net/libsoup3: import libsoup3-3.2.2

New major version of the library.
Changes compared to net/libsoup:

Changes in libsoup from 3.2.1 to 3.2.2:

* Various HTTP/2 Fixes: [Carlos Garcia Campos]
* Fix `content-sniffed` not being emitted for resources without content
* Fix leak of SoupServerConnection when stolen

Changes in libsoup from 3.2.0 to 3.2.1:

* When built against nghttp2 1.50.0+ be relaxed about header whitespace [Carlos Garcia Campos]
* Fix possible crash when cancelling an HTTP/2 message [Carlos Garcia Campos]
* Fix regresion where soup_server_message_get_socket() could return NULL [Carlos Garcia Campos]
* Fix minor memory leak [Milan Crha]

Changes in libsoup from 3.1.4 to 3.2.0:

Changes in libsoup from 3.1.3 to 3.1.4:

* Numerous improvements to HTTP/2 reliablity [Carlos Garcia Campos]
* Fix `http` proxy authentication with default proxy resolver [Carlos Garcia Campos]
* Fix undefined ``ssize_t`` with MSVC [Patrick Griffis]

Changes in libsoup from 3.1.2 to 3.1.3:

* Fix compile error when `SOUP_VERSION_MAX_ALLOWED` is defined [Patrick Griffis]

Changes in libsoup from 3.1.1 to 3.1.2:

* Replace HTTP/2 tests using Quart with internal HTTP/2 server tests [Carlos Garcia Campos]
* Improve version macros including adding ability to define `SOUP_DISABLE_DEPRECATION_WARNINGS` [Emmanuele Bassi]

Changes in libsoup from 3.0.7 to 3.1.1:

* Reintroduce some thread-safety to SoupSession (see https://libsoup.org/libsoup-3.0/client-thread-safety.html) [Carlos Garcia Campos]
* Add SoupServerMessage:tls-peer-certificate and SoupServerMessage:tls-peer-certificate-errors [Ignacio Casal Quinteiro]
* Port docs to gi-docgen [Maximiliano Sandoval R]
* Update documentation [Patrick Griffis]

Changes in libsoup from 3.0.6 to 3.0.7:

* Fix leak in SoupAuthNTLM [Milan Crha]
* Fix constructing SoupAuthNTLM objects [Milan Crha]
* Disable mutual negotiation in SoupAuthNegotiate [Michael Catanzaro]
* http2: Do not advertise the `h2` protocool for proxy connections [Carlos Garcia Campos]
* http2: Remove left-over headers when HTTP/1 redirects to HTTP/2 [Carlos Garcia Campos]
* http2: Handle HTTP_1_1_REQUIRED error [Carlos Garcia Campos]
* http2: Read request bodies synchronously for sync requests [Carlos Garcia Campos]
* http2: Properly handle server sending shut down GOAWAY [Carlos Garcia Campos]
* tests: Remove dependency on Apache's PHP module [Carlos Garcia Campos]
* tests: Depend upon Apache's http2 module [Carlos Garcia Campos]

Changes in libsoup from 3.0.5 to 3.0.6:

* Misc HTTP/2 fixes [Carlos Garcia Campos]
* Add PUT/POST support to examples/get [Carlos Garcia Campos]
* Add `--user-agent` option to examples/get [Carlos Garcia Campos]
* Misc meson improvements [Nirbheek Chauhan]
* Fix build with Visual Studio [Chun-wei Fan]

Changes in libsoup from 3.0.4 to 3.0.5:

* Misc HTTP/2 fixes [Carlos Garcia Campos]
* Fix missing files for installed-tests [Simon McVittie]
* Fix SoupServer not properly handling invalid percent encoded paths [Ignacio Casal Quinteiro]
* Fix other areas not properly handling invalid percent encoded paths [Patrick Griffis]
* Fix SoupLogger:max-body-size of 0 meaning log nothing [Patrick Griffis]

Changes in libsoup from 3.0.3 to 3.0.4:

* Fix HTTP/2 not properly handling socket timeouts [Carlos Garcia Campos]
* Improvements to test reliablity [Simon McVittie]
* Fix cross-compiling to Windows [Melroy van den Berg]
* Fix tests with development glib-networking [Patrick Griffis]
* Expose soup_uri_copy() to Vala [Jens Georg]

Changes in libsoup from 3.0.2 to 3.0.3:

* Fix various HTTP/2 issues [Carlos Garcia Campos]

* Error when libsoup2 has been loaded before libsoup3 [Patrick Griffis]

* Fix memory leak when using TLS [Mart Raudsepp]

Changes in libsoup from 3.0.1 to 3.0.2:

* Add support for multiple auth challenges in one response [Patrick Griffis]

* Fix SoupCache test failures on 32bit [Patrick Griffis]

* Don't treat `-Wincompatible-pointer-types` as error
  The `glib-mkenums` tool sometimes triggered this [Patrick Griffis]

* Improve `gssapi` dependency handling [Nirbheek Chauhan]

* Fix undefined `ssize_t` on Windows [Chun-wei Fan]

* Updated translations: Hebrew

Changes in libsoup from 3.0.0 to 3.0.1:

* Move python overrides to upstream pygobject [Patrick Griffis]

* Fix minor build warnings [Patrick Griffis]

Changes in libsoup from 2.99.9 to 3.0.0:

* Remove unused dependency on libxml [Patrick Griffis]

* Use G_DECLARE_FINAL_TYPE when applicable and bump GLib dependency
  to 2.69.1 [Patrick Griffis]

* Add PyGObject overrides for SoupMessageHeaders to behave like a dict [Patrick Griffis]

* Fix soup_message_add_status_code_handler() in Vala [Rico Tzschichholz]

* Ensure that all header values are valid UTF-8, other encodings are not
  supported [Patrick Griffis]

* Support Content-Disposition headers missing a disposition-type [Patrick Griffis]

Changes in libsoup from 2.99.8 to 2.99.9:

* Make soup_message_set_method() public API [Carlos Garcia Campos]

* Make SoupMessage:http-version read only [Carlos Garcia Campos]

* Prevent the same message being queued multiple times [Carlos Garcia Campos]

* Allow completing SoupMessage::tls-interaction with NULL certificate [Carlos Garcia Campos]

* Replace soup_server_set_ssl_cert_file() with soup_server_set_tls_certificate() [Carlos Garcia Campos]

* Add SoupServer support for client certificates [Carlos Garcia Campos]

* Fallback to building meson wrap of sqlite if not found [Xavier Claessens]

* Add soup_message_get_tls_protocol_version() and soup_message_get_tls_ciphersuite_name() API [Carlos Garcia Campos]

Changes in libsoup from 2.99.7 to 2.99.8:

* Enable HTTP/2 by default. The SOUP_FORCE_HTTP1 env var can disable for debugging
  [Carlos Garcia Campos]

* Restrict advertising brotli decoding support to HTTPS [Patrick Griffis]

* Add new API to handle client certificate authentication per-message as
  well as API to PKCS #11 PINs [Carlos Garcia Campos, Patrick Griffis]

* Add new build features `http2_tests` and `pkcs11_tests` to more granularly control
  optional test dependencies [Patrick Griffis]

Changes in libsoup from 2.99.6 to 2.99.7:

* Change SoupSession to only support one SoupSessionFeature of a given type [Patrick Griffis]

* Remove soup_session_get_features() API [Patrick Griffis]

