Re: Information regarding upcoming Gitlab Migration: clarifications

ult for new users (who use either of the 'quick config' schemes like kdesrc-build-setup or kdesrc-build --initial-setup), but to leave the built-in default unchanged. So in essence, existing kdesrc-build users (who had a folder-based layout by default unless they went out of their way to find the right option) saw no change, but new users would have that option pre-set for them in the config. Regards, - Michael Pyne

Re: Notice of upcoming changes to the behaviour of the anongit network

as explained at https://community.kde.org/Sysadmin/GitKdeOrgManual#Let_Git_rewrite_URL_prefixes). kdesrc-build configures git to understand this "kde:" prefix by default. Regards, - Michael Pyne

Re: Update on Status of Gitlab Migration

1 https://gitlab.com/ 344 kde: All of the kde: repositories use the kde:foo syntax, where the 'foo' comes from the 'repopath' parameter of the sysadmin/repo-metadata YAML files. We may need to do on-the-fly conversion of the kde: repo paths if they won't be expressible as 'kde:foo' in the future, but we should have the information needed to do this in kdesrc-build to make this happen on-the-fly. Regards, - Michael Pyne

D26839: demote plugin load errors from warning to debug level + reword

mpyne accepted this revision. mpyne added a comment. Agreed, let's commit. REPOSITORY R244 KCoreAddons BRANCH master REVISION DETAIL https://phabricator.kde.org/D26839 To: sitter, mpyne, ngraham Cc: ngraham, kde-frameworks-devel, LeGast00n, GB_2, michaelh, bruns

D26405: Add mime type for backtraces saved from DrKonqi

mpyne accepted this revision. mpyne added a comment. This revision is now accepted and ready to land. Thanks Harald. Patch looks good and I've checked that the file name pattern is what Dr. Konqi uses by default. One question though, @broulik you mentioned there's also text/x-apport, should t

D24046: Allow triggering sort from QML

mpyne added inline comments. INLINE COMMENTS > broulik wrote in personssortfilterproxymodel.h:53 > "You can ... reimplement virtual functions defined in the primary base class > hierarchy (that is, virtuals defined in the first non-virtual base class, or > in that class's first non-virtual base

D23205: [KProcessList] Optimize KProcessList::processInfo

mpyne accepted this revision. mpyne added a comment. This revision is now accepted and ready to land. I've reviewed the patch and it's good, you've addressed the issues that were noted by Aleix already, so please commit and we can address anything that might pop up as it happens. REPOSITORY

D23800: Use ECMGenerateExportHeader to manage deprecated API better

mpyne accepted this revision. mpyne added a comment. This revision is now accepted and ready to land. The change looks good and builds in various combinations of CMake options (after I installed the changeset from D23789 ). I encountered that excluding depr

D20092: New class KOSRelease - a parser for os-release files

mpyne accepted this revision. mpyne added a comment. +1 to @aacid's comment about relicensing. The code is good and easy to review. Only comment I had was to think about using a warning or error when unable to parse the file. I think that the `ID_LIKE` handling could be done by using

D20008: Fix uninitialized value in SimpleJobPrivate

This revision was automatically updated to reflect the committed changes. Closed by commit R241:920d2c4ca7d5: kjobs: Fix compiler warning for uninit value in SimpleJobPrivate. (authored by mpyne). REPOSITORY R241 KIO CHANGES SINCE LAST UPDATE https://phabricator.kde.org/D20008?vs=54642&id=54

D20008: Fix uninitialized value in SimpleJobPrivate

mpyne created this revision. mpyne added reviewers: dfaure, Frameworks, apol. Herald added a project: Frameworks. Herald added a subscriber: kde-frameworks-devel. mpyne requested review of this revision. REVISION SUMMARY As reported in bug 404474 , GC

D18137: delete copy constructor and assignment operator of some public classes

mpyne accepted this revision. mpyne added a comment. This revision is now accepted and ready to land. The KSharedDataCache change is definitely correct. I've not done much with the kdelibs migrator but I agree with the logic and the implementation. Have you tried building some of the other fr

