When opening the options page, the first entry in the NDK list widget got automatically selected. The version of that NDK entry was displayed in the status summary, and the status summary got updated when the
selection changed.
That gave the impression that the selection of an NDK entry has an
influence on the configuration of kits. That is not the case, though.
The NDK list widget is there to explore and manage available NDKs, and
the kit generation mechanism has its own criteria to determine which of
those NDKs to use for which kit.
This change removes the automatic selection of the first NDK entry, as-
well as the displaying of the NDK version in the status summary. The
translations of the status summary string are also adapted.
Fixes: QTCREATORBUG-31900
Change-Id: Ia5e8d8401bf4e3e76093dc7341198cc8c8678214
Reviewed-by: Jarek Kobus <jaroslaw.kobus@qt.io>
Primitive color palette has updated color values and some changed color
names. Light and dark tokens have updated mappings to primitive colors.
Change-Id: I764a57241bf310e0e0225eb9c750b0bef9adc31b
Reviewed-by: Cristian Adam <cristian.adam@qt.io>
We were having issues with the download of the Qt Creator packages
failing. The new version of the install-dev-package script has more
error output that might help with tracking down the issue as it
is intermittent.
Change-Id: I677bbf893215a056bea96feecef9bcd241cf85d6
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
Utilize the information from the downloaded Updates.xml more.
- Separate the package sections for "parsing", throw away "debug" ones
and limit to the right compiler spec
- When downloading a module, find the right package section, and
construct the URL via the Name and DownloadableArchives information
This way we no longer depend on hardcoded archive postfixes and hardcoded
archive URL structures.
Task-number: QTCREATORBUG-31184
Change-Id: I98912677f1618fe281cf2b8aaf989c5c6e0f3c51
Reviewed-by: Cristian Adam <cristian.adam@qt.io>
The layout of the installer components and the layout within the
archives changed.
Fixes: QTCREATORBUG-32090
Change-Id: I4218d426f06c022d1ee0f5eee6cd9fbf6cf881cb
Reviewed-by: Cristian Adam <cristian.adam@qt.io>
When determining the GCC version we did not use the same environment
that the tests actually use.
As a bonus: an attempt to narrow down UnboundLocalError that occurs
in the QXmlAttributes test in the Win 10 + MinGW + Qt5 test plan
and which I cannot reproduce locally.
Change-Id: Ib12b66091eed7b27871c0bed59d8de019ee2285b
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
Only relevant for QmlDesigner running in Qt Creator (not QDS), and for
non-QDS themes.
Fixes: QTCREATORBUG-31830
Change-Id: I075f5df41edb0ea7c4a76681c7ca1ec7c2d55389
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
The initial version of the pmr containers dumper test
had a couple of shortcomings:
* there was an inclusion of `<memory_resource>` missing
in the test program body
* some test setups use compilers that do not use C++17 by default
so the correct profile must be specified for the test
Disabled a bunch of checks that did not work with libcxx
due to incorrect reported size of
`std::pmr::polymorphic_allocator<std::pair<const int, int>>`.
Fixed the libcxx dumper for `std::map` (the ordinary, non-pmr, one).
As turned out, it was broken due to an incorrect padding calculation
for `std::pair<const int, int>`.
This has been worked around by using pair's inner types explicitly.
Amends: cb23332ae5
Task number: QTCREATORBUG-30224
Change-Id: I937295f88083ee9521a35ac7dbe22303ef52bd53
Reviewed-by: hjk <hjk@qt.io>
This used to be implicit in the following attach attempt but
apparently isn't anymore.
Change-Id: I0d8ff3db9c38fc24aed55f1f3b6514cd6315bd4c
Reviewed-by: Assam Boudjelthia <assam.boudjelthia@qt.io>
Added support for visualization of for `std::pmr` containers
(and more broadly containers with custom non-empty allocators).
The gotcha with pmr allocators in particular is that
unlike most allocators these are non-empty classes:
they contain a pointer to a memory resource.