* Numerous HTTP/2 fixes and improvements [Carlos Garcia Campos]

Changes in libsoup from 2.99.5 to 2.99.6:

* Added HTTP/2 support. Enable by setting `SOUP_ENABLE_HTTP2` env var. [Patrick Griffis, Carlos Garcia Campos]

Changes in libsoup from 2.99.4 to 2.99.5:

* Add soup_message_get_remote_address() API [Carlos Garcia Campos]

* Fix preconnect stealing a connection [Carlos Garcia Campos]

* Fix potential header issues when included in a C++ project [Patrick Griffis]

* Disabling tests also disables Autobahn tests by default [Patrick Griffis]

Changes in libsoup from 2.99.3 to 2.99.4:

* Rename SoupMessage:tls-certificate and SoupMessage:tls-certificate-errors
  to tls-peer-certificate and tls-peer-certificate-errors respectively [Carlos Garcia Campos]

* Add SoupMessageMetrics API for tracking message events [Carlos Garcia Campos]

* Add soup_message_get_connection_id() API [Patrick Griffis]

* Add WebSocket fuzzing tests with Autobahn [Diego Pino]

* Windows build fixes [Chun-wei Fan]

Changes in libsoup from 2.99.2 to 2.99.3:

* Fix soup_session_get_async_result_message() [Patrick Griffis]

* Fix SoupMessage not always emitting notify single on property change [Carlos Garcia Campos]

* Fix soup_message_query_flags() not returning a boolean [Patrick Griffis]

* Revert requiring a GPollableInputStream for request bodies [Carlos Garcia Campos]

* Rename soup_message_body_free() and soup_message_headers_free() to
  soup_message_body_unref() and soup_message_headers_unref() [Carlos Garcia Campos]

* Rename SoupMessage:options-ping property to SoupMessage:is-options-ping and add
  soup_message_new_options_ping(), soup_message_get_is_options_ping(), and
  soup_message_set_is_options_ping() [Carlos Garcia Campos]

* Add SOUP_HTTP_2_0 SoupHTTPVersion to be used later [Patrick Griffis]

Changes in libsoup from 2.99.1 to 2.99.2:

* Add new soup_session_send_and_read() API [Carlos Garcia Campos]

* Add soup_session_get_async_result_message() API [Carlos Garcia Campos]

* Remove soup_session_load_uri_bytes() and soup_session_read_uri() APIs [Carlos Garcia Campos]

* Add soup_session_preconnect_async() API [Carlos Garcia Campos]

* Add SoupSession:remote-connectable property to support proxying over unix sockets [Robert Ancell]

* Restore support to SoupLogger for logging message bodies [Daniel Kolesa]

* Update GLib dependency to 2.67.4

Changes in libsoup from 2.72.0 to 2.99.1:

* This is a major ABI breaking release that can be installed in parallel with 2.7x.
  Numerous API have been changed which is documented on https://libsoup.org

(wiz)

scribus*: fix build on NetBSD-current

ld from binutils 2.39 is more picky about library linking.

(wiz)

doc: Updated editors/xournalpp to 1.1.3

(wiz)

xournalpp: update to 1.1.3.

## 1.1.3

* Fixed a crash that occurs when trying to add/edit/test LaTeX if LaTeX is not found (#4334, #4345)
* Fixed a crash that could occur when quitting when using a touchscreen device (#4343, #4373)
* Fixed a bug that closed text fields immediately after creating them with a secondary mouse or stylus button (#3983, #4309)
* Added a setting in the preferences to disable Gtk inertial scrolling, avoiding weird scrolling on some devices (#4013, #4420)
* Linux: added an Ubuntu 22.04 release build (#4411)

## 1.1.2

Bugfixes from various community contributors.

* Fixed application window being dragged unintentionally when dragging from
  toolbox or menu bars (#3901, #3907).
* Fixed CLI export failures caused by not using the C locale (#3939, #3944).
* Fixed a bug where the Default Tool does not update the tool icon (#3918,
  #3947).
* Fixed a bug where currently editing text boxes would not unfocus when moving
  to a different page (#3765, #4027).
* Fixed stylus/mouse inputs being detected twice for some tools (#4162).
* Fixed a crash that occurs when a plugin calls `app.getDocumentStructure()`
  (#4192, #4196).
* Fixed a memory leak that occurs when using the Stroke Recognizer (#4258).
* Fixed a crash that occurs on some Linux distros when opening a save dialog
  after using the text tool (#4279, #4280).
* Fixed a potential crash in toolbar customization (#4293).
* Fixed a crash that can occur when "Apply Background to All Pages" is undoed
  and then another action is performed (#4139, #4297).
* Fixed PDF Attach Mode settings not loading correctly (#3189, #4228).
* Improved performance of Text Tool rendering by eliminating unnecessary redraws
  (#3810, #4000, #4074, #4135, #4136).
* Improved LaTeX Tool error message when running inside Flatpak by providing
  steps on how to install the TeXLive extension (#3996, #4259).
* Windows: fixed a crash in the installer (#4041, #4062).
* macOS: changed the way the Xournal++ is launched to avoid some potential
  issues (#4087, #4090, #4287).
* Linux: fixed linker errors associated with backtrace library (#3817, #3819).

(wiz)

doc: Added devel/ruby-neovim version 0.9.0

(nikita)

Import devel/ruby-neovim version 0.9.0

Packaged in wip by Jonathan Buschmann.

Ruby bindings for neovim.

This library provides Ruby support in neovim through its msgpack-rpc API.

(nikita)

Updated textproc/py-xlsxwriter, time/py-dateparser

(adam)

py-dateparser: updated to 1.1.5

1.1.5 (2022-12-29)
------------------

Improvements:

- Parse short versions of day, month, and year
- Add a test for ���in 1d���
- Update languages_info
- Add a workaround for zipimporter not having exec_module before Python 3.10
- Stabilize tests at midnight
- Add a test case for French

Cleanups:

- Remove the requirements-build file

(adam)

py-xlsxwriter: updated to 3.0.4

Release 3.0.4 - December 28 2022
--------------------------------
* Roll up release of several minor fixes.
* Drop support for EOL Python 3.4 and 3.5.

(adam)

Updated devel/meson, x11/gtk3

(adam)

gtk3: updated to 3.24.36

Overview of Changes in GTK+ 3.24.36, 12-22-2022
===============================================

This release is the first GTK 3 release to exclusively
use the meson build system.