D17647: Fix KGlobalAccel build with Qt 5.13 prerelease

This revision was automatically updated to reflect the committed changes. Closed by commit R268:c96d012b6d47: Fix KGlobalAccel build with Qt 5.13 prerelease. (authored by mpyne). REPOSITORY R268 KGlobalAccel CHANGES SINCE LAST UPDATE https://phabricator.kde.org/D17647?vs=47743&id=47775 REVI

D17647: Fix KGlobalAccel build with Qt 5.13 prerelease

mpyne created this revision. mpyne added reviewers: Frameworks, graesslin. Herald added a project: Frameworks. Herald added a subscriber: kde-frameworks-devel. mpyne requested review of this revision. REVISION SUMMARY After a recent system upgrade to Qt git `dev` branch (for upcoming 5.13), kgl

D17446: cmake macros: Port away from deprecated ECM var in kcoreaddons_add_plugin

This revision was automatically updated to reflect the committed changes. Closed by commit R244:1285c4bf5552: cmake macros: Port away from deprecated ECM var in kcoreaddons_add_plugin (authored by mpyne). REPOSITORY R244 KCoreAddons CHANGES SINCE LAST UPDATE https://phabricator.kde.org/D1744

D17446: cmake macros: Port away from deprecated ECM var in kcoreaddons_add_plugin

mpyne created this revision. mpyne added reviewers: Frameworks, elvisangelaccio, apol, dfaure. Herald added a project: Frameworks. Herald added a subscriber: kde-frameworks-devel. mpyne requested review of this revision. REVISION SUMMARY As reported in bug 401888

D17078: Make it possible to use KAboutData/License/Person from QML

mpyne added a comment. In D17078#363886 , @apol wrote: > In D17078#363772 , @anthonyfieroni wrote: > > > You can't change parameters in methods or constructors, it's a BIC. > > > https://commun

Re: Regression in Frameworks - DBus Hangs

ktop they are being run within. The only reference of any sort that I see in either KF5 or kde/workspace for "quitOnLastWindowClosed" is in kxmlgui (for KMainWindow). There aren't many recent commits so anyone who can reproduce the hang should have a quick job verifying if reverting some of those would help or not. [1] Using git -P log --since="8 weeks ago" -G quitOnLastWindow --oneline for each git repository Regards, - Michael Pyne

Re: Password field security and information leaking

ication piece only. But even there, if the application is able to use a secure password entry widget, it could perhaps be possible to be able to safely set or update authentication credentials in the same app config U/I (i.e. have the app itself serve the pinentry-qt role). Regards, - Michael Pyne

D16189: kio_help: Fix crash in QCoreApplication when accessing help://

This revision was automatically updated to reflect the committed changes. Closed by commit R241:cf3b0d102855: kio_help: Fix crash in QCoreApplication when accessing help://. (authored by mpyne). REPOSITORY R241 KIO CHANGES SINCE LAST UPDATE https://phabricator.kde.org/D16189?vs=43569&id=4359

Re: KDE4 build question: turning off hidden visibility?

MAKE_CXX_FLAGS, though try to do this in a way that maintains the existing KDE settings. Regards, - Michael Pyne

D16189: kio_help: Fix crash in QCoreApplication when accessing help://

mpyne created this revision. mpyne added reviewers: Frameworks, sitter. Herald added a project: Frameworks. Herald added a subscriber: kde-frameworks-devel. mpyne requested review of this revision. REVISION SUMMARY A recent commit d428fc8e6447ede81f1e1911d0b66b39265672f3

D8256: Add _XOPEN_SOURCE to C definitions

mpyne added a comment. Sorry, if I'd been able to check this earlier I'd have warned against pushing, this was the kind of concern I was addressing with my earlier comment about FreeBSD. I had run into a lot of those issues when trying to get KF5 to build on Alpine. Adding an _XOPEN_SOU

D15076: Build failures with KSyntaxHighlighting 5.49

mpyne added a comment. The ECM documentation points to a specific change that may be related: https://api.kde.org/ecm/kde-module/KDECMakeSettings.html#build-settings Using the `KDE_SKIP_BUILD_SETTINGS` option that is described just leads to different build failures for me, however. I thi

D15076: Build failures with KSyntaxHighlighting 5.49

mpyne added a comment. I don't know the cause myself but the ECM version works up until 5.38.0 in my own testing. So presumably the change in behavior is something introduced in that release of ECM? REPOSITORY R55 Cantor REVISION DETAIL https://phabricator.kde.org/D15076 To: asemke, #k

