digikam: update to 7.9.0.

4 years of development.

(wiz)

doc: Updated security/pam-yubico to 2.27

(wiz)

pam-yubico: update to 2.27.

* Version 2.27 (released 2021-04-09)

** Add always_prompt configuration option.

** Add client certificate support for ldap.

** Add starttls support for ldap.

** Add ldap_bind_as_user support.

** Parsing, cleanliness and string fixes.

** Documentation and spelling fixes.

* Version 2.26 (released 2018-04-20)

** Make sure to close authfile (CVE-2018-9275).

** Fix compiler warnings.

** Open file descriptors with O_CLOEXEC.

** Use mkostemp() instead of mkstemp().

* Version 2.25 (released 2018-03-27)

** Documentation updates.

** Only do OTP validation if it's a token that might be valid.

** Return early in case user has no valid tokens.

** Ldap, compare values only with yubi_attr attributes.

** Add nullok parameter.

* Version 2.24 (released 2016-11-25)

** Debug mode changed, allows file output with debug_file.

** Fixup returning user-unknown correctly.

* Version 2.23 (released 2016-06-15)

** Fix an issue where a failure to set permissions was wrongly outputted.

* Version 2.22 (released 2016-05-23)

** Documentation improvements.

** Retain ownership and permission of challenge files (issue #92).

** Make dependency on yubico-c-client 2.15 clearer.

* Version 2.21 (released 2016-02-19)

** Add proxy support for yubico-c-client.

** Check that conv is set before trying to use it
fixes a crash bug with the osx loginwindow.

** Add building of a mac installer.

* Version 2.20 (released 2015-09-22)

** Add cainfo option to allow usage of a cabundle instead of path.

** Support comments in authfile.

** For challenge response with system-wide directory, write the files as root
instead of the user.

* Version 2.19 (released 2015-03-23)

** Add new ldap functionality
ldap_bind_user and ldap_bind_password for authenticated binds
ldap_filter for using subtree search and a filter
ldap_cacertfile to use a specific cacert for ldaps

* Version 2.18 (released 2015-02-12)

** Fix a memory leak of the pam response data.

** Add more tests.

** Add version flag to ykpamcfg.

(wiz)

py-yubikey-manager: fix PKGNAME

(wiz)

doc: Updated security/ykman to 5.0.1

(wiz)

doc: Updated security/py-yubikey-manager to 5.0.1

(wiz)

py-yubikey-manager, ykman: update to 5.0.1

* Version 5.0.1 (released 2023-01-17)
** Bugfix: Fix the interactive confirmation prompt for some CLI commands.
** Bugfix: OpenPGP Signature PIN policy values were swapped.
** Bugfix: FIDO: Handle discoverable credentials that are missing name or displayName.
** Add support for Python 3.11.
** Remove extra whitespace characters from CLI into command output.

* Version 5.0.0 (released 2022-10-19)
** Various cleanups and improvements to the API.
** Improvements to the handling of YubiKeys and connections.
** Command aliases for ykman 3.x (introduced in ykman 4.0) have now been dropped.
** Installers for ykman are now provided for Windows (amd64) and MacOS (universal2).
** Logging has been improved, and a new TRAFFIC level has been introduced.
** The codebase has been improved for scripting usage, either directly as a Python
    module, or via the new "ykman script" command.
    See doc/Scripting.adoc, doc/Library_Usage.adoc, and examples/ for more details.
** PIV: Add support for dotted-string OIDs when parsing RFC4514 strings.
** PIV: Drop support for signing certificates and CSRs with SHA-1.
** FIDO: Credential management commands have been improved to deal with ambiguity
    in certain cases.
** OATH: Access Keys ("remembered" passwords) are now stored in the system keyring.
** OpenPGP: Commands have been added to manage PINs.

* Version 4.0.9 (released 2022-06-17)
** Dependency: Add support for python-fido2 1.x
  ** Fix: Drop stated support for Click 6 as features from 7 are being used.

* Version 4.0.8 (released 2022-01-31)
** Bugfix: Fix error message for invalid modhex when programing a YubiOTP credential.
** Bugfix: Fix issue with displaying a Steam credential when it is the only account.
** Bugfix: Prevent installation of files in site-packages root.
** Bugfix: Fix cleanup logic in PIV for protected management key.
** Add support for token identifier when programming slot-based HOTP.
** Add support for programming NDEF in text mode.
** Dependency: Add support for Cryptography <= 38.

* Version 4.0.7 (released 2021-09-08)
** Bugfix release: Fix broken naming for "YubiKey 4", and a small OATH issue with
    touch Steam credentials.

* Version 4.0.6 (released 2021-09-08)
** Improve handling of YubiKey device reboots.
** More consistently mask PIN/password input in prompts.
** Support switching mode over CCID for YubiKey Edge.
** Run pkill from PATH instead of fixed location.

* Version 4.0.5 (released 2021-07-16)
** Bugfix: Fix PIV feature detection for some YubiKey NEO versions.
** Bugfix: Fix argument short form for --period when adding TOTP credentials.
** Bugfix: More strict validation for some arguments, resulting in better error messages.
** Bugfix: Correctly handle TOTP credentials using period != 30 AND touch_required.
** Bugfix: Fix prompting for access code in the otp settings command (now uses "-A -").

* Version 4.0.3 (released 2021-05-17)
** Add support for fido reset over NFC.
** Bugfix: The --touch argument to piv change-management-key was ignored.
** Bugfix: Don't prompt for password when importing PIV key/cert if file is invalid.
** Bugfix: Fix setting touch-eject/auto-eject for YubiKey 4 and NEO.
** Bugfix: Detect PKCS#12 format when outer sequence uses indefinite length.
** Dependency: Add support for Click 8.

* Version 4.0.2 (released 2021-04-12)
** Update device names.
** Add read_info output to the --diagnose command, and show exception types.
** Bugfix: Fix read_info for YubiKey Plus.

* Version 4.0.1 (released 2021-03-29)
** Add support for YK5-based FIPS YubiKeys.
** Bugfix: Fix OTP device enumeration on Win32.

* Version 4.0.0 (released 2021-03-02)
** Drop support for Python < 3.6.
** Drop reliance on libusb and libykpersonalize.
** Support the "fido" and "otp" subcommands over NFC (using the --reader flag)
** New "ykman --diagnose" command to aid in troubleshooting.
** New "ykman apdu" command for sending raw APDUs over the smart card interface.
** Restructuring of subcommands, with aliases for old versions (to be removed
    in a future release).
** Major changes to the underlying "library" code:
  *** New "yubikit" package added for custom development and advanced scripting.
  *** Type hints added for a large part of the "public" API.
** OpenPGP: Add support for KDF enabled YubiKeys.
** Static password: Add support for FR, IT, UK and BEPO keyboard layouts.

* Version 3.1.2 (released 2021-01-21)
** Bugfix release: Fix dependency on python-fido2 version.

(wiz)

doc: Updated security/py-fido2 to 1.1.0

(wiz)

py-fido2: update to 1.1.0.

NetBSD support by riastradh@, thanks!

Version 1.1.0 (released 2022-10-17)

    Bugfix: Fix name of "crossOrigin" in CollectedClientData.create().
    Bugfix: Some incorrect type hints in the MDS3 classes were fixed.
    Stricter checking of dataclass field types.
    Add support for JSON-serialization of WebAuthn data classes.
    This changes the objects dict representation to align with new additions in the
    WebAuthn specification. As this may break compatibility, the new behavior
    requires explicit opt-in until python-fido2 2.0 is released.
    Update server example to use JSON serialization.
    Server: Add support for passing RegistrationResponse/AuthenticationResponse (or
    their deserialized JSON data) to register_complete/authenticate_complete.
    Add new "hybrid" AuthenticatorTransport.
    Add new AuthenticatorData flags, and use 2-letter names as in the WebAuthn spec
    (long names are still available as aliases).