* GtkLabel:
- Tweak selection behavior

* GtkEmojiChooser:
- Properly handle empty recent section

* GtkFileChooser:
- Make ~ key work regardless of dead keys

* build:
- Improve handling of flaky tests
- Drop the autotools build

* Wayland:
- Fix problems with X<>Wayland DND
- Revert cursor changes from 3.24.35
- Fix handling of surrounding text in input

* Windows:
- Improve handling of line ends in clipboard

* MacOS:
- Use NSTrackingArea

* Translation updates:
Abkhazian
Basque
Brazilian Portuguese
Catalan
Interlingue
Persian

(adam)

meson: updated to 1.0.0

1.0.0
Compiler check functions prefix kwargs accepts arrays
Flags removed from cpp/objcpp warning level 1
Developer environment improvements
Deprecate java.generate_native_headers, rename to java.native_headers
rust.bindgen accepts a dependency argument
String arguments to the rust.bindgen include_directories argument
The Rust module is stable
in operator for strings
warning-level=everything option

(adam)

doc: Updated www/wiki-tui to 0.6.1

(pin)

www/wiki-tui: update to 0.6.1

Bug Fixes
- Fix error when no results are available by @Builditluc in #105
- Fix multiple views not displaying the correct border by @Builditluc in #118
- Fix large release size by @Builditluc in #119
- Fix short articles having a limited size by @Builditluc in #120
- Fix external links not being recognized by @Builditluc in #121

Other Changes
- Bump cursive to 0.20 by @Builditluc in #106
- Change GitHub workflow badge routes by @Builditluc in #107
- Update contributors badge style by @Builditluc in #108
- Update lockfile dependencies by @Builditluc in #122
- Fix spelling error by @Builditluc in #123

(pin)

doc: Updated meta-pkgs/xfce4 to 4.18.0nb1

(gutteridge)

xfce4: reflect correct current minimums for Xfce 4.18

(gutteridge)

doc: Updated meta-pkgs/xfce4-extras to 4.18.0nb6

(gutteridge)

xfce4-extras: bump a few packages now we're at 4.18

(gutteridge)

doc: Updated devel/xfce4-dev-tools to 4.18.0

(gutteridge)

xfce4-dev-tools: update to 4.18.0

Change log:

4.18.0 (2022-12-15)
======
- build: Bump requirements for Xfce 4.18

4.17.2 (2022-12-01)
======
- Add new libsoup3 to the builder
- xfce-do-release: unification of update version 'sed' command
- xfce-do-release: match multiple spaces for git version replace
- bump copyright year to 2022

(gutteridge)

doc: Updated sysutils/xfce4-thunar to 4.18.1

(gutteridge)

xfce4-thunar: update to 4.18.1

Change log:

4.18.1 (2022-12-22)
======
- Fix several thumbnailing issues
- 'replace' option fixed for copy with '*.partial~' (Issue #974)
- Preserve trusted state when launchers are copied
- Dont reset zoom-level for directory specific settings (issue #970)
- Prevent build failure for NixOS (Issue #972)
- Prevent location bar crash (Issue #968)
- Translation Updates:
  Albanian, Arabic, Armenian, Armenian (Armenia), Asturian, Basque,
  Belarusian, Bengali, Bulgarian, Catalan, Chinese (China), Chinese
  (Hong Kong), Chinese (Taiwan), Croatian, Czech, Danish, Dutch,
  Eastern Armenian, English (Australia), English (United Kingdom),
  Esperanto, Estonian, Finnish, French, Galician, German, Greek,
  Hebrew, Hungarian, Icelandic, Indonesian, Interlingue, Italian,
  Japanese, Kazakh, Korean, Latvian, Lithuanian, Malay, Norwegian
  Bokm鐃緒申l, Norwegian Nynorsk, Occitan (post 1500), Persian (Iran),
  Polish, Portuguese, Portuguese (Brazil), Romanian, Russian, Serbian,
  Slovak, Slovenian, Spanish, Swedish, Telugu, Thai, Ukrainian, Urdu,
  Urdu (Pakistan), Uyghur, Vietnamese

(gutteridge)

xfce4-mousepad: add a comment about a CSD-only dependency

(gutteridge)

doc: Updated x11/xfce4-settings to 4.18.1

(gutteridge)

xfce4-settings: update to 4.18.1

Change log:

4.18.1 (2022-12-22)
======
- xfsettingsd: Make CSD (client side decorations) for dialogs opt-in
- Translation Updates:
  Albanian, Arabic, Armenian (Armenia), Asturian, Basque, Belarusian,
  Belarusian (Tarask), Bulgarian, Catalan, Chinese (China), Chinese
  (Hong Kong), Chinese (Taiwan), Croatian, Czech, Danish, Dutch,
  Eastern Armenian, English (Australia), English (Canada), English
  (United Kingdom), Estonian, Finnish, French, Galician, Georgian,
  German, Greek, Hebrew, Hungarian, Icelandic, Indonesian, Interlingue,
  Italian, Japanese, Kabyle, Kazakh, Korean, Lithuanian, Malay,
  Norwegian Bokm鐃緒申l, Norwegian Nynorsk, Occitan (post 1500), Polish,
  Portuguese, Portuguese (Brazil), Romanian, Russian, Serbian, Slovak,
  Slovenian, Spanish, Swedish, Thai, Turkish, Ukrainian, Uyghur

(gutteridge)

doc: Updated sysutils/xfce4-systemload-plugin to 1.3.2

(gutteridge)

xfce4-systemload-plugin: update to 1.3.2

Change log:

1.3.2 (2022-11-03)
=====
- Remove code obsoleted by libxfce4util-4.17.2
- Fix GLib warnings occurring with newer xfce4-dev-tools
- Fix intltool lock file bug in make distcheck
- Code cleanup: Update variable declarations, extend maximum line length
- Code cleanup: Replace gboolean with bool where possible
- Code cleanup: Remove void from (void)
- Code cleanup: Remove typedef keywords
- Add BSD-2 license badge to README
- Rename all files from C to C++
- Bump required Xfce4 libs from 4.12 to 4.16
- Update copyright notices, copyright years and author emails
- Cleanup Makefile.am
- Update configure.ac.in syntax
- Document rgba_equal() function
- Don't save default colors to Xfconf configuration file
- Move swap to be next to mem
- Update default colors
- Decouple the program order from the visual order of monitors
- Bump required XDT version to 4.14
- Avoid G_GNUC_FALLTHROUGH to enable compilation with GLib earlier than 2.60
- Translation Updates:
  Albanian, Arabic, Basque, Belarusian, Bulgarian, Catalan, Chinese
  (China), Chinese (Taiwan), Croatian, Czech, Danish, Dutch, Eastern
  Armenian, English (Australia), English (Canada), English (United
  Kingdom), Estonian, Finnish, French, Galician, German, Greek, Hebrew,
  Hungarian, Icelandic, Indonesian, Interlingue, Italian, Japanese,
  Korean, Lithuanian, Malay, Norwegian Bokm鐃緒申l, Occitan (post 1500),
  Polish, Portuguese, Russian, Serbian, Slovak, Slovenian, Spanish,
  Swedish, Thai, Turkish, Ukrainian, Uyghur

(gutteridge)

xfce4-cpugraph-plugin: note specific libxfce4util dependency

(Ride previous update.)

(gutteridge)

doc/TODO: -xfce-4.18, +xfce4-screenshooter

xfce4-screenshooter is blocked, as it requires libsoup >= 3.0.