D13627: [KSharedDataCache] Assume lock before flush changes

mpyne added a comment. I'd say to leave the code alone if this the change isn't needed to fix a bug. REPOSITORY R244 KCoreAddons REVISION DETAIL https://phabricator.kde.org/D13627 To: anthonyfieroni, davidedmundson, dfaure, #frameworks, hein, mpyne Cc: mpyne, hein, kde-frameworks-devel,

D14908: Bindings: Add support for scoped enums

mpyne accepted this revision. mpyne added a comment. This revision is now accepted and ready to land. LGTM and there's been plenty of time for a NAK. I say let's commit. REPOSITORY R240 Extra CMake Modules BRANCH scoped_enum REVISION DETAIL https://phabricator.kde.org/D14908 To: bruns

D14984: Explicitly request Qt 5.7's QtQuick to make use of Connections.enabled

This revision was automatically updated to reflect the committed changes. Closed by commit R169:b23e642e5ac4: Explicitly request Qt 5.7's QtQuick to make use of Connections.enabled (authored by mpyne). REPOSITORY R169 Kirigami CHANGES SINCE LAST UPDATE https://phabricator.kde.org/D14984?vs=4

D14908: Bindings: Add support for scoped enums

mpyne added a comment. As long as older versions of LLVM result in the Python binding generation step being skipped completely (we already check for libclang in the find-package for this, right?), I don't think the dependency increase is a problem. REPOSITORY R240 Extra CMake Modules REV

D14984: Explicitly request Qt 5.7's QtQuick to make use of Connections.enabled

mpyne created this revision. mpyne added reviewers: Frameworks, Plasma, mart. Herald added a project: Kirigami. Herald added a subscriber: plasma-devel. mpyne requested review of this revision. REVISION SUMMARY I recently rebuilt my Qt, KF5, and Plasma 5 environments from scratch. I noticed tha

D13627: [KSharedDataCache] Assume lock before flush changes

mpyne accepted this revision. mpyne added a comment. This change has worked fine for me in my testing (including KDE games and Plasma themes) so I think it's fine to commit. REPOSITORY R244 KCoreAddons REVISION DETAIL https://phabricator.kde.org/D13627 To: anthonyfieroni, davidedmundson

D14908: Bindings: Add support for scoped enums

mpyne added a comment. LGTM but then I don't use the Python bindings. Have you already tried the kcoreaddons build to see if this makes it work where it had failed before? REPOSITORY R240 Extra CMake Modules REVISION DETAIL https://phabricator.kde.org/D14908 To: bruns, #frameworks Cc: m

D13627: [KSharedDataCache] Assume lock before flush changes

mpyne added a comment. The patch looks OK but I'm a bit worried about trying to lock the cache at the same time we're potentially trying to delete a KSharedDataCache so I'm going to recompile and do some stress and sanity tests first. REPOSITORY R244 KCoreAddons REVISION DETAIL https://

D12233: Avoid manipulation of lists with quadratic complexity

mpyne accepted this revision. mpyne added a comment. I don't use Baloo directly but I've taken a look from a code sanity perspective and agree with @dhaumann that the change is a proper port of the existing logic. REPOSITORY R293 Baloo REVISION DETAIL https://phabricator.kde.org/D12233

D13627: [KIconThemes] Isolate private data from race conditions

mpyne added a comment. In D13627#287837 , @anthonyfieroni wrote: > https://phabricator.kde.org/source/kiconthemes/browse/master/src/kiconloader.cpp$603 > So the problem maybe not here as IconLoader loads icons but in icon writer > @davided

D11583: Add platform detection and adjustment to kded

This revision was automatically updated to reflect the committed changes. Closed by commit R297:88eda28b3567: Add platform detection and adjustment to kded (authored by graesslin, committed by mpyne). REPOSITORY R297 KDED CHANGES SINCE LAST UPDATE https://phabricator.kde.org/D11583?vs=30243&

D11290: [UDevManager] Already filter for subsystem before querying

mpyne added a comment. I never had the bug myself, I just grepped through the attached log to see if it made sense or not. :) I'll close the bug for now, if it doesn't fix the issue then it'll just get reopened. REPOSITORY R245 Solid REVISION DETAIL https://phabricator.kde.org/D11290