Version 1.0.0 (released 2022-06-08)

    First stable release.

Version 1.0.0rc1 (released 2022-05-02)

    Release Candidate 1 of first stable release.
    Require Python 3.7 or later.
    APIs have updated to align with WebAuthn level 2.
    Several CTAP 2.1 features have been implemented.

Version 0.9.3 (released 2021-11-09)

    Bugfix: Linux - Don't fail device discovery when hidraw doesn't support HIDIOCGRAWUNIQ (Linux kernels before 5.6).

Version 0.9.2 (released 2021-10-14)

    Support the latest Windows webauthn.h API (included in Windows 11).
    Add product name and serial number to HidDescriptors.
    Remove the need for the uhid-freebsd dependency on FreeBSD.

Version 0.9.1 (released 2021-02-03)

    Add new CTAP error codes and improve handling of unknown codes.

Version 0.9.0 (released 2021-01-20)

WARNING: Backwards-incompatible changes!

    Server: Attestation is now done in two parts (to align better with the spec):
    First, type-specific validation is done to provide a trust chain.
    Second, validation of the trust chain is done.
    Client: API changes to better support extensions.
        Fido2Client can be configured with Ctap2Extensions to support.
        Client.make_credential now returns a AuthenticatorAttestationResponse,
        which holds the AttestationObject and ClientData, as well as any client
        extension results for the credential.
        Client.get_assertion now returns an AssertionSelection object, which is
        used to select between multiple assertions, resulting in an
        AuthenticatorAssertionResponse, which holds the ClientData, assertion
        values, as well as any client extension results for the assertion.
    Renames: The CTAP1 and CTAP2 classes have been renamed to Ctap1 and Ctap2,
    respectively. The old names currently work, but will be removed in the
    future.
    ClientPin: The ClientPin API has been restructured to support multiple PIN
    protocols, UV tokens, and token permissions.
    CTAP 2.1 PRE: Several new features have been added for CTAP 2.1, including
    Credential Management, Bio Enrollment, Large Blobs, and Authenticator Config.
    HID: The platform specific HID code has been revamped and cleaned up.

(wiz)

py-keyring: add some dependencies, following its setup.cfg

Bump PKGREVISION.

(wiz)

doc: Added security/py-secretstorage version 3.3.1

(wiz)

security/Makefile: + py-secretstorage

(wiz)

security/py-secretstorage: import py-secretstorage-3.3.1

Packaged by bsiegert, K.I.A.Derouiche, and myself for wip.

This module provides a way for securely storing passwords and other secrets.

It uses D-Bus Secret Service API that is supported by GNOME Keyring
(since version 2.30) and KSecretsService.

SecretStorage supports most of the functions provided by Secret Service,
including creating and deleting items and collections, editing items,
locking and unlocking collections (asynchronous unlocking is also supported).

(wiz)

doc: Added devel/py-makefun version 1.15.1

(wiz)

devel/Makefile: + py-makefun

(wiz)

devel/py-makefun: import py-makefun-1.15.1

makefun helps you create functions dynamically, with the signature
of your choice. It was largely inspired by decorator and functools,
and created mainly to cover some of their limitations.

The typical use cases are:

* creating signature-preserving function wrappers - just like
functools.wraps but with accurate TypeError exception raising when
user-provided arguments are wrong, and with a very convenient way
to access argument values.

* creating function wrappers that have more or less arguments that
the function they wrap. A bit like functools.partial but a lot more
flexible and friendly for your users. For example, I use it in my
pytest plugins to add a requests parameter to users' tests or
fixtures when they do not already have it.

* more generally, creating functions with a signature derived from
a reference signature,

* or even creating functions with a signature completely defined
at runtime.

(wiz)

doc: Updated devel/fossil to 2.21

(js)

Update devel/fossil to 2.21

Changes for version 2.21 (2023-02-25)

  * Users can request a password reset. This feature is disabledby default. Use the new self-pw-reset property to enable it. New web pages /resetpw and /reqpwreset added.
  * Add the fossil repack command (together with fossil all repack) as a convenient way to optimize the size of one or all of the repositories on a system.
  * Add the ability to put text descriptions on ticket report formats.
  * Upgrade the test-find-pivot command to the merge-base command.
  * The /chat page can now embed fossil-rendered views of wiki/markdown/pikchr file attachments with the caveat that such embedding happens in an iframe and thus does not inherit styles and such from the containing browser window.
  * The fossil all remote subcommand added to "fossil all".
  * Passwords for remembered remote repositories are now stored as irreversible hashes rather than obscured clear-text, for improved security.
  * Add the "nossl" and "nocompress" options to CGI.
  * Update search infrastructure from FTS4 to FTS5.
  * Add the /deltachain page for debugging purposes.
  * Writes to the database are disabled by default if the HTTP request does not come from the same origin. This enhancement is a defense in depth measure only; it does not address any known vulnerabilities.
  * Improvements to automatic detection and mitigation of attacks from malicious robots.

(js)

devel/argp: Fix compilation

(js)

Add filesystems/glusterfs8 as the last 32 bit version

(js)

chat/matrix-synapse: Remove pinning for frozendict

(js)

Updated databases/sqlite3, databases/sqlite3-docs, databases/sqlite3-tcl, devel/lemon

(adam)

sqlite3: updated to 3.41.1

version 3.41.1 (2023-03-10):

Provide compile-time options -DHAVE_LOG2=0 and -DHAVE_LOG10=0 to enable SQLite to be compiled on systems that omit the standard library functions log2() and log10(), repectively.
Ensure that the datatype for column t1.x in "CREATE TABLE t1 AS SELECT CAST(7 AS INT) AS x;" continues to be INT and is not NUM, for historical compatibility.
Enhance PRAGMA integrity_check to detect when extra bytes appear at the end of an index record.
Fix various obscure bugs reported by the user community. See the timeline of changes for details.

(adam)

doc/TODO: + ImageMagick-7.1.1.3, a2ps-4.15, cmake-3.25.3.

(wiz)

doc: Updated textproc/R-markdown to 1.5

(mef)

(textproc/R-markdown) Updated 1.1 to 1.5

# CHANGES IN markdown VERSION 1.5

- Values of meta variables `title`, `author`, and `date` (if provided)
  will be transformed to the target output format before they are
  passed into templates.

- Fixed the bug that the default CSS was not added to HTML output.

- Removed dependency on the **mime** package.

- Added experimental support for HTML slides:
  `markdown::mark_html(..., meta = list(css = c('default', 'slides'),
  js = 'slides'))`. If you prefer knitting `Rmd` documents in RStudio,
  you may use the output format:

  ```yaml
  output:
    markdown::html_format:
      meta:
        css: [default, slides]
        js: [slides]
  ```

  See https://yihui.org/en/2023/01/minimal-r-markdown/ for a demo.

# CHANGES IN markdown VERSION 1.4

- Empty `\title{}` in LaTeX output will be removed (along with `\maketitle`).

- highlight.js is loaded from
  https://www.jsdelivr.com/package/gh/highlightjs/cdn-release by
  default now. This means more languages are supported (not only R),
  but also means syntax-highlighting will not work offline at the
  moment (it will be improved in future).

- MathJax failed to load in the previous version. The bug has been
  fixed now.

- Removed the function `markdownExtensions()`.