(gutteridge)

doc: Updated x11/xfce4-session to 4.18.0

(gutteridge)

xfce4-session: update to 4.18.0

Change log:

4.18.0 (2022-12-15)
======
- build: Bump requirements for Xfce 4.18
- Translation Updates:
  Russian, Turkish

4.17.2 (2022-12-01)
======
- logout-dialog: New Xfconf property "ShowSave" (!29)
- bump copyright year to 2022

4.17.1 (2022-11-01)
======

- Translation Updates:
  Portuguese

4.17.0 (2022-09-14)
======
- settings: Revert type hint change
- Check getgrgid(3) return value to avoid crash
- build: Fix intltool lock file problem during make distcheck
- Remove pre-defined lockers
- Refresh list of screensavers (#17)
- Do not mind output of screensaver, if successful
- Run custom command in a separate shell
- Avoid extra pause by redirecting stdin
- Update COPYING
- Increase switch user timeout (!25)
- xflock4: Apply initial delay before running a screensaver/locker (#29)
- xflock4: Update copyrights
- xflock4: make the script executable
- Add labels to inline toolbar buttons
- Use correct icon for Close button
- Translation Updates:
  Albanian, Arabic, Armenian (Armenia), Asturian, Azerbaijani, Basque,
  Belarusian, Belarusian (Tarask), Bengali, Bulgarian, Catalan, Chinese
  (China), Chinese (Hong Kong), Chinese (Taiwan), Croatian, Czech,
  Danish, Dutch, Eastern Armenian, English (Australia), English
  (Canada), English (United Kingdom), Estonian, Finnish, French,
  Galician, German, Greek, Hebrew, Hindi, Hungarian, Icelandic,
  Indonesian, Interlingue, Italian, Japanese, Kazakh, Korean,
  Lithuanian, Malay, Norwegian Bokm鐃緒申l, Norwegian Nynorsk, Occitan
  (post 1500), Persian (Iran), Polish, Portuguese, Portuguese (Brazil),
  Romanian, Russian, Serbian, Slovak, Slovenian, Spanish, Swedish,
  Thai, Turkish, Ukrainian, Uyghur, Vietnamese

(gutteridge)

doc: Updated sysutils/xfce4-cpugraph-plugin to 1.2.7

(gutteridge)

xfce4-cpugraph-plugin: update to 1.2.7

Fixes builds now that we have libxfce4util 4.18.

Change log:

1.2.7 (2022-11-03)
=====
- Remove code obsoleted by libxfce4util-4.17.2
- Fix GLib warnings occurring with newer xfce4-dev-tools
- Fix intltool lock file bug in make distcheck
- Translation Updates:
  Belarusian, Czech, Malay, Portuguese, Portuguese (Brazil), Russian,
  Turkish, Ukrainian

(gutteridge)

Updated net/cppzmq, mail/dovecot2, mail/dovecot2-pigeonhole

(adam)

dovecot2: updated to 2.3.20

v2.3.20

+ Add dsync_features=no-header-hashes. When this setting is enabled and
  one dsync side doesn't support mail GUIDs (i.e. imapc), there is no
  fallback to using header hashes. Instead, dsync assumes that all mails
  with identical IMAP UIDs contains the same mail contents. This can
  significantly improve dsync performance with some IMAP servers that
  don't support caching Date/Message-ID headers.
+ lua: HTTP client has more settings now, see
  https://doc.dovecot.org/admin_manual/lua/#dovecot.http.client
+ replicator: "doveadm replicator status" command now outputs when the
  next sync is expected for the user.
- LAYOUT=index: duplicate GUIDs were not cleaned out. Also the list
  recovery was not optimal.
- auth: Assert crash would occur when iterating multiple userdb
  backends.
- director: Logging into director using master user with
  auth_master_user_separator character redirected user to a wrong
  backend, unless master_user_separator setting was also set to the same
  value. Merged these into auth_master_user_separator.
- dsync: Couldn't always fix folder GUID conflicts automatically with
  Maildir format. This resulted in replication repeatedly failing
  with "Remote lost mailbox GUID".
- dsync: Failed to migrate INBOX when using namespace prefix=INBOX/,
  resulting in "Remote lost mailbox GUID" errors.
- dsync: INBOX was created too early with namespace prefix=INBOX/,
  resulting a GUID conflict. This may have been resolved automatically,
  but not always.
- dsync: v2.3.18 regression: Wrong imapc password with dsync caused
  Panic: file lib-event.c: line 506 (event_pop_global):
  assertion failed: (event == current_global_event)
- imapc: Requesting STATUS for a mailbox with imapc and INDEXPVT
  configured did not return correct (private) unseen counts.
- lib-dict: Process would crash when committing data to redis without
  dict proxy.
- lib-mail: Corrupted cached BODYSTRUCTURE caused panic during FETCH.
  Fixes: Panic: file message-part-data.c: line 579 (message_part_is_attachment):
  assertion failed: (data != NULL). v2.3.13 regression.
- lib-storage: mail_attribute_dict with dict-sql failed when it tried to
  lookup empty dict keys.
- lib: ioloop-kqueue was missing include breaking some BSD builds.
- lua-http: Dovecot Lua HTTP client could not resolve DNS names in mail
  processes, because it expected "dns-client" socket to exist in the
  current directory.
- oauth2: Using %{oauth2:name} variables could cause useless
  introspections.
- pop3: Sending POP3 command with ':' character caused an assert-crash.
  v2.3.18 regression.
- replicator: Replication queue had various issues, potentially causing
  replication requests to become stuck.
- stats: Invalid Prometheus label names were created with specific

v0.5.20 of Pigeonhole
* No changes - release done to keep version numbers synced.

(adam)

cppzmq: updated to 4.9.0

v4.9.0

Type-safe socket.get(sockopt::type)
GCC C++14 fix
Support ZMQ_ONLY_FIRST_SUBSCRIBE
Add pkg-config file and instructions
Fix missing include
Fix monitor test failures
Improved documentation

(adam)

doc: Added devel/breezy-qbrz version 0.4.1.0.1651

(rhialto)

devel/breezy-qbrz: Successor of qbzr, but with Python 3.x and Qt 5.

A GUI for the Breezy version control system.

No tagged releases, distfiles, nor release notes available.
Version 0.4.1 from version.txt.

(rhialto)

gcc9-libs: bump PKGREVISION to follow gcc9 bump.

(tsutsui)

gcc9: pull a hack for the __stack_chk_fail_local() on NetBSD.

Taken from src/external/gpl3/gcc/dist/gcc/targhooks.c in the NetBSD HEAD:
http://cvsweb.netbsd.org/bsdweb.cgi/src/external/gpl3/gcc/dist/gcc/targhooks.c#rev1.2
via pkgsrc/lang/gcc49 in Attic:
http://cvsweb.netbsd.org/bsdweb.cgi/pkgsrc/lang/gcc49/patches/Attic/patch-gcc_targhooks.c#rev1.1

This fixes an error during make configure in pkgsrc/x11/qt6-qtbase
on NetBSD/i386 9.3 as mentioned in PR/57151.

Bump PKGREVISION.

(tsutsui)

doc/TODO: + ImageMagick-7.1.0.57, MesaLib-22.3.2.

(wiz)

doc: Updated net/dbip-country-lite to 2023.01

(fcambus)

dbip-country-lite: update to 2023.01.

No upstream ChangeLog.

(fcambus)

doc: Updated net/dbip-city-lite to 2023.01

(fcambus)

dbip-city-lite: update to 2023.01.

No upstream ChangeLog.

(fcambus)

doc: Updated net/dbip-asn-lite to 2023.01

(fcambus)

dbip-asn-lite: update to 2023.01.

No upstream ChangeLog.

(fcambus)

(databases/R-RSQLite) Removed redundent TEST_DEPENDS+

(mef)

doc: Updated databases/R-RSQLite to 2.2.20

(mef)

(databases/R-RSQLite) Updated 2.2.7 to 2.2.20

# RSQLite 2.2.20 (2022-12-21)

## Features

- Upgrade bundled SQLite to 3.40.0 (#446).

## Tests

- Fix test on non-UTF-8 locales.

## Internal

- GitPod and cmake integration.

# RSQLite 2.2.19 (2022-11-22)

## Bug fixes

- Add missing boost headers for icpc compiler (@lolow, #442, #443).

## Internal

- Use testthat edition 3 (#438).