D11290: [UDevManager] Already filter for subsystem before querying

mpyne added a comment. Any reason not to close the bug 391738? I can do so but wanted to make sure it wasn't intentionally left open. REPOSITORY R245 Solid REVISION DETAIL https://phabricator.kde.org/D11290 To: broulik, #frameworks, davidedmundson Cc: mpyne, ngraham, michaelh

D10989: Check for nullptr in indexForNode

mpyne added a comment. I'm not sure the job could have been deleted before `KDialogJobUiDelegate::showErrorMessage()` since much of the call stack before then involves making method calls on the job. KDirLister::handleError(KIO::Job*) (kdirlister.cpp:81) makes a call to `job->error()` righ

D10168: Use nullptr for passing a null pointer to crc32

mpyne accepted this revision. mpyne added a comment. This revision is now accepted and ready to land. Actually `-Wzero-as-null-pointer-constant` is probably a good idea, and this is a correct fix. A decade ago when everything was going from 32 to 64-bits, we ran into a JuK crash bug because

D10450: Generate a custom target in kcoreaddons_desktop_to_json

mpyne added a comment. OK then, I think @kossebau is right in that this is a dependency issue in the `lookandfeel` part of plasma-desktop. The `kcm_lookandfeel` target declares the JSON dependency (with the CMake macro) in time for CMake to care about it and ensure the generated build syste

D10450: Generate a custom target in kcoreaddons_desktop_to_json

mpyne added a comment. In D10450#208453 , @kossebau wrote: > So just to make sure we are all on the same page: for what I have understood meanwhile is what is missing but needed here is a dependency rule between > a) the generated JSON file (

D10450: Generate a custom target in kcoreaddons_desktop_to_json

mpyne added a comment. What about this? I can't change the diff since I didn't create the RR, but this seems to cause the required dependency rules to be added and works for me to build plasma-desktop. The only real addition is the `add_dependencies` call. I tried this without the `add_cu

D10450: Generate a custom target in kcoreaddons_desktop_to_json

mpyne added a comment. Yes, I think I agree with @rakuco. Especially since the fix for D10485 ended up being reverted. I still think a separate fix is needed for kcm_lookandfeel, but the issue is that the `kcoreaddons_desktop_to_json` macro generates a

D10450: Generate a custom target in kcoreaddons_desktop_to_json

mpyne added a comment. So what's the conclusion here? Is this only a bug in kcm_lookandfeel or do we think that some follow-on changes are still required in `kcoreaddons_desktop_to_json`? REPOSITORY R244 KCoreAddons REVISION DETAIL https://phabricator.kde.org/D10450 To: tcberner, #fre

D9823: Optimize: use QMetaObject::invokeMethod with functor

mpyne accepted this revision. mpyne added a comment. LGTM REPOSITORY R244 KCoreAddons REVISION DETAIL https://phabricator.kde.org/D9823 To: mwolff, dfaure, apol, mpyne Cc: mpyne, #frameworks

D9334: Remove cmake 3.10+ warning for users of K_PLUGIN_FACTORY and K_PLUGIN_FACTORY_WITH_JSON

mpyne added a comment. In https://phabricator.kde.org/D9334#180881, @kossebau wrote: > In https://phabricator.kde.org/D9334#180830, @mpyne wrote: > > > In fact this appears to force files containing `K_PLUGIN_FACTORY*` into being evaluated by CMake's AUTOMOC (other warning fixes remov

D9334: Remove cmake 3.10+ warning for users of K_PLUGIN_FACTORY and K_PLUGIN_FACTORY_WITH_JSON

mpyne added a comment. In fact this appears to force files containing `K_PLUGIN_FACTORY*` into being evaluated by CMake's AUTOMOC (other warning fixes removed files from consideration by CMake AUTOMOC). CMake itself appears to have had an interface change for AUTOMOC between 3.8 and 3.9

D9277: Remove cmake warning about generating moc file

mpyne added a comment. @aacid you might be thinking of my KI18n fix for ui*.h files in https://phabricator.kde.org/D9118, but I'm sure there's other examples. I'll point out that Kevin looked at my patch as well and recommended something like `set_source_files_properties(${_header} ...)`