# CHANGES IN markdown VERSION 1.3

- Switched the underlying Markdown rendering engine from the C library
  **sundown** (which has been deprecated for a decade) to the R
  package **commonmark** (thanks, @jeroen, yihui/knitr#1329).

- The functions `renderMarkdown()` and `markdownToHTML()` have been
  renamed to `mark()` and `mark_html()`, respectively. The old names
  are still kept in this package for backward-compatibility.

- Removed the arguments `stylesheet` and `fragment.only` in
  `mark_html()`. For `stylesheet`, please use the argument `meta =
  list(css = ...)` to provide the CSS stylesheet. For `fragment.only`,
  please use `mark_html(template = FALSE)` or `mark_html(options =
  '-standalone')` instead of `fragment.only = TRUE`. Currently these
  old arguments are still accepted internally, but may be deprecated
  and dropped in the long run.

- The `file` argument of `mark()` and `mark_html()` can also take a
  character vector of Markdown text now.

- Removed functions `rendererExists()`, `rendererOutputType()`, and
  `registeredRenderer()`. They were primarily for internal use.

- Deprecated the function `markdownExtensions()`. All extensions
  should be specified via the `options` argument of functions like
  `mark()`, e.g., `mark(options = '+table+tasklist')`. See all options
  on the help page `?markdown::markdown_options`.

- Renamed `markdownHTMLOptions()` to `markdown_options()`.

# CHANGES IN markdown VERSION 1.2

- Fixed the warnings "a function declaration without a prototype is
  deprecated in all versions of C" (#94).

(mef)

doc: Updated textproc/R-stringr to 1.5.0

(mef)

(textproc/R-stringr) Updated 1.4.0 to 1.5.0

# stringr 1.5.0

## Breaking changes

* stringr functions now consistently implement the tidyverse recycling rules
  (#372). There are two main changes:

    *  Only vectors of length 1 are recycled. Previously, (e.g.)
      `str_detect(letters, c("x", "y"))` worked, but it now errors.

    *  `str_c()` ignores `NULLs`, rather than treating them as length 0
        vectors.

    Additionally, many more arguments now throw errors, rather than warnings,
    if supplied the wrong type of input.

* `regex()` and friends now generate class names with `stringr_` prefix (#384).

* `str_detect()`, `str_starts()`, `str_ends()` and `str_subset()` now error
  when used with either an empty string (`""`) or a `boundary()`. These
  operations didn't really make sense (`str_detect(x, "")` returned `TRUE`
  for all non-empty strings) and made it easy to make mistakes when programming.

## New features

* Many tweaks to the documentation to make it more useful and consistent.

* New `vignette("from-base")` by @sastoudt provides a comprehensive comparison
  between base R functions and their stringr equivalents. It's designed to
  help you move to stringr if you're already familiar with base R string
  functions (#266).

* New `str_escape()` escapes regular expression metacharacters, providing
  an alternative to `fixed()` if you want to compose a pattern from user
  supplied strings (#408).

* New `str_equal()` compares two character vectors using unicode rules,
  optionally ignoring case (#381).

* `str_extract()` can now optionally extract a capturing group instead of
  the complete match (#420).

* New `str_flatten_comma()` is a special case of `str_flatten()` designed for
  comma separated flattening and can correctly apply the Oxford commas
  when there are only two elements (#444).

* New `str_split_1()` is tailored for the special case of splitting up a single
  string (#409).

* New `str_split_i()` extract a single piece from a string (#278, @bfgray3).

* New `str_like()` allows the use of SQL wildcards (#280, @rjpat).

* New `str_rank()` to complete the set of order/rank/sort functions (#353).

* New `str_sub_all()` to extract multiple substrings from each string.

* New `str_unique()` is a wrapper around `stri_unique()` and returns unique
  string values in a character vector (#249, @seasmith).

* `str_view()` uses ANSI colouring rather than an HTML widget (#370). This
  works in more places and requires fewer dependencies. It includes a number
  of other small improvements:

    * It no longer requires a pattern so you can use it to display strings with
      special characters.
    * It highlights unusual whitespace characters.
    * It's vectorised over both string` and `pattern` (#407).
    * It defaults to displaying all matches, making `str_view_all()` redundant
      (and hence deprecated) (#455).

* New `str_width()` returns the display width of a string (#380).

* stringr is now licensed as MIT (#351).

## Minor improvements and bug fixes

* Better error message if you supply a non-string pattern (#378).

* A new data source for `sentences` has fixed many small errors.

* `str_extract()` and `str_exctract_all()` now work correctly when `pattern`
  is a `boundary()`.

* `str_flatten()` gains a `last` argument that optionally override the
  final separator (#377). It gains a `na.rm` argument to remove missing
  values (since it's a summary function) (#439).

* `str_pad()` gains `use_width` argument to control whether to use the total
  code point width or the number of code points as "width" of a string (#190).

* `str_replace()` and `str_replace_all()` can use standard tidyverse formula
  shorthand for `replacement` function (#331).

* `str_starts()` and `str_ends()` now correctly respect regex operator
  precedence (@carlganz).

* `str_wrap()` breaks only at whitespace by default; set
  `whitespace_only = FALSE` to return to the previous behaviour (#335, @rjpat).

* `word()` now returns all the sentence when using a negative `start` parameter
  that is greater or equal than the number of words. (@pdelboca, #245)

# stringr 1.4.1

Hot patch release to resolve R CMD check failures.

(mef)

doc: Updated www/R-httpuv to 1.6.9

(mef)

(mef)

doc: Updated graphics/R-rgl to 1.0.1

(mef)

(graphics/R-rgl) Updated 0.107.14 to 1.0.1

# rgl 1.0.1

## Major changes

* The long promised deprecations of the `rgl.*` functions
have happened.  Now deprecated: `rgl.abclines`,
`rgl.bbox`, `rgl.bg`, `rgl.clear`, `rgl.clipplanes`,
`rgl.close`, `rgl.light`, `rgl.lines`,
`rgl.linestrips`, `rgl.material`,  `rgl.open`,
`rgl.planes`, `rgl.points`, `rgl.quads`,
`rgl.select3d`, `rgl.set`, `rgl.setAxisCallback`,
`rgl.sprites`, `rgl.surface`, `rgl.texts`,
`rgl.triangles`, and `rgl.viewpoint`.
* A vignette "Deprecating the `rgl.*` interface"
has been added.
* Also deprecated: `elementId2Prefix`, `writeWebGL`

## Minor changes

* Since `rgl.material` is deprecated and no
longer contains the list of material types in its
argument list, `rgl.material.names` and `rgl.material.readonly` have been
added.
* Similarly, `rgl.par3d.names` and `rgl.par3d.readonly`
contain lists of properties that may be set or queried
in `par3d()`.
* The flexibility improvements for `surface3d()` in
0.111.6 were incomplete.
* Argument `flip` has been added to `surface3d()` to allow
front and back to be switched.

# rgl 0.111.6

## Minor changes

* Added a panning example to the help page for `setUserCallbacks()`.
* Replaced all calls to `sprintf` from C/C++ code with calls to
`snprintf`.
* `surface3d` and `rgl.surface` are now more flexible,
allowing any of the 3 coordinates to be a vector or matrix
as long as at least one is a matrix.
* `material3d` can now specify an `id` to query properties
for individual objects.
* Since `rgl.material` is soon to be deprecated and no
longer contain the list of material types in its
argument list, `rgl.material.names` and `rgl.material.readonly` have been
added.
* Similarly, `rgl.par3d.names` and `rgl.par3d.readonly`
contain lists of properties that may be set or queried
in `par3d()`.
* Made some examples conditional on interactive use
to save time on CRAN.