# RSQLite 2.2.18 (2022-10-04)

- Upgrade bundled SQLite to 3.39.4 (#435).

# RSQLite 2.2.17 (2022-09-09)

- Upgrade bundled SQLite to 3.39.3 (#433).

# RSQLite 2.2.16 (2022-08-16)

- Upgrade bundled SQLite to 3.39.2 (#431).

# RSQLite 2.2.15 (2022-07-15)

- Upgrade bundled SQLite to 3.39.1 (#429).

# RSQLite 2.2.14 (2022-05-05)

- Upgrade bundled SQLite to version 3.38.5 (#421, #423).

# RSQLite 2.2.13 (2022-04-29)

- Upgrade bundled SQLite to version 3.38.3 (#418).

# RSQLite 2.2.12 (2022-04-02)

## Features

- Upgrade bundled SQLite to version 3.38.2 (#415).

## Internal

- Prefer `empty()` over `size() == 0` (#410, @MichaelChirico).
- Refactor to improve code consistency (#411, @MichaelChirico).
- Use more parsimonious header to avoid CRAN warnings.

# RSQLite 2.2.11 (2022-03-07)

## Features

- `initExtension()` gains `extension` argument (#407, #408) with options
  `"math"`, `"regexp"`, `"series"` and `"csv"` (#274, #389, @mgirlich).
- Upgrade bundled SQLite to version 3.38.0 (#405).

## Bug fixes

- Skip failing test on CRAN.

# RSQLite 2.2.10 (2022-02-16)

## Features

- Upgrade bundled SQLite to version 3.37.2 (#403).

## Documentation

- Explain reason for bundling SQLite sources.
- Add known SQLite authors to `DESCRIPTION`.

## Internal

- Add test for interrupt behavior in `sqliteSetBusyHandler()` (#401).
- Make method definition more similar to S3. All `setMethod()` calls
  refer to top-level functions (#396).

# RSQLite 2.2.9 (2021-12-06)

## Features

- Upgrade bundled SQLite to version 3.37.0 (#392).
- `dbConnect()` now supports URIs that start with `file://` for the
  `dbname` argument. The SQLite code is now built with the
  `SQLITE_USE_URI` (#390, @tschoonj).

## Bug fixes

- `dbBegin()`, `dbCommit()` and `dbRollback()` support keywords in
  savepoint names (#386).

## Internal

- Adapt to DBItest 1.7.2 (#385).
- Enable all DBItest tests, passing now.

# RSQLite 2.2.8 (2021-08-21)

- Upgrade bundled SQLite to version 3.36.0 (#374).
- Fix build on CentOS 7 (#367).
- Busy callback emits message instead of warning (#355).
- Avoid using memoise at build time (#371).

(mef)

doc: Updated math/R-tibble to 3.1.8

(mef)

(devel/R-tibble) Updated 3.1.4 to 3.1.8

# tibble 3.1.8

## Documentation

- Better reporting for error calls from `vec_as_location()` (#1237).

- Mention `median()` in Recovery section of `vignette("numbers")` (#1197).

# tibble 3.1.7

## Breaking change

- `trunc_mat()` now returns a value with a different structure. This
  is considered an implementation detail that can change in the
  future, do not rely on it. The only guarantee is that calling
  `print()` will display the input like a tibble (#1059).

## Documentation

- Avoid listing `dim_desc()` in reexports.
- Add more examples for data frame and matrix columns (#978, #1012).

## Internal

- Require rlang 1.0.1 and pillar 1.7.0 (#1063).
- Prefer `class` over `.subclass` in `rlang::error_cnd()` (#1015, #1060).

# tibble 3.1.6

- `set_num_opts()` and `set_char_opts()` are reexported from pillar (#959).

- `view()` uses `rlang::expr_deparse(width = Inf)` to avoid errors
  with long `|>` pipes (#957).

- `new_tibble()` checks that the `nrow` argument is nonnegative and
  less than 2^31 (#916).

- `tbl_sum.tbl_df()` has an ellipsis in its formals for extensibility.

# tibble 3.1.5

- Avoid necessity to set `"tibble.view_max"` option for lazy tables (#954).

- Avoid blanket import for lifecycle package for compatibility with
  upcoming rlang (#955, @romainfrancois).

(mef)

doc: Updated misc/R-pillar to 1.8.1

(mef)

(misc/R-pillar) Updated 1.6.2 to 1.8.1

# pillar 1.8.1

## Features

- New `pillar.advice` option to turn off advice in the footer, see `?pillar_options`. Now off by default in non-interactive mode (#577).

# pillar 1.8.0

## Display

- Column names that are abbreviated in the header gain a footnote and are printed in full in the footer (#483),
after the extra columns (#548). If a column name in the header is abbreviated, all backticks are removed (#525). The new `"pillar.superdigit_sep"` option that determines the string used to separate footnote from column name in the footer (#553).

- The default value of the `pillar.min_title_width` option is changed to 5. This means that effectively the width of a pillar is decided only by the data. Use `options(pillar.min_title_width = 15)` to restore the previous default, see also `?pillar_options` for details (#531).

- Offer advice in the footer on how to print all columns or rows (#567).

- Avoid aligning `NA` inside quotes for very short character vectors (#562).

## Features

- Pick up `"pillar_focus"` attribute on printing to define focus columns (#549).

- New `ctl_new_rowid_pillar()` generic and default method for customizing the appearance of row IDs (#260, #550, @nbenn).

## Bug fixes

- Fix printing of `Surv` and `Surv2` objects (#561).

- Fix wording for corner case `max_extra_cols = 1` (#535).

- Remove excess underlines for `bit64::integer64()` data of different magnitude (#517, #529).

## Documentation

- `ctl_new_pillar_list()` is documented on a separate help page (#516).

- Remove outdated detail in `?tbl_sum` (@IndrajeetPatil, #565).

## Chore

- Drop crayon dependency (#559).

- Import ellipsis from rlang (#554).

- Skip test that requires lubridate if it's not installed (#505, @MichaelChirico).

## API

- Soft-deprecate `colonnade()`, `squeeze()` and `extra_cols()` (#496).

- Require rlang 1.0.1 (#512).

# pillar 1.7.0

## Breaking changes

- `colonnade()` is now soft-deprecated (#485).
- `expect_known_display()` and `is_vector_s3()` are now deprecated (#460, #501).
- `new_pillar()` deprecates `extra` argument (#497).

## Features

- Focus columns specified via the `focus` argument to `tbl_format_setup()` are kept in their original place and shown with the maximum width and with the "type" component underlined (#465).

## Bug fixes

- Update `s3_register()` to use new implementation from rlang, this fixes CRAN checks related to `scale_type()` (#462).

## Internal

- Single pillars constructed with `pillar()` use only as much width as required when printing (#484).

# pillar 1.6.5

## Breaking changes

- New `ctl_new_pillar_list()`, supersedes `ctl_new_compound_pillar()` (#433).

## Features

- If some but not all sub-columns of a data frame or matrix column are shown, the names and types of the remaining columns are displayed in the footer (#365, #444).
- `num(fixed_exponent = ...)` is now represented with the fixed exponent in the pillar header, and in the title in ggplot2 (#307).
- `tbl_format_setup()` gains `focus` argument that expects a character vector of column names. Focus columns are moved to the front and separated from the main columns by a subtle vertical line (#384).
- New `scale_x_num()` and `scale_y_num()`. If a column created with `num()` is used in a ggplot, the x and y scale will be formatted automatically according to to the specification (#400, #404).
- List columns omit size information if horizontal space is insufficient (#392).
- If the column title of a backticked column is abbreviated, the trailing backtick is still printed (#391).
- `new_pillar_shaft_simple()` gains `short_formatted` argument that contains the data to be used if horizontal space is insufficient (#389).
- Default `obj_sum()` method returns abbreviation in attribute of return value (#390).

