Do not set FUSERMOUNT_DIR since we already set it though
NIX_CFLAGS_COMPILE.
Setting it twice results in a lot of warnings like this one:
<command-line>:0:0: warning: "FUSERMOUNT_DIR" redefined
which makes the build look potentially broken.
This doesn't affect the result but fuse3 will now build without any
warnings and building fuse yields less warnings.
GCC 8.x is required to avoid the following error:
cc1plus: error: -Werror=class-memaccess: no option -Wclass-memaccess
If we build without "-Wno-error=class-memaccess" the build should fail
and the upstream requirements were raised from GCC 7 to GCC 8 anyway:
https://github.com/telegramdesktop/tdesktop/commit/b2f821d3deb65ade7416d28a18148176f6c40b34
The two additional patches are required to avoid build errors like this:
/build/source/Telegram/ThirdParty/qtlottie/src/bodymovin/bmbase.cpp:123:12:
error: invalid use of incomplete type 'class QDebug'
qWarning()
^
In file included from
/nix/store/1qk60bhxcf95iicms7mkyjq3h7lz136p-qtbase-5.12.0-dev/include/QtCore/qglobal.h:1200,
from
/nix/store/1qk60bhxcf95iicms7mkyjq3h7lz136p-qtbase-5.12.0-dev/include/QtCore/qalgorithms.h:43,
from
/nix/store/1qk60bhxcf95iicms7mkyjq3h7lz136p-qtbase-5.12.0-dev/include/QtCore/qlist.h:43,
from
/nix/store/1qk60bhxcf95iicms7mkyjq3h7lz136p-qtbase-5.12.0-dev/include/QtCore/QList:1,
from
/build/source/Telegram/ThirdParty/qtlottie/src/bodymovin/bmbase.h:31,
from
/build/source/Telegram/ThirdParty/qtlottie/src/bodymovin/bmbase.cpp:29:
/nix/store/1qk60bhxcf95iicms7mkyjq3h7lz136p-qtbase-5.12.0-dev/include/QtCore/qlogging.h:57:7:
note: forward declaration of 'class QDebug'
class QDebug;
^~~~~~
Changelog:
https://cryptography.io/en/latest/changelog/#v2-7
Important changes:
- BACKWARDS INCOMPATIBLE: We no longer distribute 32-bit manylinux1
wheels. Continuing to produce them was a maintenance burden.
- BACKWARDS INCOMPATIBLE: Removed the
cryptography.hazmat.primitives.mac.MACContext interface.
The CMAC and HMAC APIs have not changed, but they are no longer
registered as MACContext instances.
"Most of the new features for this release are in the devlink and rdma utilities.
And most of the bug fixes are in fixing the output format glitches
that resulted from converting most of the tools to have JSON
output." [0]
File changes:
+share/man/man8/devlink-health.8.gz
nix path-info -S:
5.0.0 46661416
5.1.0 46677880
[0]: https://www.spinics.net/lists/netdev/msg569915.html
I switched to fetchFromGitLab for more transparency and because 2.10 is
not yet available on any Debian mirrors.
A drawback is that any dates in the man pages will now always be
set to 1970-01-01, because fetchFromGitLab removes the timestamps from
$src (not manually but by unpacking the archive).
The two directories KDB and PTree do not exist before the SKS DB is
build for the first time. If /var/db/sks is empty and the module is
enabled via "services.sks.enable = true;" the following error will
occur:
...-unit-script-sks-db-pre-start[xxx]:
ln: failed to create symbolic link 'KDB/DB_CONFIG': No such file or directory
To avoid this both links have to be created after the DB is build.
Note: Creating the directories manually might be better but the initial
build might be skipped as a result:
unit-script-sks-db-pre-start[xxxxx]: KeyDB directory already exists. Exiting.
unit-script-sks-db-pre-start[xxxxx]: PTree directory already exists. Exiting.
This change was only a temporary workaround and isn't required anymore,
since /etc/systemd/system/system.slice should not be present on any
recent NixOS system (which makes this change a no-op).
This reverts commit 7098b0fcdf.
This change will load all configuration files from /etc, to make it easy
to override them, but fallback to /nix/store/.../etc/sway/config to make
Sway work out-of-the-box with the default configuration on non NixOS
systems.
Unfortunately the changes in ab5dcc7068
introduced a typo (took me a while to spot that...) that broke the
whole module (or at least the sks-db systemd unit).
The systemd unit was failing with the following error message:
...-unit-script-sks-db-pre-start[xxx]: KDB/DB_CONFIG exists but is not a symlink.
Changelog:
https://cryptography.io/en/latest/changelog/#v2-6-1
Important changes:
- BACKWARDS INCOMPATIBLE: Removed
cryptography.hazmat.primitives.asymmetric.utils.encode_rfc6979_signature
and
cryptography.hazmat.primitives.asymmetric.utils.decode_rfc6979_signature,
which had been deprecated for nearly 4 years. Use
encode_dss_signature() and decode_dss_signature() instead.
- BACKWARDS INCOMPATIBLE: Removed cryptography.x509.Certificate.serial,
which had been deprecated for nearly 3 years. Use serial_number
instead.
This should make the management easier. The package cryptography_vectors
contains the test vectors for cryptography and should therefore always
have the same version. By linking the version of cryptography_vectors to
cryptography, this simply cannot be forgotten.
Michael Weiss