## Bug fixes

* Default mouse modes used when a window is opened by an `rgl.*`
call (which is not recommended!) now match
the defaults in `rgl::r3dDefaults`.
* Missing values could cause `surface3d()` to segfault.
* The C source code for `gl2psGetFileFormat` missed declaring
a prototype.

# rgl 0.110.2

## Major changes

* Material property `"blend"` has been added, to allow
various kinds of blending for semi-transparent objects
(issue #245).

## Minor changes

* The `Buffer` object now handles reading of sparse
accessors.
* Low level drawing of primitives has been made more
memory efficient.  This is only likely to make a
noticeable change with very large objects, where R
was running out of memory because of unnecessary
duplication. (Related to issue #260.)
* Recycling of x, y and z vectors in several functions
is more consistent.
* The `polygon3d()` function now chooses coordinates
automatically, as `triangulate()` does (PR #262.)
* The `mtext3d()` and related functions such as
`title3d()` now accept language objects
other than expressions, as `plotmath3d()` always has
(issue #273).

## Bug fixes

* The bounding box could be calculated incorrectly
if data all had large values (issue #250).
* Shiny displays failed to load the shaders (issue #249).
* `transform3d()` failed due to missing argument (issue #253).
* `readOBJ()` is now more flexible in what kinds of
separators it will accept. (issue #258).
* Failure to initialize could cause a segfault.
* On non-macOS platforms, gray-scale textures failed
to display, with a message about an invalid enumerant.
* The third coordinate for `adj` that was added in 0.108.3
was not rendered properly in `rglwidget()` displays of
text.  This sometimes caused text to disappear when it
was near the far limit of the display (issue #269).
* The X11 error fix in 0.109.6 could result in R
freezing in `Rcmdr`.
* Low level drawing functions are now more consistent
about returning an invisible `NULL` if asked to plot zero
items, rather than raising an error or crashing (issue #274).
* Calling `axis3d()` with no ticks or labels no longer triggers
an error, it now silently returns `NULL`.

# rgl  0.109.6

## Minor changes

* `rglwidget()` displays now act on "pointer" events,
not just "mouse" events, so they should be more usable
on touch screens and tablets (PR #240).

## Bug fixes

* Plotting `scene3d()` objects didn't handle suppressed
axes properly, drawing the default axis instead (issue
#241).
* On some systems using X11, `rgl` would segfault when
the "fixed" font was not found.
* X11 errors could cause R to abort.

# rgl  0.109.2

## Major changes

* Changes to support glTF animation:
  - Handling of `embedding = "modify"` for the model matrix
    has changed.  Now the centering step is only done for
    `embedding = "replace"`.  In addition, various bugs
    have been fixed.
  - If a subscene has no lights defined, the lights from the parent
    are used.
  - `plot.rglscene()` now ends with the root subscene as
    current.  It also allows specification of `open3d()`
    parameters in a list.
  - The `MATn` types in `Buffer` are returned as arrays with
    dim `c(n, n, count)`.
  - The `plot3d.rglscene` method now passes `...` to `open3d()`.
  - The `setUserShaders()` function now allows arrays of 4x4 matrices as "uniforms", and allows additional textures to be specified.
* `sprites3d()` now has the option of
`rotating = TRUE`, to allow 3D sprites to rotate with
the scene.
* Added `getShaders()` function to get shaders used in WebGL.
* Now detects if `rgl` is running within `reprex::reprex()`
and if so arranges that a screenshot will be included in the
output.
* Added default shaders to be used in `rglwidget()`, rather than
constructing them on the fly.  This incompatibly affects the use
of lights and clipping planes with user shaders:  their data
is now stored in arrays rather than multiple numbered variables.

## Minor changes

* Now that `pkgdown` 2.0.0 has been released, a number
of internal workarounds to support the development version
have been removed.
* Added `as.mesh3d()` methods for `"rglsubscene"` and `"rglscene"`.
* `open3d()` now handles `useNULL` and `silent` arguments
passed in `params`.
* Controls passed to `playwidget()` may now include a
component specifying HTML dependencies.
* Added `rglwidgetClass.readAccessor()` method to let other
code use the buffering.
* Changed the internal organization of bounding box calculations.
* All functions that produce meshes now accept
material properties.  Newly modified to do so using the `...`
argument:  `cylinder3d()`, and `getBoundary3d()`.
* Updated the system requirements and installation instructions.
* Solid bounding box decorations now try harder to display 3 faces (issue #206).
* Now that `webshot2` is on CRAN, instructions for
installing it from Github have been removed.
* Sometimes `webshot2` snapshots are very slow, so
the default for the `webshot` argument to `snapshot3d()`
now depends on the `RGL_USE_WEBSHOT` environment
variable, using `TRUE` if it is unset. (Reported by Prof. B. D. Ripley.)
* If the Chrome browser is not found, `snapshot3d(webshot = TRUE)` now issues a warning and
reverts to using `rgl.snapshot()`.
* Buffers now use "normalized integers" to store
color or texture coordinate values that lie between 0
and 1 when it saves some space.
* At the request of CRAN, the `akima` package is no
longer suggested.

## Bug fixes

* `as.mesh3d.rglobject()` didn't handle objects with indices
properly.
* In WebGL, the front vs back calculation sometimes
got the wrong result (issue #164).
* `pop3d(tag = x)` did not always find the objects with `tag == x` if they were not in the current subscene.
* The default values for `front` and `back` in `rgl.material`
and `material3d` are now `"filled"`, as documented in some
places.
* The `fog` setting wasn't handled properly by `bg3d()`.
* Numerous cases of partial argument matching were fixed
(suggestion of Henrik Bengtsson in issue #170.)
* Argument `col` is accepted as a synonym for `color` in `material3d()` and `rgl.material()`.
* `planes3d()` objects were not displayed consistently
in `rgl` windows and WebGL displays, because the bounding
boxes were not computed consistently (issue #169).
* Some initialization wasn't done properly in Shiny apps,
so they failed after a redraw (issue #173).
* Buffers are now optional, as they don't work with
Shiny scene changes (also issue #173).
* The NULL device would sometimes miscalculate the
bounding box.
* `selectpoints3d(closest = TRUE)` selected too many points
when multiple objects were in the scene.
* Clearing nested subscenes could cause a segfault and crash.
* In `knitr` and `rmarkdown`, blank plots could be shown
when `par3d(skipRedraw=TRUE)` was set (issue #188).
* Objects drawn with `sprites3d()` weren't lit correctly
in WebGL (issue #189).
* Objects with textures were sometimes drawn more than once, both
before the texture loaded and after.  This was most noticeable for
objects with user textures.
* Axis mode `"pretty"` got lost when scenes were redrawn.
* Tick labels were sometimes lost in WebGL displays and
`snapshot3d()` results (issue #197).
* The new material properties from 0.107.10 and 0.108.3
were not handled properly by `plotmath3d()`.
* `rglMouse()` did not set the default value of the drop-down
selector properly (issue #213).
* `merge.mesh3d()`, used by `filledContour3d()`, didn't handle
colors properly (issue #212).
* `bg3d(sphere = TRUE)` has been fixed (issue #207).
* Textures were not appearing on spheres, and front-back
differences weren't being rendered (issue #217).
* When "knitting" within RStudio under R 4.2.0 on
Windows, `rgl` scenes didn't appear (reported by
Dieter Menne.) A workaround has been added.
* In `rglwidget()`, axis labels were not always
displayed, and did not move with solid bounding box
decorations properly (issue #206).
* On some systems, `lines3d()` using both missing values
and transparency did not draw properly (issue #234,
originally reported by Gaspar Jekely).
* The `rglShared()` example failed when `crosstalk`
was uninstalled.