## Bug fixes

- Extra columns in footer show backticks again if they are non-syntactic (#393).
- Fixed some cases for combinations of printed width and `getOption("width")` (#432).
- Fix support for `nanotime::nanotime()` classes (#378, #380).

## Documentation

- `?num` and `?char` now point to tibble (#382).

## Internal

- Use eager registration via `NAMESPACE` for own methods for classes from other packages.
- Reworked formatting routine, now using a visitor-based approach with in-order iteration over all pillars. The only visible changes are that usage of free space (in the case of limited space) has slightly improved (#435).
- Prepared removal of dependency on the crayon package (#233, #406).
- Use snapshot variants, requires testthat >= 3.1.1 (#387).
- Replace internal `"pillar_vertical"` class with `glue::as_glue()` (#279).

# pillar 1.6.4

## Bug fixes

- Fix printing for some tibbles where a fixed-width column is followed by a column with variable width (#366).
- Avoid nested backtick blocks in vignette.

## Breaking changes

- `num()` requires an integerish `digits` argument (#362).

## Documentation

- Link to tibble vignettes and documentation pages.

# pillar 1.6.3

- Avoid blanket import for lifecycle package for compatibility with upcoming rlang (#368, @romainfrancois).

(mef)

(math/R-tibble) +TEST_DEPENDS+= R-mockr, R-lubridate, R-bench

(mef)

(devel/R-purrr) +TEST_DEPENDS+= R-lubridate (2 ERRORs now)

(mef)

doc: Updated archivers/R-zip to 2.2.2

(mef)

doc: Updated databases/R-dbplyr to 2.2.1

(mef)

(databases/R-dbplyr) Updated 2.1.1 to 2.2.1

# dbplyr 2.2.1

* Querying Oracle databases works again. Unfortunately, the fix requires every
  column to be explicitly selected again (@mgirlich, #908).

* `semi_join()` and `anti_join()` work again for Spark (@mgirlich, #915).

* `str_c()` is now translated to `||` in Oracle (@mgirlich, #921).

* `sd()`, `var()`, `cor()` and `cov()` now give clear error messages on
  databases that don't support them.

* `any()` and `all()` gain default translations for all backends.

# dbplyr 2.2.0

## New features

* SQL formatting has been considerably improved with new wrapping and indenting.
  `show_query()` creates more readable queries by printing the keywords in blue
  (@mgirlich, #644). When possible dbplyr now uses `SELECT *` instead of
  explicitly selecting every column (@mgirlich).

* Added support for `rows_insert()`, `rows_append()`, `rows_update()`,
  `rows_patch()`, `rows_upsert()`, and `rows_delete()` (@mgirlich, #736).

* Added `copy_inline()` as a `copy_to()` equivalent that does not need write
  access (@mgirlich, #628).

* `remote_query()`, `show_query()`, `compute()` and `collect()` have an
  experimental `cte` argument. If `TRUE` the SQL query will use common table
  expressions instead of nested queries (@mgirlich, #638).

* New `in_catalog()`, which works like `in_schema()`, but allows creation of
  table identifiers consisting of three components: catalog, schema, name
  (#806, @krlmlr).

## Improvements to SQL generation

* When possible, dbplyr now uses `SELECT *` instead of explicitly selecting
  every column (@mgirlich).

* New translation for `cut()` (@mgirlich, #697).

* Improved translations for specific backends:
  * `as.Date()` for Oracle (@mgirlich, #661).
  * `case_when()` with a final clause of the form `TRUE ~ ...` uses `ELSE ...`
    for SQLite (@mgirlich, #754).
  * `day()`, `week()`, `isoweek()`, and `isoyear()` for Postgres (@mgirlich, #675).
  * `explain()` for ROracle (@mgirlich).
  * `fill()` for SQL Server (#651, @mgirlich) and RPostgreSQL (@mgirlich).
  * `quantile()` for SQL Server (@mgirlich, #620).
  * `str_flatten()` for Redshift (@hdplsa, #804)
  * `slice_sample()` for MySQL/MariaDB and SQL Server (@mgirlich, #617).
  * `union()` for Hive (@mgirlich, #663).

* The backend function `dbplyr_fill0()` (used for databases that lack
  `IGNORE NULLS` support) now respects database specific translations
  (@rsund, #753).

* Calls of the form `stringr::foo()` or `lubridate::foo()` are now evaluated in
  the database, rather than locally (#197).

* Unary plus (e.g. `db %>% filter(x == +1)`) now works (@mgirlich, #674).

* `is.na()`, `ifelse()`, `if_else()`, `case_when()`, and `if()`
  generate slightly more compact SQL (@mgirlich, #738).

* `if_else()` now supports the `missing` argument (@mgirlich, #641).

* `n()` now respects the window frame (@mgirlich, #700).

* `quantile()` no longer errors when using the `na.rm` argument (@mgirlich, #600).

* `remote_name()` now returns a name in more cases where it makes sense
  (@mgirlich, #850).

* The partial evaluation code is now more aligned with `dtplyr`. This makes it
  easier to transfer bug fixes and new features from one package to the other.
  In this process the second argument of `partial_eval()` was changed to a lazy
  frame instead of a character vector of variables (@mgirlich, #766).
  Partially evaluated expressions with infix operations are now correctly
  translated. For example `translate_sql(!!expr(2 - 1) * x)` now works
  (@mgirlich, #634).

## Minor improvements and bug fixes

* New `pillar::tbl_format_header()` method for lazy tables: Printing a lazy
  table where all rows are displayed also shows the exact number of rows in the
  header. The threshold is controlled by `getOption("pillar.print_min")`,
  with a default of 10 (#796, @krlmlr).

* The 1st edition extension mechanism is formally deprecated (#507).

* `across()`, `if_any()` and `if_all()` now defaults to `.cols = everything()`
  (@mgirlich, #760). If `.fns` is not provided `if_any()` and `if_all()` work
  like a parallel version of `any()`/`any()` (@mgirlich, #734).

* `across()`, `if_any()`, and `if_all()` can now translate evaluated lists
  and functions (@mgirlich, #796), and accept the name of a list of functions
  (@mgirlich, #817).

* Multiple `across()` calls in `mutate()` and `transmute()` can now access
  freshly created variables (@mgirlich, #802).

* `add_count()` now doesn't change the groups of the input (@mgirlich, #614).

* `compute()` can now handle when `name` is named by unnaming it first
  (@mgirlich, #623), and now works when `temporary = TRUE` for Oracle
  (@mgirlich, #621).

* `distinct()` now supports `.keep_all = TRUE` (@mgirlich, #756).

* `expand()` now works in DuckDB (@mgirlich, #712).

* `explain()` passes `...` to methods (@mgirlich, #783), and
  works for Redshift (@mgirlich, #740).

* `filter()` throws an error if you supply a named argument (@mgirlich, #764).

* Joins disambiguates columns that only differ in case (@mgirlich, #702).
  New arguments `x_as` and `y_as` allow you to control the table alias
  used in SQL query (@mgirlich, #637). Joins with `na_matches = "na"` now work
  for DuckDB (@mgirlich, #704).