D9128: [knewstuff] Respect global variable

mpyne added a comment. OK, if no one else has objections I'd say to commit. REPOSITORY R304 KNewStuff REVISION DETAIL https://phabricator.kde.org/D9128 To: anthonyfieroni, leinir, dfaure, mpyne Cc: mpyne, #frameworks

D9128: [knewstuff] Respect global variable

mpyne accepted this revision. mpyne added a comment. This revision is now accepted and ready to land. I think the change looks good. But if the cache here is holding a *weak* pointer to a `Cache` then every user of this s_cache should be checking for null pointers each time, right? We can d

D9118: ki18n cmake macros: Mark UIC-generated .h files to skip AUTOMOC by default

This revision was automatically updated to reflect the committed changes. Closed by commit R249:6e3b70843566: cmake: Mark UIC-generated .h files to skip AUTOMOC by default. (authored by mpyne). CHANGED PRIOR TO COMMIT https://phabricator.kde.org/D9118?vs=23312&id=23478#toc REPOSITORY R249 KI

D9120: Fix a memory leak

mpyne added a comment. I think `~KuitStaticData()` *does* delete domainSetups. But deleting that QHash doesn't delete the underlying values. That means the real leak is at "Kuit::setupForDomain(QByteArray const&) (kuitmarkup.cpp:504)". If that's the case, then adding a destructor to th

Re: kcrash, fork, and stdout/stderr

le) to have kdeinit handle the restart if it was responsible for the initial launch. Regards, - Michael Pyne

D9118: ki18n cmake macros: Mark UIC-generated .h files to skip AUTOMOC by default

mpyne added a comment. In https://phabricator.kde.org/D9118#174892, @aacid wrote: > Will this cause trouble with older cmake versions? Good question, I haven't tried it. I've verified that `set_property` at least is present in 2.8.12

D9118: ki18n cmake macros: Mark UIC-generated .h files to skip AUTOMOC by default

mpyne created this revision. mpyne added reviewers: Frameworks, Build System, Localization. Restricted Application added a project: Frameworks. REVISION SUMMARY CMake in 3.10.0 has started warning that `.h` files which are automatically generated have historically been left out of AUTOMOC, and

D8256: Add _XOPEN_SOURCE to C definitions

mpyne added a comment. As I mentioned on the KDE bug, there is a possibility that changes like this would break FreeBSD. FreeBSD operates by a model of exporting symbols by default for legacy/BSD-specific needs. However if a feature flag is set, it limits its exports to *only* that require

D8992: Warn about errors when parsing json files

mpyne accepted this revision. This revision is now accepted and ready to land. REPOSITORY R244 KCoreAddons BRANCH master REVISION DETAIL https://phabricator.kde.org/D8992 To: apol, #frameworks, mpyne

D8871: Don't look for /etc/kderc every single time

mpyne accepted this revision. mpyne added a comment. This revision is now accepted and ready to land. +1 REPOSITORY R237 KConfig BRANCH master REVISION DETAIL https://phabricator.kde.org/D8871 To: apol, #frameworks, mpyne Cc: mpyne

Re: Adding build info attribute to KAboutData to improve bug report data?

ame is appropriate, like "setBugReportDetail" (since this is intended to support improved bug reports) or "setSourceRevision" (which describes what we're actually capturing). Just my thoughts, happy to support it going in however makes the mose sense. Regards, - Michael Pyne

D8207: Use https for KDE urls

mpyne added a comment. I support the idea, but I think there's a few more spots where we can convert http:// to https://. I ran `ag --only-matching --no-heading 'http:.*k.*\.org'` in the kcoreaddons source directory, with the current version of this patch already applied, and got the follow

D8002: Improve DesktopFileParser heuristic for finding old-style service desktop entries

This revision was automatically updated to reflect the committed changes. Closed by commit R244:4bdc7bb199ec: desktoptojson: Improve legacy service type detection heuristic. (authored by mpyne). CHANGED PRIOR TO COMMIT https://phabricator.kde.org/D8002?vs=19963&id=20361#toc REPOSITORY R244 K

D8002: Improve DesktopFileParser heuristic for finding old-style service desktop entries

mpyne added inline comments. INLINE COMMENTS > graesslin wrote in desktopfileparser.cpp:547 > Be careful with for and qt containers. This causes a detach. Either use the > old variant or wrap the list in qAsConst ( which is new in 5.7) OK. I think we settled on Qt 5.6 for now at the request of