Tested locally with CDB and GDB (MinGW).
Task number: QTCREATORBUG-30224
Change-Id: If0ceb8a46905151a98f27d3e3ff11386406a9737
Reviewed-by: hjk <hjk@qt.io>
Reviewed-by: David Schulz <david.schulz@qt.io>
Currently hint and information diagnostics are rendered as warnings. Add
and use a new color for those type of diagnostics.
Fixes: QTCREATORBUG-31878
Change-Id: I2fdd01213b32bd5dc433b925a1a44c0db480f422
Reviewed-by: Alessandro Portale <alessandro.portale@qt.io>
This introduces a field "DisplayName" to PluginSpec and the plugin spec
json files. If present in the spec, PluginSpec::displayName returns it,
or otherways falls back to more established fields.
Fixes: QTCREATORBUG-31761
Change-Id: I4f4c4f3e33b17dfb2ec63644b1f50b3b9c6c024a
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
In the Figma-based design system, the "2024" light and dark theme color
tokens reference colors from a primitives layer. Qt Creator was so far
missing this indirection.
This change adds the primitive layer as .ini file and replaces the #rgb
values in the themes with references to the new layer.
Fixes: DSGNSYSTM-9
Change-Id: I9ef358dee1f2e4c0e54f9f3ad97c6f3ae4618475
Reviewed-by: Marcus Tillmanns <marcus.tillmanns@qt.io>
Makes it easier to reason what 'False' and 'True' in the message means.
Change-Id: I0e8fc1c9d2db7837a8334dc1400fa4d3b54abf06
Reviewed-by: Jarek Kobus <jaroslaw.kobus@qt.io>
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
A path was considered relative (only) when it started with a slash,
not e.g. "C:/...". This led to display of the "./" "relative path
marker" in these cases. "remote" windows paths regularly occur
for baremetal executables on a Windows host.
Change-Id: I082ae1cfb473ce2009d830fa98759d65a45735b5
Reviewed-by: Marcus Tillmanns <marcus.tillmanns@qt.io>
Allows to embed a widget into a line that will do its best
to stay in position.
Change-Id: I72590814057eb92bd17978a63bb4f7132410212f
Reviewed-by: David Schulz <david.schulz@qt.io>
This change reorganizes the display names of the themes.
The new themes and related files are renamed.
Task-number: QTCREATORBUG-31747
Change-Id: I5b52e640b5590cc263a68b68c7a8538fc84c9926
Reviewed-by: hjk <hjk@qt.io>
Fixed dumping of the standard library iterator types
in case the binary is compiled with _GLIBCXX_DEBUG switch.
Some non-stdlib tests that also used to fail
after enabling _GLIBCXX_DEBUG due to a crash in
`QCborMap::fromVariantMap` were adjusted as well.
Full list of affected tests:
* StdMap - fix
* StdSet - fix
* QC41700 - indirect fix (this test uses map iterators internally)
* QCbor - workaround (QCborMap::fromVariantMap is simply avoided)
* QJson - workaround (same as above)
All in all, enabling _GLIBCXX_DEBUG in the test should not produce
more failures compared to running without it.
*Note: tested locally with mingw 13.10.*
Fixes: QTCREATORBUG-31876
Change-Id: Ie55a40dc9a1aadb20795ad43abd0597047c429e4
Reviewed-by: hjk <hjk@qt.io>
This replaces all occurrences of
value('Plugins').indexOf('MyPlugin') >= 0
with
isPluginRunning('myplugin')
And combines OR'ed sequences with AnyPluginRunning.
Tests for (Boot2Qt || Boot2QtQdb) are reduced to just boot2qt, since the
other one is history.
Change-Id: I21206116e2263579349c77ba3914f8ebfe1f089e
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
Fixes the dock widget splitter color in Debugger.
Uses more toke colors.
Change-Id: I38a52f48a049d36898eeffaaa32033c343ec2c05
Reviewed-by: Marcus Tillmanns <marcus.tillmanns@qt.io>