# rgl  0.108.3.2

## Bug fixes

* Changes introduced in 0.100.50 lacked checks; these caused
segfaults in Windows with R 4.2.0 and RStudio (issue #208).
* A typo caused problems loading fonts on some systems.

# rgl  0.108.3

## Major changes

* Added `getBoundary3d()` function to extract the boundary
edges of a mesh.
* Added material property `tag`, a string associated
with each object.  The value is reported by `ids3d(tags = TRUE)` and
may be used to select objects in most functions that use ids,
but otherwise is
largely ignored by `rgl`.  The `tagged3d()` function returns
information on tags.
* Primitive types (points, lines, segments, triangles, quads)
can now accept an `indices` parameter, similar to the
indices in `mesh3d` objects.
* Added `Buffer` object, based on glTF design, for holding binary
data for `rglwidget()`.

## Minor changes

* Allowed for a third coordinate in `text3d()`'s `adj`
parameter.
* Added support for `adj`, `pos` and `offset` to
`sprites3d()`.
* Added support for `pos` values of `0` (at specified
location), `5` (in front of it), and `6` (behind it) in
`text3d()`, `sprites3d()` and `plotmath3d()`.
* `crosstalk` is now a Suggested package, rather than
a required one.
* The `Makevars.ucrt` file has been modified with
contributions from Tomas Kalibera to work with his `winutf8`
build of R.
* `bgplot3d()` no longer pauses for each page when running
examples.
* `deldir` version 1.0-2 is incompatible with `rgl`.  Added
the `checkDeldir()` function to avoid running it.
* `shade3d()` treated texture coordinates like colors, and
  duplicated the first one for the whole face when `meshColor = "faces"` was chosen.
  Instead, they are now treated like vertex coordinates.
  (Reported by Michael Sumner in issue #145).
* Corrected the documentation and made the implementations
of `asHomogeneous()`, `asEuclidean()` etc. more consistent.
* An `as.rglscene()` generic has been added, though no methods
are defined in this package.
* `downlit` 0.4.0 has been released with support for `rgl`, so instructions
for installing the devel version have been removed.

## Bug fixes

* Fixed rendering of text as sprites3d() objects.
* Added `--static` flag to configure script for FreeType
  installation.  (Suggestion of Simon Urbanek and Prof. Brian Ripley.)
* `shade3d()`, `wire3d()` and `dots3d()` overrode
  `"front"` and `"back"` material settings in mesh objects.
* `rglwidget()` handling of bounding box decorations had
  several bugs.
* `rgl` could not find routines in the DLL on some Windows
installs (Issue 148.)
* Some cases where allocations were not protected have been fixed.

(mef)

doc: Updated www/R-htmlwidgets to 1.6.1

(mef)

(www/R-htmlwidgets) Updated 1.5.4 to 1.6.1

htmlwidgets 1.6.1
------------------------------------------------------

### Bug fixes

* Closed #456: Fixed an issue where widgets were no longer being
  resized properly when rendered in a standalone fashion. (#458)

htmlwidgets 1.6.0
-------------------------------------------------------

### Potentially breaking changes

* `shinyWidgetOutput()` and `sizingPolicy()` both gain a new `fill`
  parameter. When `TRUE` (the default), the widget's container element
  is allowed to grow/shrink to fit it's parent container so long as
  that parent is opinionated about its height and has been marked with
  `htmltools::bindFillRole(x, container = TRUE)`. (#442)

  * The primary motivation for this is to allow widgets to grow/shrink
    by default [inside
    `bslib::card_body_fill()`](https://rstudio.github.io/bslib/articles/cards.html#responsive-sizing)

  * Widgets that aren't designed to fill their container in this way
    should consider setting `sizingPolicy(fill =
    FALSE)`/`shinyWidgetOutput(fill = FALSE)` and/or allowing users to
    customize these settings (i.e., add a `fill` argument to the
    `customWidgetOutput()` function signature).

* `shinyWidgetOutput()`'s `reportSize` argument now defaults to
  `TRUE`. This way, calling `shiny::getCurrentOutputInfo()` inside a
  `shinyRenderWidget()` context will report the current height and
  width of the widget.

### Improvements

* Closed #433 and #440: `saveWidget(selfcontained=TRUE)` now uses the
  `{rmarkdown}` package to discover and call pandoc, which fixes a
  couple existing issues and helps "future proof" this code path from
  future changes to pandoc.

* Closed #257 and #358: `saveWidget(selfcontained=TRUE)` now correctly
  prevents HTML from being interpreted as markdown. (#401)

(mef)

doc: Updated textproc/R-htmltools to 0.5.4

(mef)

(textproc/R-htmltools) Updated 0.5.2 to 0.5 4

# htmltools 0.5.4

## New Features

* Added  a  new  `bindFillRole()`  function  for  modifying  `tag()`
  object(s) into tags  that are allowed to grow and  shrink when their
  parent is  opinionated about  their height.  See `help(bindFillRole,
  "htmltools")`  for  documentation  and examples.  Note  the  primary
  motivation for  adding these functions  is to power  `{bslib}`'s new
  `card()`        API        (in        particular,        [responsive
  sizing](https://rstudio.github.io/bslib/articles/cards.html#responsive-sizing))
  as  well  as  the  new `fill`  arguments  in  `shiny::plotOutput()`,
  `shiny::imageOutput()`,                        `shiny::uiOutput()`,
  `htmlwidgets::sizingPolicy()`,                                  and
  `htmlwidgets::shinyWidgetOutput()`. (#343)

## Bug fixes

* Closed #331: `copyDependencyToDir()` creates `outputDir`
  recursively, which happens in Quarto or when `lib_dir` points to a
  nested directory. (@gadenbuie, #332)

* Closed #346: `tagQuery()`'s `$remove()`, `$after()`, `$before()`,
  `$replaceWith()` had a bug that prevented expected behavior when
  sibling children values where not tag elements. (#348)

# htmltools 0.5.3

## Breaking changes

* Closed #305: `htmlPreserve()` no longer uses _inline_ code blocks
  for Pandoc's raw attribute feature when used inside a _non_-inline
  knitr/rmarkdown code chunk, and as a result, in this case, an
  additional `<p>` tag is no longer wrapped around the HTML
  content. (#306)

## Bug fixes

* Closed #301: `tagQuery()` was failing to copy all `tagList()` html
  dependencies within nest child tag lists. `tagQuery()` will now
  relocate html dependencies as child objects. (#302)

* Closed #290: htmltools previously did not specify which version of
  fastmap to use, and would fail to install with an old version of
  fastmap. (#291)

* `copyDependencyToDir()` no longer creates empty directories for
  dependencies that do not have any files. (@gadenbuie, #276)

* Closed #320: `copyDependencyToDir()` now works with dependencies
  with specified attributes. (@dmurdoch, #321)

(mef)

doc: Updated math/R-compositions to 2.0.5

(mef)

(math/R-compositions) Updated 2.0.4 to 2.0.5

Version 2.0-5
  * Adapted to new rgl 1.0.1 interface

(mef)

py-markups: add a comment noting 4.0.0 drops Python < 3.9

(gutteridge)

doc: Updated devel/py-looseversion to 1.1.2

(gutteridge)

py-looseversion: update to 1.1.2

### 1.1.2 (22 Feb 2023)

- 2023.02.22
  - Revert unintended change in internal version representation

### 1.1.1 (19 Feb 2023)

- 2023.02.19
  - Restructure package so stubs get installed and detected

### 1.1.0 (19 Feb 2023)

- 2023.02.19
  - Add type annotations and stubs.