D8002: Improve DesktopFileParser heuristic for finding old-style service desktop entries

mpyne added a comment. Any objections? If not I'll try to commit in the next few days. REPOSITORY R244 KCoreAddons REVISION DETAIL https://phabricator.kde.org/D8002 To: mpyne, #frameworks

D8002: Improve DesktopFileParser heuristic for finding old-style service desktop entries

mpyne created this revision. mpyne added a reviewer: Frameworks. Restricted Application added a project: Frameworks. REVISION SUMMARY When we adopted JSON plugin metadata for services, we also included compatibility code so that we could continue to load the older Desktop Entry format for KDE

D7281: KCoreAddons: Enforce hidden symbol visibility in nested private classes

mpyne added a comment. In https://phabricator.kde.org/D7281#135053, @kossebau wrote: > @mpyne: Beware, I have such patches prepared for all of KF :) OK, I'll hold off on this one pending your KF5-wide patchset ;) REPOSITORY R244 KCoreAddons REVISION DETAIL https://phabricator

D7281: KCoreAddons: Enforce hidden symbol visibility in nested private classes

mpyne created this revision. Restricted Application added a project: Frameworks. REVISION SUMMARY Inspired by https://phabricator.kde.org/D7279, manually declare nested Private classes to have hidden visibility (by default, they inherit the visibility of the class that they are nested under).

D7170: Fix errorneous whitespace handling in Desktop Entry parsing from DesktopFileParser

This revision was automatically updated to reflect the committed changes. Closed by commit R244:077ef6797eb6: desktopfileparser: Fix non-compliant key/value parsing. (authored by mpyne). REPOSITORY R244 KCoreAddons CHANGES SINCE LAST UPDATE https://phabricator.kde.org/D7170?vs=17913&id=18069

D7170: Fix errorneous whitespace handling in Desktop Entry parsing from DesktopFileParser

mpyne updated this revision to Diff 17913. mpyne added a comment. Joins the regex for the Type= line with its use. REPOSITORY R244 KCoreAddons CHANGES SINCE LAST UPDATE https://phabricator.kde.org/D7170?vs=17792&id=17913 REVISION DETAIL https://phabricator.kde.org/D7170 AFFECTED FILES

D7170: Fix errorneous whitespace handling in Desktop Entry parsing from DesktopFileParser

mpyne marked an inline comment as done. REPOSITORY R244 KCoreAddons REVISION DETAIL https://phabricator.kde.org/D7170 To: mpyne, #frameworks, arichardson Cc: apol

D7169: Fix errorneous whitespace handling in Desktop Entry parsing from KConfigIni

This revision was automatically updated to reflect the committed changes. Closed by commit R237:11ee92d92b1e: kconfigini: Strip leading whitespace when reading entry values. (authored by mpyne). REPOSITORY R237 KConfig CHANGES SINCE LAST UPDATE https://phabricator.kde.org/D7169?vs=17791&id=1

D7170: Fix errorneous whitespace handling in Desktop Entry parsing from DesktopFileParser

mpyne created this revision. Restricted Application added a project: Frameworks. REVISION SUMMARY As reported in bug 310674, the Desktop Entry spec says that whitespace surrounding the = sign that separates the key and value in .desktop files should be ignored when interpreting the key and val

D7169: Fix errorneous whitespace handling in Desktop Entry parsing from KConfigIni

mpyne created this revision. Restricted Application added a project: Frameworks. REVISION SUMMARY As reported in bug 310674, the Desktop Entry spec says that whitespace surrounding the `=` sign that separates the key and value in .desktop files should be ignored when interpreting the key and v

D6597: Add magic define to enable setgroups(2)

This revision was automatically updated to reflect the committed changes. Closed by commit R241:15451d505fdb: Fix build on musl libc by exposing setgroups(2). (authored by mpyne). REPOSITORY R241 KIO CHANGES SINCE LAST UPDATE https://phabricator.kde.org/D6597?vs=16422&id=17270 REVISION DETA