* `mutate()` and `transmute()` use named windows if a window definition is
  used at least twice and the backend supports named windows (@mgirlich, #624).

* `mutate()` now supports the arguments `.keep`, `.before`, and `.after`
  (@mgirlich, #802).

* `na.rm = FALSE` only warns once every 8 hours across all functions (#899).

* `nesting()` now supports the `.name_repair` argument (@mgirlich, #654).

* `pivot_longer()` can now pivot a column named `name` (@mgirlich, #692),
  can repair names (@mgirlich, #694), and can work with multiple `names_from`
  columns (@mgirlich, #693).

* `pivot_wider(values_fn = )` and `pivot_longer(values_transform = )`
  can now be formulas (@mgirlich, #745).

* `pivot_wider()` now supports the arguments `names_vary`, `names_expand`, and
  `unused_fn` (@mgirlich, #774).

* `remote_name()` now returns a name in more cases where it makes sense
  (@mgirlich, #850).

* `sql_random()` is now exported.

* `ungroup()` removes variables in `...` from grouping (@mgirlich, #689).

* `transmute()` now keeps grouping variables (@mgirlich, #802).

(mef)

(archivers/R-zip) Updated 2.2.0 to 2.2.2

# zip 2.2.2

* No user visible changes.

# zip 2.2.1

* No user visible changes.

(mef)

doc: Updated devel/R-repr to 1.1.4

(mef)

(devel/R-repr) Updated 1.1.3 to 1.1.4

repr 1.1.4
* See the [GitHub releases](https://github.com/IRkernel/repr/releases) page

    [#144] Fix html filter to also work with R 4.1.1 html (@renkun-ken)
    [#149] Switch CI to GH actions by (@flying-sheep)

(mef)

doc: Updated devel/R-pkgcache to 2.0.4

(mef)

(devel/R-pkgcache) Updated 2.0.3 to 2.0.4

# pkgcache 2.0.4

* Update R version -> Bioconductor version mapping. R 4.2.x now maps to
  Bioconductor 3.16.

(mef)

doc: Updated www/R-webfakes to 1.1.6

(mef)

(www/R-webfakes) Updated 1.1.3 to 1.1.6

# webfakes 1.1.6

* `response$send_file()` now handles `root = "/"` and absolute paths
  better on Windows.

* `new_app_process()` and `local_app_process()` are now faster,
  because the app object they need to copy to the subprocess is smaller.

# webfakes 1.1.5

* `mw_etag()` now handles the `If-None-Match` header properly, and sets
  the status code of the response to 304, and removes the response body.

# webfakes 1.1.4

* No user visible changes.

(mef)

doc: Updated devel/R-purrr to 1.0.0

(mef)

(devel/R-purrr) Update 0.3.5 to 1.0.0

# purrr 1.0.0

## Breaking changes

### Core purpose refinements

* `cross()` and all its variants have been deprecated in favour of
  `tidyr::expand_grid()`. These functions were slow and buggy and we
  no longer think they are the right approach to solving this problem.
  See #768 for more information.

* `update_list()` (#858) and `rerun()` (#877), and the use of tidyselect
  with `map_at()` and friends (#874) have been deprecated. These functions
  use some form of non-standard evaluation which we now believe is a poor
  fit for purrr.

* The `lift_*` family of functions has been deprecated. We no longer believe
  these to be a good fit for purrr because they rely on a style of function
  manipulation that is very uncommon in R code (#871).

* `prepend()`, `rdunif()`, `rbernoulli()`, `when()`, and `list_along()` have
  all been deprecated (#925). It's now clear that they don't align with the
  core purpose of purrr.

* `splice()` is deprecated because we no longer believe that automatic
  splicing makes for good UI. Instead use `list2()` + `!!!` or
  `list_flatten()` (#869).

### Mapping

* Use of map functions with expressions, calls, and pairlists has been
  deprecated (#961).

* All map `_raw()` variants have been deprecated because they are of limited
  use and you can now use `map_vec()` instead (#903).

* In `map_chr()`, automatic conversion from logical, integer, and double to
  character is now deprecated. Use an explicit `as.character()` if needed
  (#904).

### Deprecation next steps

* `as_function()` and the `...f` argument to `partial()` are no longer
  supported. They have been defunct for quite some time.

* Soft deprecated functions: `%@%`, `reduce_right()`, `reduce2_right()`,
  `accumulate_right()` are now fully deprecated. Similarly, the
  `.lazy`, `.env`, and `.first` arguments to `partial()`,
  and the `.right` argument to `detect()` and `detect_index()`
  are fully deprecated. Removing elements with `NULL` in `list_modify()` and
  `list_merge()` is now fully deprecated.

* `is_numeric()` and `is_scalar_numeric()` have been removed. They have
  been deprecated since purrr 0.2.3 (Sep 2017).

* `invoke_*()` is now deprecated. It was superseded in 0.3.0 (Jan 2019) and
  3.5 years later, we have decided to deprecate it as part of the API
  refinement in the 1.0.0 release.

* `map_call()` has been removed. It was made defunct in 0.3.0 (Jan 2019).

## New features

* `*_at()` can now take a function (or formula) that's passed the vector of
  element names and returns the elements to select.

* New `map_vec()`, `map2_vec()`, and `pmap_vec()` work on all types of vectors,
  extending `map_lgl()`, `map_int()`, and friends so that you can easily work
  with dates, factors, date-times and more (#435).

* New `keep_at()` and `discard_at()` that work like `keep()` and `discard()`
  but operation on element names rather than element contents (#817).

* Some mapping functions have now a `.progress` argument to create a
  progress bar. See `?progress_bars` (#149).

* purrr is now licensed as MIT (#805).

* `modify()`, `modify_if()`, `modify_at()`, and `modify2()` are no longer
  generics. We have discovered a simple implementation that no longer requires
  genericity and methods were only provided by a very small number of packages
  (#894).

* purrr now uses the base pipe (`|>`) and anonymous function short hand (`\(x)`),
  in all examples. This means that examples will no longer work in R 4.0 and
  earlier so in those versions of R, the examples are automatically converted
  to a regular section with a note that they might not work (#936).

* When map functions fail, they now report the element they failed at (#945).

* New `modify_tree()` for recursively modifying nested data structures (#720).

### Flattening and simplification

* New `list_c()`, `list_rbind()`, and `list_cbind()` make it easy to
  `c()`, `rbind()`, or `cbind()` all of the elements in a list.

* New `list_simplify()` reduces a list of length-1 vectors to a simpler atomic
  or S3 vector (#900).

* New `list_transpose()` which automatically simplifies if possible (#875).

* `accumulate()` and `accumulate2()` now both simplify the output if possible
  using vctrs. New arguments `simplify` and `ptype` allow you to control the
  details of simplification (#774, #809).

* `flatten()` and friends are superseded in favour of `list_flatten()`,
  `list_c()`, `list_cbind()`, and `list_rbind()`.

* `*_dfc()` and `*_dfr()` have been superseded in favour of using the
  appropriate map function along with `list_rbind()` or `list_cbind()` (#912).

* `simplify()`, `simplify_all()`, and `as_vector()` have been superseded in
  favour of `list_simplify()`. It provides a more consistent definition of
  simplification (#900).

* `transpose()` has been superseded in favour of `list_transpose()` (#875).
  It has built-in simplification.