(gutteridge)

py-music21: s/BUILD_DEPENDS/TOOL_DEPENDS/

(gutteridge)

Drop a few dead XEmacs master site mirrors.

(hauke)

gpsd: fix build with ruby 3.2

(wiz)

doc: Updated misc/wthrr to 0.10.0

(pin)

misc/wthrr: update to 0.10.0

What's Changed
- feat: weekday forcasts #75
- fix: time_indicator disable option #76
- fix: daytime icons in graph #77
- fix border displacement with other precip. units than probability 0473b20
- feat: historical weather #78

(pin)

doc: Updated sysutils/ripdrag to 0.3.0

(pin)

sysutils/ripdrag: update to 0.3.0

- Implement DropTarget the (almost)right way

(pin)

doc: Updated mail/nmh to 1.8

(leot)

nmh: Update to 1.8

Changes:
===
1.8
===
Welcome to nmh, the new version of the classic MH mail handling system.
It's been nearly five years since the last release of nmh, and there have
been a number of significant changes since that last release.  Long-time
MH and nmh users should read carefully the NOTABLE CHANGES section, as
there are some significant changes to nmh behavior.  Otherwise, please
see the README and INSTALL files for help on getting started with nmh.

This release is dedicated to Norman Z. Shapiro, co-designer of the MH
Message Handling System.  MH is the predecessor of nmh.  Norm was an
active supporter of nmh development until he passed away in October of
2021.  We are most grateful to Norm for his stewardship of MH and nmh.
    https://en.wikipedia.org/wiki/Norman_Shapiro

For news of future releases, subscribe to the low-volume
    https://lists.nongnu.org/mailman/listinfo/nmh-announce

---------------
NOTABLE CHANGES
---------------

- Support for Content-MD5 header fields, MIME content cache functionality,
  and the message/partial MIME type have been removed.
- Gmail OAuth2/XOAUTH support for desktop applications has been effectively
  dropped, so nmh no longer supports it.  nmh support for Gmail API access
  is experimental, please post to nmh-workers@nongnu.org if you'd like to
  help with test and development.
- repl(1) -convertargs now allows editing of the composition draft between
  translation and any encoding of text content.  Because encoding can wrap
  long lines, the use of a paragraph formatter has been removed from
  mhn.defaults.

------------
NEW FEATURES
------------

- The default editor has been changed from 'vi' to 'prompter', to align with
  historical practice and reduce packaging dependencies on external programs.
- A new -checkbase64 switch has been added to mhfixmsg(1).
- inc(1)/msgchk(1) now support STARTTLS for the POP protocol.
- All TLS-supported protocols now will send the SNI (server name indicator)
  TLS extension.
- A new mh-format function %(ordinal) has been implemented to output the
  appropriate ordinal suffix for numbers.  Example: "%(num 22)%(ordinal)"
  will output "22nd".
- show and mhl now decode more addresses in header fields.
- Added warning from all programs that read the profile if the profile
  contains a post entry, which is ignored, but does not contain a
  postproc entry.  In other words, if you get this warning and want
  to suppress it, your options include:
  1) Remove your post profile entry.
  2) Make your post profile entry a comment by prepending it with the #:
    comment indicator.
  3) Add a postproc entry that points to the post that you use.  That can
    be viewed with "mhparam postproc".
- scan(1) -file argument can be a Maildir directory.
- Updated mhn.defaults to prefer mpv(1) over xv(1) and replace mpeg_play(1),
  and to use it for all video types, not just video/mpeg.  And prefer all
  other searched-for pdf viewers over acroread(1).
- Added mhshow-suffix-video.mp4 to mhn.defaults, for use by mhshow(1) and
  send(1).
- Removed support from mhn.defaults for application/x-ivs and text/richtext.
- Changed interpretation of argument to mhfixmsg(1) -decodeheaderfieldbodies
  switch to specify character set of the decoded field bodies.
- repl(1) -convertargs now allows editing of the composition draft between
  translation and any encoding of text content.
- install-mh(1) now enables the mh-draft(5) draft folder facility.

-----------------
OBSOLETE FEATURES
-----------------

- The generation and verification of a Content-MD5 field has been removed
  without deprecation.  The related -check and -nocheck options now error.
- The MIME content cache functionality has been mostly non-functional since
  the development on nmh, and consequently all of the content caching code
  and related switches (-cache/-rcache/-wcache) have been removed.
- Support for generating and reassembling message/partial messages has been
  removed; it seems that this has been broken since 1.5 and there is very
  little support across MUAs.
- Marked Gmail OAuth2/XOAUTH support as being unsupported.
- Support for the MHPDEBUG environment variable was removed.  It was
  deprecated in nmh 1.7.  The pick(1) -debug switch replaced it.
- The 'libdir' mhparam(1) component was removed.  It was deprecated in
  nmh 1.7, when it was replaced by a new 'libexecdir' component.

---------
BUG FIXES
---------

- Fixed bcc to work with sendmail/pipe, and better documented that dcc
  doesn't work with it [Bug 55700].
- An -attendee switch has been added to mhical(1), for use when more than one
  (or zero) attendees match a user's mailbox.
- Fixed inc(1) and %(me) function escape to not obey Local-Mailbox profile
  component.
- Fixed source charset in mhfixmsg textcharset verbose output.
- Fixed mhfixmsg charset determination of content added with -reformat.
- Fixed file descriptor leak in mhfixmsg when run on multiple input files.
- Fixed mhfixmsg(1) -decodeheaderfilebodies to support mixed encoded/undecoded.
- Fixed memory corruption in post(1) and inc(1) when using XOAUTH2,
  with 4 or more entries in the oauth-authservice file.
- Added alias expansion to From: address for use by sendfrom.
- Removed extra space added before header field bodies by dist(1) to $mhdraft.
- Fixed display of iCalendar object with multiple VEVENTS using a VTIMEZONE.
- Fixed allowable encodings with MIME message types to get closer to RFC 2046.
- Detect other files regardless of backup prefix [Bug #49476].
- Copy if hard link by refile(1) or send(1) fails with EACESS [Bug 56575].

(leot)

doc: Updated databases/mariadb106-client to 10.6.12

(nia)

mariadb106: update to 10.6.12

Notable Items

  InnoDB

    * Full-text index corruption with system versioning (MDEV-25004)
    * innodb_undo_log_truncate=ON recovery and backup fixes (MDEV-29999,
      MDEV-30179, MDEV-30438)
    * Upgrade after a crash is not supported (MDEV-24412)
    * Remove InnoDB buffer pool load throttling (MDEV-25417)
    * InnoDB shutdown hangs when the change buffer is corrupted (MDEV-30009)
    * innodb_fast_shutdown=0 fails to report change buffer merge progress
      (MDEV-29984)
    * mariadb-backup --backup --incremental --throttle=... hangs
      (MDEV-29896)
    * Crash after recovery, with InnoDB: Tried to read (MDEV-30132)
    * Trying to write ... bytes at ... outside the bounds (MDEV-30069)
    * TRUNCATE breaks FOREIGN KEY locking (MDEV-29504, MDEV-29849)
    * INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION.NAME is NULL for undo
      tablespaces (MDEV-30119)
    * Fixed hangs and error handling in B-tree operations (MDEV-29603,
      MDEV-30400)

  Galera

    * Fixes for cluster wide write conflict resolving (MDEV-29684)

  Replication

    * Parallel slave applying in binlog order is corrected for admin class
      of commands including ANALYZE (MDEV-30323)
    * Seconds_Behind_Master is showed now more precisely at the slave
      applier start, including in the delayed mode (MDEV-29639)
    * mysqlbinlog --verbose is made to show the type of compressed columns
      (MDEV-25277)
    * Deadlock is resolved on replica involving BACKUP STAGE BLOCK_COMMIT
      and a committing user XA (MDEV-30423)

  JSON

    * JSON_PRETTY added as an alias for JSON_DETAILED (MDEV-19160)

  General

    * Infinite sequence of recursive calls when processing embedded CTE
      (MDEV-30248)
    * Crash with a query containing nested WINDOW clauses (MDEV-30052)
    * Major performance regression with 10.6.11 (MDEV-29988)

(nia)

doc: Updated misc/s6-portable-utils to 2.3.0.1

(schmonz)

Update to 2.3.0.1. From the changelog:

- s6-test removed.
- New multicall binary: s6-portable-utils.
- Bugfixes.