D6598: Declare _POSIX_SOURCE to ensure kill(2) is declared

This revision was automatically updated to reflect the committed changes. Closed by commit R303:c42f18a7fc1a: Use proper feature flag to include kill(2). (authored by mpyne). CHANGED PRIOR TO COMMIT https://phabricator.kde.org/D6598?vs=16423&id=17269#toc REPOSITORY R303 KInit CHANGES SINCE

D6596: Fix dev_t declaration for Solid UDisks2 backend

This revision was automatically updated to reflect the committed changes. Closed by commit R245:2382c3f8d366: Require sys/types.h for dev_t unconditionally. (authored by mpyne). REPOSITORY R245 Solid CHANGES SINCE LAST UPDATE https://phabricator.kde.org/D6596?vs=16420&id=17066 REVISION DETA

D6597: Add magic define to enable setgroups(2)

mpyne added a comment. With the help of the KDE sysadmins, I've tested this to build on FreeBSD. BSD-specific defines remain available on FreeBSD unless you start to use `_POSIX_C_SOURCE`, `_ANSI_SOURCE`, or `_XOPEN_SOURCE`. So `_BSD_SOURCE` and `_DEFAULT_SOURCE` appear to be safe for Free

D6596: Fix dev_t declaration for Solid UDisks2 backend

mpyne added a comment. If there are objections to this please let me know in next few days. A similar Differential request to this has already been opened and abandoned. I also have similar fixes/changes in https://phabricator.kde.org/D6597 and https://phabricator.kde.org/D6598 for whic

D6789: API dox: add note about moc include need with K_PLUGIN_FACTORY(_WITH_JSON)

mpyne accepted this revision. This revision is now accepted and ready to land. REPOSITORY R244 KCoreAddons BRANCH addnoteaboutmocincludewithpluginfactory REVISION DETAIL https://phabricator.kde.org/D6789 To: kossebau, #frameworks, mpyne

D6672: add KAboutLicense::spdx and introduce orLater qualification

mpyne accepted this revision. This revision is now accepted and ready to land. REPOSITORY R244 KCoreAddons BRANCH spdx REVISION DETAIL https://phabricator.kde.org/D6672 To: sitter, sebas, mpyne Cc: #frameworks

D6672: add KAboutLicense::spdx and introduce orLater qualification

mpyne added a comment. In https://phabricator.kde.org/D6672#126360, @sitter wrote: > Ah drat. We could still get rid of the second private ctor though, by moving the defaults from the private to the public and calling the full private ctor: Yes, that would be fine if you want to

D6672: add KAboutLicense::spdx and introduce orLater qualification

mpyne added a comment. I like delegating constructors, but we can't use them in Frameworks yet. :( The current compiler requirement policy is that we require a C++ compiler supported in the last 3 Qt minor releases (see https://community.kde.org/Frameworks/Policies#Frameworks_Qt_requirem

D6700: disable compiler warning on testing deprecated function

mpyne accepted this revision. This revision is now accepted and ready to land. REPOSITORY R244 KCoreAddons BRANCH no-testwarning REVISION DETAIL https://phabricator.kde.org/D6700 To: sitter, mpyne Cc: #frameworks

D6672: add KAboutLicense::spdx and introduce orLater qualification

mpyne requested changes to this revision. mpyne added a comment. This revision now requires changes to proceed. Still a couple of things: 1. The included new autotest fails for me, because of using `OrLaterVersion` as a default version restriction. `OnlyThisVersion` is used elsewhere as

D6672: add KAboutLicense::spdx and introduce orLater qualification

mpyne requested changes to this revision. mpyne added a comment. This revision now requires changes to proceed. AFAICT we do need to maintain BIC here even for private ctors. The inline comment has more detail. Other than that and with using a flag enum instead of a `bool` I like the patch

D6631: Fix compilation with strict libc (such as musl)

mpyne created this revision. Restricted Application added a project: Plasma. Restricted Application added a subscriber: plasma-devel. REVISION SUMMARY Many of the functions we use implicitly are not part of the C standard library so we have to specific that we're pulling from POSIX, X/Open,

D6598: Declare _POSIX_SOURCE to ensure kill(2) is declared

mpyne created this revision. Restricted Application added a project: Frameworks. REVISION SUMMARY Believe it or not, but the `kill(2)` system call is not specified to be available by default in C. It's part of BSD, POSIX, Single UNIX Spec, etc., but you do have to declare one of the various f