### Tidyverse consistency

* `_lgl()`, `_int()`, `_int()`, and `_dbl()` now use the same (strict) coercion
  methods as vctrs (#904). This means that:

    * `map_chr(TRUE, identity)`, `map_chr(0L, identity)`, and
      `map_chr(1L, identity)` are deprecated because we now believe that
      converting a logical/integer/double to a character vector should require
      an explicit coercion.

    * `map_int(1.5, identity)` now fails because we believe that silently
      truncating doubles to integers is dangerous. But note that
      `map_int(1, identity)` still works since no numeric precision is lost.

    * `map_int(c(TRUE, FALSE), identity)`, `map_dbl(c(TRUE, FALSE), identity)`,
      `map_lgl(c(1L, 0L), identity)` and `map_lgl(c(1, 0), identity)` now
      succeed because 1/TRUE and 0/FALSE should be interchangeable.

* `map2()`, `modify2()`, and `pmap()` now use tidyverse recycling rules where
  vectors of length 1 are recycled to any size but all others must have
  the same length (#878).

* `map2()` and `pmap()` now recycle names of their first input if
  needed (#783).

* `modify()`, `modify_if()`, and `modify_at()` have been reimplemented using
  vctrs principles. This shouldn't have an user facing impact, but it does
  make the implementation much simpler.

### Plucking

* `vec_depth()` is now `pluck_depth()` and works with more types of input
  (#818).

* `pluck()` now requires indices to be length 1 (#813). It also now reports
  the correct type if you supply an unexpected index.

* `pluck()` now accepts negative integers, indexing from the right (#603).

* `pluck()` and `chuck()` now fail if you provide named inputs to ... (#788).

* `pluck()` no longer replaces 0-length vectors with `default`; it now
  only applies absent and `NULL` components (#480).

* `pluck<-`/`assign_in()` can now modify non-existing locations (#704).

### Setting with NULL

* `pluck<-`/`assign_in()` now sets elements to `NULL` rather than removing them
  (#636). Now use the explicit `zap()` if you want to remove elements.

* `modify()`, `modify2()`, and `modify_if()` now correctly handle `NULL`s
  in replacement values (#655, #746, #753).

* `list_modify()`'s interface has been standardised. Modifying with `NULL`
  now always creates a `NULL` in the output (#810)

### `list_` functions`

* New `list_assign()` which is similar to `list_modify()` but doesn't work
  recursively (#822).

* `list_modify()` no longer recurses into data frames (and other objects built
  on top of lists that are fundamentally non-list like) (#810). You can
  revert to the previous behaviour by setting `.is_node = is.list`.

## Minor improvements and bug fixes

* `capture_output()` correctly uses `conditionMessage()` instead of directly
  interrogating the `message` field (#1010).

* `modify()` no longer works with calls or pairlists.

* `modify_depth()` is no longer a generic. This makes it more consistent
  with `map_depth()`.

* `map_depth()` and `modify_depth()` have a new `is_node` argument that
  allows you to control what counts as a level. The default uses
  `vec_is_list()` to avoid recursing into rich S3 objects like linear models
  or data.frames (#958, #920).

* `map_depth()` and `modify_depth()` now correctly recurse at depth 1.

* `as_mapper()` is now around twice as fast when used with character,
  integer, or list (#820).

* `possibly()` now defaults `otherwise` to NULL.

* `modify_if(.else)` is now actually evaluated for atomic vectors (@mgirlich,
  #701).

* `lmap_if()` correctly handles `.else` functions (#847).

* `every()` now correctly propagates missing values using the same
  rules as `&&` (#751). Internally, it has become a wrapper around
  `&&`. This makes it consistent with `&&` and also with `some()`
  which has always been a wrapper around `||` with the same
  propagation rules.

* `every()` and `some()` now properly check the return value of their
  predicate function. It must now return a `TRUE`, `FALSE`, or `NA`.

* Greatly improved performance of functions created with `partial()` (#715).
  Their invocation is now as fast as for functions creating manually.

* `partial()` no longer inlines the function in the call stack. This
  fixes issues when `partial()` is used with `lm()` for instance (#707).

(mef)

(devel/R-cli) Updated to 3.5.0

(mef)

(devel/R-cli) Updated 3.4.1 to 3.5.0

# cli 3.5.0

* New `keypress()` function to read a single key press from a terminal.

* New function `pretty_print_code()` to print function objects with syntax
  highlighting at the R console.

* `col_*` and `bg_*` functions how handle zero-length input correctly (#532).

* New function `ansi_collapse()` to collapse character vectors into a single
  string.

* `ansi_strtrim()` now handles some edge cases better, when `ellipsis` has
  length zero, and when it is wider than `width`.

* New `hash_file_md5()` function to calculate the MD5 hash of one or more
  files.

(mef)

py-sip6: add PYTHON_SELF_CONFLICT

(adam)

doc: Updated sysutils/pciutils to 3.9.0

(msaitoh)

py-qt-builder: update CATEGORIES

(adam)

py-sip6: update CATEGORIES and HOMEPAGE

(adam)

pciutils: Update to 3.9.0.

2022-11-20  Martin Mares <mj@ucw.cz>

* Released as 3.9.0.

* We decode Compute Express Link (CXL) capabilities.

* The tree mode of lspci is now compatible with filtering options.

* When setpci is used with a named register, it checks whether
  the register is present in the particular header type.

* Linux: The intel-conf[12] back-ends prefer to use ioperm() instead
  of iopl() to gain access to I/O ports.

* Windows: We have two new back-ends thanks to Pali Roh叩r.
  One uses the NT SysDbg interface, the other uses kldbgdrv.sys
  (which is a part of the Microsoft WinDbg tool).

* Windows: We support building libpci as a DLL. Also, Windows
  binaries now include meta-data with version.

* Hurd: The Hurd back-end works again.

* mmio-conf1(-ext): Added a new back-end implementing the intel-conf1
  interface over MMIO. This is useful on some ARM machines, but it
  requires manual configuration of the MMIO addresses.

* As usually, updated pci.ids to the current snapshot of the database.

(msaitoh)

CHANGES for knightcap-brain

(dholland)

knightcap-brain: Fix wrong version globs that failed to exclude -10.

Also, these should be BROKEN_ON rather than NOT_FOR.

(dholland)

Updated textproc/utf8-cpp, textproc/ugrep

(adam)

ugrep: updated to 3.9.3

ugrep v3.9.3
Fixed a PCRE2 library-related issue with -P (--perl-regexp) matching binary files.

(adam)

utf8-cpp: updated to 3.2.3

3.2.3
A minor release that includes a fix for https://github.com/nemtrif/utfcpp/issues/92[: Clang 10 Compiler warnings](e3e57c8) and build system improvements

(adam)

Updated net/py-smbc, time/py-ciso8601

(adam)

py-ciso8601: updated to 2.3.0

Version 2.3.0

* Added Python 3.11 support
* Fix the build for PyPy2
* Added missing `fromutc` implementation for `FixedOffset`.
* Removed improper ability to call `FixedOffset`'s `dst`, `tzname` and `utcoffset` without arguments
* Fixed: `datetime.tzname` returns a `str` in Python 2.7, not a `unicode`
* Change `METH_VARARGS` to `METH_O`, enhancing performance.
* Added support for ISO week dates,
* Added support for ordinal dates

(adam)

py-smbc: updatedd to 1.0.24

1.0.24
Unknown changes

(adam)

make available for NetBSD platforms going forwards to 99

(plunky)