(schmonz)

doc: Updated net/s6-networking to 2.5.1.3

(schmonz)

Update to 2.5.1.3. From the changelog:

- Bugfixes.

(schmonz)

doc: Updated sysutils/s6 to 2.11.3.0

(schmonz)

Update to 2.11.3.0. From the changelog:

- New s6-svc -Q command.
- -0167 options to s6-ioconnect are now deprecated.
- Internal changes to instances and servicedir reserved names.
- Bugfixes.

(schmonz)

doc: Updated lang/execline to 2.9.2.1

(schmonz)

Update to 2.9.2.1. From the changelog:

- New multicall binary: execline.
- Bugfixes.

(schmonz)

doc: Updated devel/skalibs to 2.13.1.0

(schmonz)

Update to 2.13.1.0. From the changelog:

- New macro: SKALIBS_PATH_MAX, in posixishard.h.
  Defined to 4096 on systems that do not define PATH_MAX.
- New function: sals().

(schmonz)

doc: Updated devel/mob to 4.3.1

(schmonz)

Update to 4.3.1. From the changelog:

- Fix: `mob done` will now print the command `git merge --squash --ff
  mob/main` just once and not twice.

(schmonz)

doc: Updated math/R-classInt to 0.4.9

(mef)

(math/R-classInt) Updated 0.4.8 to 0.4.9

## Version 0.4-9

- #41 issues. The maximum and minimum breaks are set to \code{+Inf}
  and \code{-Inf} to avoid errors induced in the earlier version
  where breaks could cease to be strictly ascending. The
  \code{legacy=} argument with value \code{TRUE} may be used to
  revert to the previous behaviour.

(mef)

gstreamer1: add PLIST.Darwin

(tnn)

doc: Updated devel/R-mockr to 0.2.1

(mef)

(devel/R-mockr) Updated 0.2.0 to 0.2.1

# mockr 0.2.1 (2023-01-30)

## Bug fixes

- More careful querying of functions to be mocked, to avoid errors for `.onLoad()` when testing interactively (#29).

## Chore

- Change maintainer e-mail address.

(mef)

doc: Updated math/R-igraph to 1.4.1

(mef)

(math/R-igraph) Adding TEST_DEPENDS+= , but still missing more

(mef)

(devel/R-igraph) Updated 1.13.1 to 1.14.1

# igraph 1.4.1

## Bug fixes

- `console()` now works again and provides a Tcl/Tk based UI where igraph can post status messages and progress info (#664).

- Fix errors when printing long vertex names (#677, @ahmohamed).

- Fix regression that broke builds on some systems (e.g., GCC version 5 or earlier), introduced in igraph 1.4.0 (#670, #671).

- `fit_hrg()` does not crash any more when called with a graph that has less than three vertices.

## Documentation

- Various improvements (#663, @maelle; #667).

## Internal

- Fix warning about `yyget_leng()` returning wrong type when using LTO (#676).

- Don't mention C++11 or C++17 for best compatibility with both newest R and older compilers, while still requesting a C++ compiler for linking.

- Don't ignore `build/` when building the package because the vignette index is built there.

- Skip plot test entirely on R-devel.

- Avoid submodules for building igraph (#674).

- Makevars cleanup (#671).

- Add Zenodo configuration file.

# igraph 1.4.0

## Breaking changes

- Breaking change: Allow change of attribute type when setting attribute for all vertices or edges; only attributes of length 1 or the length of the target set allowed (#633).

## Added

- `tkplot()` gained a `palette` argument and it is now using the same palette as `plot()` by default, for sake of consistency.
- `plot.igraph()` gained a `loop.size` argument that can be used to scale the common radius of the loop edges.

## Fixed

- The default maximum number of iterations for ARPACK has been increased to 3000 to match that of the igraph C core.
- Rare convergence problems have been corrected in `cluster_leading_eigen()`.
- All ARPACK-based functions now respect random seeds set in R when generating a random starting vector.
- `igraph_version()` returned an invalid value in 1.3.4, this is now corrected.
- The value of `par(xpd=...)` is now restored after plotting a graph.
- Fixed a bug in `as.dendrogram.communities()` for large dendrograms, thanks
  to @pkharchenko (see PR #292).
- Fixed two bugs in `graph_from_incidence_matrix()` that prevented the creation of directed graphs with `mode="all"` from dense or sparse matrices.
- `dfs()` accidentally returned zero-based root vertex indices in the result object; this is now fixed and the indices are now 1-based.
- `as_graphnel()` does not duplicate loop edges any more.
- `convex_hull()` now returns the vertices of the convex hull with 1-based indexing.
- Some `rgl.*()` function calls in the codebase were replaced with equivalent `*3d()` function calls in preparation for upcoming deprecations in `rgl` (see PR #619)
- `plot.igraph()` does not use the `frame=...` partial argument any more when calling `plot.default()`. The default `NULL` value of `frame.plot` is now also handled correctly.
- `hub_score()` and `authority_score()` considered self-loops only once on the diagonal of the adjacency matrix of undirected graphs, thus the result was not identical to that obtained by `eigen_centrality()` on loopy undirected graphs. This is now corrected.
- `distances()` no longer ignores the `mode` parameter when `algorithm='johnson'`.

## Deprecated

- `automorphisms()` was renamed to `count_automorphisms()`; the old name is still available, but it is deprecated.

## Other

- Documentation improvements.
- The Github repository was now moved to a single-branch setup where the package can be built from the `main` branch directly.
- Added igraph extended tutorial as an R vignette (#587).
- igraph now has a homepage based on `pkgdown` thanks to @maelle (see #645). This will eventually become the official homepage.

# igraph 1.3.5

Added:

- `mark.groups=...` argument of `plot.igraph()` now accepts `communities`
  objects

Fixed:

- Negative degree exponents are not allowed any more in `sample_pa()` and
  `sample_aging_pa()`.

- Package updated to be compatible with Matrix 1.5.

Other:

- Documentation improvements and fixes.

# igraph 1.3.4

Added:

- `sample_asym_pref()` now returns the generated types of the vertices in the
  vertex attributes named `outtype` and `intype`.

Fixed:

- `layout_nicely()` does not recurse infinitely any more if it is assigned to
  the `layout` attribute of a graph

- `layout_nicely()` now ignores edge weights when there are non-positive
  edge weights. This is needed because igraph 1.3.3 started validating edge
  weights in `layout_with_fr()` and `layout_with_drl()`, resulting in
  errors when `layout_nicely()` was used on weighted graphs with negative
  weights. Since `layout_nicely()` is the default layout algorithm for
  `plot()`, most users were not even aware that they were using the FR or
  DrL layouts behind the scenes. Now the policy is that `layout_nicely()`
  attempts to get the job done without errors if possible, even if that means
  that edge weights must be ignored. A warning is printed if this is the case.