D6597: Add magic define to enable setgroups(2)

mpyne created this revision. Restricted Application added a project: Frameworks. REVISION SUMMARY The KIO KPAC DHCP helper uses `setgroups(2)`. But this is apparently not specified in POSIX due to being a "privileged call", so it is only available as an extension, typically needing `_BSD_SOUR

D6596: Fix dev_t declaration for Solid UDisks2 backend

mpyne created this revision. Restricted Application added a project: Frameworks. REVISION SUMMARY Compiling Solid on a musl-based Linux environment (long story...) gives an error compiling UDisks2 support because it can't find the `dev_t` type. Looking up `dev_t`, it appears that POSIX req

D6376: Fix double delete crash during shutdown

mpyne added a comment. In https://phabricator.kde.org/D6376#119359, @rjvbb wrote: > But what's the official stance on deleting objects (widgets) that have a parent and are thus capable of auto-cleanup? AFAIK one can still delete them explicitly, and if they're reparented during that proc

D5966: Fix race-condition in KRandom-seeding.

mpyne added a comment. The changeset is fine to go in (once the duplicated semicolon is fixed). As for C++11, I'm not aware that Qt requires C++11 as a blanket policy. Instead they require specific compilers, and then we can use the C++11 features supported by that compiler. Accord

D5966: Fix race-condition in KRandom-seeding.

mpyne added a comment. Would it be better to use `QThreadStorage` (to record if the thread has been seeded) instead of a `QSet`? It would allow us to more easily adopt C++11's `thread_local` once we remove support for older compilers. I also looked into whether we could adopt the higher-qu

D5439: API dox: more info about KAboutData's orgDomain/desktopFileName properties

mpyne accepted this revision. mpyne added a comment. I think the API dox fixes are fine to go in. We can add i18n-related warnings in a separate patch as you suggest. REPOSITORY R244 KCoreAddons REVISION DETAIL https://phabricator.kde.org/D5439 To: kossebau, #frameworks, aacid, mpyne,

D5439: API dox: more info about KAboutData's orgDomain/desktopFileName properties

mpyne added a comment. Is there any reason why we cannot just do the needed `setlocale(LC_ALL, "")` within ki18n itself (e.g. in the `KLocalizedStringPrivateStatics` ctor which is used to support translation of `i18n` calls? `QCoreApplication` is going to do this anyways so making it happen

D5430: KDECompilerSettings: Pass -Wvla & -Wdate-time

mpyne accepted this revision. mpyne added a comment. This revision is now accepted and ready to land. I think these warning flags make sense as default warning flags. I don't agree that `__DATE__` and `__TIME__` should be avoided as a rule, but projects that find this valuable can disable th

D5392: Fix build with Qt 5.9

mpyne added a comment. I can confirm that the patch works with KHTML built on Qt 5.8.1, and that Konqueror still finds KHTML fine after. REPOSITORY R313 KHtml REVISION DETAIL https://phabricator.kde.org/D5392 To: apol, #frameworks Cc: mpyne, ltoscano, #frameworks

D5104: Error page, correctly load the image (with a real URL)

mpyne accepted this revision. mpyne added a comment. This revision is now accepted and ready to land. In https://phabricator.kde.org/D5104#97246, @ltoscano wrote: > Thanks for checking: with this patch you should not see the error that I reported in the summary/commit message. Is it the c

D5104: Error page, correctly load the image (with a real URL)

mpyne added a comment. The patch looks fine to me, and gives a file:// that successfully opens in a separate Konq tab... but the image still doesn't show in the actual error page. I'm not sure of the cause though and have no time to research further tonight. :( REPOSITORY R313 KHtml REV

D5104: Error page, correctly load the image (with a real URL)

mpyne added a comment. Will try to look later today but won't be able to at this point. REPOSITORY R313 KHtml REVISION DETAIL https://phabricator.kde.org/D5104 To: ltoscano Cc: mpyne, #frameworks

D5079: debug when disabling kcrash due to env vars

mpyne accepted this revision. mpyne added a comment. This revision is now accepted and ready to land. LGTM REPOSITORY R285 KCrash BRANCH master REVISION DETAIL https://phabricator.kde.org/D5079 To: sitter, dfaure, mpyne Cc: mpyne, #frameworks

  1   2   3   4   >