa-doc/alsa-lib/group___control.html#gad48d44da8e3bfe150e928267008b8ff5
The functions snd_card_get_name and snd_card_get_longname allocate a
String and return 0 if success otherwise the negative error code.
There is a need to check the return value before using the allocated String.
This commit fixes this by checking the return value and exiting if the
return value is not zero.
Ref: #4910
Co-Authored-By: Zero0one1 <summerlinasity@gmail.com>
Co-Authored-By: Krzmbrzl <krzmbrzl@gmail.com>
The fields of the rlimit struct were assumed to be of type long int when
printing via the qWarning function. This seems to be correct on e.g.
Ubuntu but at least on Alpine Linux this is not true.
In that case the wrong format specifiers are being used for printing
these values, which causes undefined behavior.
The fix is to always copy the values into variables of type long long
unsigned int and then printing these. The code also adds an assertion
making sure that within this copy operation the values won't overflow
the variable's size.
Fixes#4897
The cmake-based fork (mumble-voip/ice) recently has
received a new commit making changes that obviously also changed the SHA
checksum when installing this Ice version via vcpkg.
The mentioned commit is
mumble-voip/ice@a9d68a8
This commit therefore updates the mentioned hash in the corresponding
portfile.
The cmake-based fork (https://github.com/mumble-voip/ice/) recently has
received a new commit making changes that obviously also changed the SHA
checksum when installing this Ice version via vcpkg.
The mentioned commit is
a9d68a8d1d
This commit therefore updates the mentioned hash in the corresponding
portfile.
The functions snd_card_get_name and snd_card_get_longname allocate a
String for the caller that needs to be freed once it is no longer needed.
This was not done in the current implementation and thus this was causing
a memory leak.
This commit fixes this by explicitly freeing the resources after usage.
Fixes#4910
Ref: alsa-project.org/alsa-doc/alsa-lib/group___control.html#ga547dcead4a72f24db4a0f9f530f75228
The functions snd_card_get_name and snd_card_get_longname allocate a
String for the caller that needs to be freed once it is no longer needed.
This was not done in the current implementation and thus this was causing
a memory leak.
This commit fixes this by explicitly `free`ing the resources after usage.
Fixes#4910
Co-Authored-By: Zero0one1 <summerlinasity@gmail.com>
Co-Authored-By: Krzmbrzl <krzmbrzl@gmail.com>
On our self-hosted macOS runner we were seeing issues due to the
workspace not being cleaned up after each run, causing the next one to
fail.
This commit makes sure that Azure will clean the entire workspace of any
remaining files, before starting a new job.
Ref: docs.microsoft.com/en-us/azure/devops/pipelines/process/phases?view=azure-devops&tabs=yaml#workspace
The fields of the rlimit struct were assumed to be of type long int when
printing via the qWarning function. This assumption is correct for glibc
(used by e.g. Ubuntu) but it is not true for musl (used by e.g. Alpine
Linux).
In that case the wrong format specifiers are being used for printing
these values, which causes undefined behavior.
The fix is to always copy the values into variables of type long long
unsigned int and then printing these. The code also adds an assertion
making sure that within this copy operation the values won't overflow
the variable's size.
Fixes#4897
The current implementation for echo cancellation using Apple's acoustic
echo cancellation system has been reported to cause major problems
(users no longer being able to transmit any audio) and therefore we
disable it by default.
Furthermore we mark the option as experimental so that (hopefully) only
experienced users will use it.
Ref: #4912
The current implementation for echo cancellation using Apple's acoustic
echo cancellation system has been reported to cause major problems
(users no longer being able to transmit any audio) and therefore we
disable it by default.
Ref: #4912
openssl.org/docs/man1.1.1/man3/RAND_bytes.html:
If the entropy source fails or is not available, the CSPRNG will enter an error state and refuse to generate random bytes. For that reason, it is important to always check the error return value of RAND_bytes() and RAND_priv_bytes() and not take randomness for granted.
https://www.openssl.org/docs/man1.1.1/man3/RAND_bytes.html:
If the entropy source fails or is not available, the CSPRNG will enter an error state and refuse to generate random bytes. For that reason, it is important to always check the error return value of RAND_bytes() and RAND_priv_bytes() and not take randomness for granted.
Also remove unnecessary cast.
The previous dialog was simply a message box with a bunch of HTML in
order to give a little structure to the contents. This approach is very
limiting in terms of UI flexibility though and therefore this commit
replaces the old HTML approach with a dedicated dialog class that uses
proper UI elements instead.
While doing so, the ordering and grouping of information was also
changed in order to make it more suitable for the every-day-user.
Fixes#4733
The previous dialog was simply a message box with a bunch of HTML in
order to give a little structure to the contents. This approach is very
limiting in terms of UI flexibility though and therefore this commit
replaces the old HTML approach with a dedicated dialog class that uses
proper UI elements instead.
While doing so, the ordering and grouping of information was also
changed in order to make it more suitable for the every-day-user.
Fixes#4733
In this mode the sametext heuristic is disabled which _should_ fix the
issue of the Weblate PR's CI failing because of Weblate not applying
such heuristics.