# igraph 1.3.3

Added:

- `reverse_edges()` reverses specific or all edges in a graph.

- Single-bracket indexing of `V()` and `E()` resolves attribute names in the
  indexing expressions by default (for instance, `E(g)[weight > x]` matches
  edges with a weight larger than a threshold). This can be problematic if the
  attribute masks one of the variables in the local evaluation context.
  We now have a pronoun called `.env` (similarly to `rlang::.env`) that
  allows you to force attribute name lookup to the calling environment. For
  sake of completeness, we also provide `.data` (similarly to `rlang::.data`)
  to force attribute name lookup to the vertex / edge attributes only. These
  pronouns are automatically injected into the environment where the indexing
  expression is evaluated.

Deprecated:

- Names of functions that can be used inside a `V()` or `E()` indexing start
  with a dot since igraph 1.1.1; however, the old dotless names did not print
  a deprecation warning so this may have gone unnoticed for years. We are
  introducting a deprecation warning for `nei()`, `innei()`, `outnei()`,
  `inc()`, `from()` and `to()` inside single-bracket indexing of vertex and
  edge sequences and will remove the old variants soon.

# igraph 1.3.2

The C core is updated to 0.9.9, fixing a range of bugs.

Fixed:

- The length of size-zero `communities` objects is now reported correctly.
- `layout_with_kk()` would fail to produce reasonable results with the default initial
  coordinates. This has been corrected, however, this function no longer produces
  precisely the same output for a given graph as before. To restore the previous
  behaviour, use `layout_with_kk(g, coord=layout_in_circle(g))` in 2D or
  `layout_with_kk(g, dim=3, coord=layout_on_sphere(g))` in 3D.
- Indexing an `igraph.vs` object with `v[x, na_ok=T]` now correctly handles the
  `na_ok` argument in all cases; previous versions ignored it when `x` was
  a single number.

Other:

- Documentation improvements and fixes.

(mef)

doc: Updated math/R-nimble to 0.13.1

(mef)

(math/R-nimble)  Updated 0.12.2 to 0.13.1

(From: nimble/inst/NEWS.md)
#                            CHANGES IN VERSION 0.13.1 (December 2022)

## BUG FIXES

- Fix bug in MCMC sampler inclusion/exclusion of predictive nodes from
  target node dependencies (PR #1248).

#                            CHANGES IN VERSION 0.13.0 (November 2022)

## USER LEVEL CHANGES

- Thoroughly revamp handling of predictive nodes in MCMC sampling. If MCMC
results identical to previous versions of NIMBLE are needed in models with
posterior predictive nodes, set
`nimbleOptions(MCMCusePredictiveDependenciesInCalculations = TRUE)`
and `nimbleOptions(MCMCorderPosteriorPredictiveSamplersLast = FALSE)`.

    -- MCMC samplers, by default, will now exclude predictive dependencies
    from internal sampler calculations.  This can be reverted to the old behavior
    of including predictive dependencies in calculations using
    `nimbleOptions(MCMCusePredictiveDependenciesInCalculations = TRUE)`.

    -- At the time of `buildMCMC`, all `posterior_predictive` samplers are
    automatically reordered to operate last among all samplers. Doing so,
    posterior predictive samples are generated conditional on the other values
    in the MCMC sample.  This reordering can be disabled using
    `nimbleOptions(MCMCorderPosteriorPredictiveSamplersLast = FALSE)`
    (but doing so without also setting
    `nimbleOptions(MCMCusePredictiveDependenciesInCalculations = TRUE)` could
    result in samples that are invalid in terms of the joint posterior
    distribution (but with valid samples marginally).

    -- Removal of the `posterior_predictive_branch` sampler.  Filling the same
    role, the `posterior_predictive` sampler now updates all nodes downstream
    of its `target` node.  Assignment of the `posterior_predictive` sampler
    happens automatically during MCMC configuration, unless
    `nimbleOptions(MCMCusePosteriorPredictiveSampler = FALSE)`.

    -- Automatic determination of "predictive" model nodes, which are all
    stochastic non-data nodes that have no data nodes anywhere in their
    downstream dependencies. Tracking of predictive nodes is done
    automatically, but maybe be disabled using
    `nimbleOptions(determinePredictiveNodesInModel = FALSE)`.

    -- New arguments `includePredictive` (default value `TRUE`) and
    `predictiveOnly` (default value `FALSE`), for both the `getNodeNames`
    and the `getDependencies` methods of model objects.  These specify whether
    any predictive nodes are included in the results, and whether only
    predictive nodes are included, respectively.

    -- The MCMC configuration object will issue a warning message if there are
    stochastic non-data nodes which will not undergo MCMC sampling.  This
    warning can be disabled using
    `nimbleOptions(MCMCwarnUnsampledStochasticNodes = FALSE)`.

- Add option to WAIC system (via `controlWAIC`) to allow additional burnin (in
addition to standard MCMC burnin) before calculating online WAIC, thereby
allowing inspection of initial samples without forcing them to be used for WAIC
(PR #1244).

- For MCMC configuration `addSampler` method, change name of the
`scalarComponents` argument to `expandComponents` (PR #1215).

- Add new `default` argument for the `addSampler` method of MCMC configuration
objects.  When `default = TRUE`, default samplers (conjugate, or otherwise) will
be added to the specified nodes.  The addition of this argument provides an
entry point to the logic of default sampler determination and assignment,
without creating a new MCMC configuration object (PR #1215).

- Add new `nodes` argument for the `addSampler` method of MCMC configuration
objects.  Nodes specified in `nodes` automatically undergo expansion according
to `expandNodeNames` prior to sampler assignment, allowing for easier assignment
of samplers to multiple nodes (PR #1215).

- `rcar_normal` issues an informative error message when invoked from the R
command line (PR #1243).

- Warn users of unused constants during model building (PR #1242).

- Add `replaceSamplers` method to MCMC configuration objects to simplify
modifying how a node is sampled (PR #1222).

- Convert `NEWS` to Markdown format for proper rendering in browser
(issue #1231).

- Indicate model code that produces warnings about unknown nimbleFunctions
(issue #370).

## BUG FIXES

- Avoid error occurring when a model variable name starts with "logProb"
(PR #1240).

- Avoid error occurring when a model variable is named "i" (PR #1239).

- Prevent infinite recursion in particular cases in conjugacy checking
(PR #1228).

- Fix bug in simulating from `dcar_normal` nodes when multiple nodes passed to
simulate (issue #1238).

- Fix error message about duplicate node declarations (PR #1233).

- Fix another issue with long variable names (PR #1217).

- Fix warning related to `dataNodes` in WAIC.

## DEVELOPER LEVEL CHANGES

- Remove use of bitwise `|` and `&` operators in C++ code, per CRAN request.

- Refactor `nimbleMCMC` to pull out model creation (PR #1223).

- Fix an issue with nested `nimbleList`s on MacOS (PR #1213).

(mef)

doc: Added security/py-otp version 2.8.0

(wiz)

security/Makefile: + py-otp

(wiz)

security/py-otp: import py-otp-2.8.0

Packaged for wip by K.I.A.Derouiche and myself.

PyOTP is a Python library for generating and verifying
one-time passwords. It can be used to implement two-factor
(2FA) or multi-factor (MFA) authentication methods in web
applications and in other systems that require users to log in.

(wiz)

doc/TODO: + py-fido2-1.1.0, ykman-5.0.1.

(wiz)

CHANGES-2022: add some missing entries

(wiz)

doc: Updated audio/gospt to 0.0.46

(pin)