## 3.1.5
- Fixed Python 2/3 incompatibility with `itertools.izip_longest()`.
## 3.1.4
- Added `BigAutoField` to support 64-bit auto-incrementing primary
keys.
- Use Peewee-compatible datetime serialization when exporting JSON
from a `DataSet`. Previously the JSON export used ISO-8601 by default.
- Added `Database.batch_commit` helper to wrap iterators in chunked
transactions.
## 3.1.3
- Fixed issue where scope-specific settings were being updated
in-place instead of copied.
- Fixed bug where setting a `ForeignKeyField` did not add it to the
model's "dirty" fields list.
- Use pre-fetched data when using `prefetch()` with `ManyToManyField`.
- Use `JSON` data-type for SQLite `JSONField` instances.
- Add a `json_contains` function for use with SQLite `json1` extension.
- Various documentation updates and additions.
ssl:
- Added new API functions to facilitate cipher suite
handling
erts, observer:
- More crash dump info such as: process binary virtual
heap stats, full info for process causing out-of-mem
during GC, more port related info, and dirty scheduler
info.
inets:
- Add support for unix domain sockets in the http client.
changes in version 1.28:
* The formerly internal yat2m tool is now installed for a native
build.
* The new files gpgrt.m4 and gpgrt-config are now installed. They
can be used instead of gpg-error.m4 and gpg-error-config.
* New logging functions similar to those used by GnuPG.
* New helper functions for platform abstraction.
I'm pleased to announce the 18.0.1 release of xf86-video-ati, the Xorg
driver for ATI/AMD Radeon GPUs supported by the radeon kernel driver.
This release supports xserver versions 1.13-1.19. It also works with
xserver 1.20 RC1, so unless something unexpected happens, it should work
with xserver 1.20 as well.
This is a bug-fix release addressing issues in 18.0.0. While those
issues shouldn't affect most users, I recommend that all users of 18.0.0
update to 18.0.1.
* The Xorg process could crash when multiple primary screens are
configured in xorg.conf.
* TearFree could trigger debugging messages in the pixman library
I'm pleased to announce the 18.0.1 release of xf86-video-amdgpu, the
Xorg driver for AMD Radeon GPUs supported by the amdgpu kernel driver.
This release supports xserver versions 1.13-1.19. It also works with
xserver 1.20 RC1, so unless something unexpected happens, it should work
with xserver 1.20 as well.
This is a bug-fix release addressing issues in 18.0.0. While those
issues shouldn't affect most users, I recommend that all users of 18.0.0
update to 18.0.1.
* The Xorg process could enter an infinite loop after a server reset (in
configurations where Xorg doesn't terminate when the last client
disconnects)
* The Xorg process could crash when multiple primary screens are
configured in xorg.conf.
* TearFree could trigger debugging messages in the pixman library
This is to reflect the behaviour documented in netpgp(1).
Originally submitted on tech-pkg@ as:
[PATCH 09/11] Output signatures to the standard output for "-"
Only modified for consistency with the coding style; as also applied in
NetBSD's src repository.
Originally submitted on tech-pkg@ as:
[PATCH 07/11] Correct option "--armor"
[PATCH 08/11] Also document alternate option "--detach"
As also applied in NetBSD's src repository.
Originally submitted on tech-pkg@ as:
[PATCH 04/11] Do not use random data for pass-phrases on EOF
Only modified for consistency with the coding style; as also applied in
NetBSD's src repository.
Tested on NetBSD/amd64.
This also fixes a crash when the pass-phrase entered is empty.
Originally submitted on tech-pkg@ as:
[PATCH 02/11] Do not truncate pass-phrases without a newline character
Only modified for consistency with the coding style; as also applied in
NetBSD's src repository.
Tested on NetBSD/amd64.
Originally submitted on tech-pkg@ as:
[PATCH 06/11] Do not ask for a passphrase when empty
Only modified for consistency with the coding style; as also applied in
NetBSD's src repository.
Tested on NetBSD/amd64.
1.1.0:
+ Changed from distutils to setuptools because it's the future
+ Implement RFC 7601 SHOULD to ignore unknown method identifiers (2.7.6):
+ Discard unknown ptypes and associated properties
+ Added tests to document errors raised by different kinds of broken header
fields
Changes:
4.8
---
Features:
- Add key bindings for moving to the word start or end: Mod1+b/^Left and
Mod1+f/^Right.
- Set class name on menu window (WM_CLASS).
Fixes:
- Fix a possible free of a uninitialize variable in paste().
Changes:
0.8
---
There are many features and code that has been rewritten. Particulary
noteworthy is now a more clear separation of the X11/Xft code and the terminal
emulator logic.
Some peole have shown interest in writing a Wayland port, writing specific
(fuzz) tests for the terminal emulator or reusing the terminal emulator code.
Pkgsrc changes:
* Add libunbound.pc to PLIST.
Upstream changes:
Features
- auth-zone provides a way to configure RFC7706 from unbound.conf,
eg. with auth-zone: name: "." for-downstream: no for-upstream: yes
fallback-enabled: yes and masters or a zonefile with data.
- Aggressive use of NSEC implementation. Use cached NSEC records to
generate NXDOMAIN, NODATA and positive wildcard answers.
- Accept tls-upstream in unbound.conf, the ssl-upstream keyword is
also recognized and means the same. Also for tls-port,
tls-service-key, tls-service-pem, stub-tls-upstream and
forward-tls-upstream.
- [dnscrypt] introduce dnscrypt-provider-cert-rotated option,
from Manu Bretelle.
This option allows handling multiple cert/key pairs while only
distributing some of them.
In order to reliably match a client magic with a given key without
strong assumption as to how those were generated, we need both key and
cert. Likewise, in order to know which ES version should be used.
On the other hand, when rotating a cert, it can be desirable to only
serve the new cert but still be able to handle clients that are still
using the old certs's public key.
The `dnscrypt-provider-cert-rotated` allow to instruct unbound to not
publish the cert as part of the DNS's provider_name's TXT answer.
- Update B root ipv4 address.
- make ip-transparent option work on OpenBSD.
- Fix#2801: Install libunbound.pc.
- ltrace.conf file for libunbound in contrib.
- Fix#3598: Fix swig build issue on rhel6 based system.
configure --disable-swig-version-check stops the swig version check.
Bug Fixes
- Fix#1749: With harden-referral-path: performance drops, due to
circular dependency in NS and DS lookups.
- [dnscrypt] prevent dnscrypt-secret-key, dnscrypt-provider-cert
duplicates
- Better documentation for cache-max-negative-ttl.
- Fixed libunbound manual typo.
- Fix#1949: [dnscrypt] make provider name mismatch more obvious.
- Fix#2031: Double included headers
- Document that errno is left informative on libunbound config read
fail.
- iana port update.
- Fix#1913: ub_ctx_config is under circumstances thread-safe.
- Fix#2362: TLS1.3/openssl-1.1.1 not working.
- Fix#2034 - Autoconf and -flto.
- Fix#2141 - for libsodium detect lack of entropy in chroot, print
a message and exit.
- Fix#2492: Documentation libunbound.
- Fix#2882: Unbound behaviour changes (wrong) when domain-insecure is
set for stub zone. It no longer searches for DNSSEC information.
- Fix#3299 - forward CNAME daisy chain is not working
- Fix link failure on OmniOS.
- Check whether --with-libunbound-only is set when using --with-nettle
or --with-nss.
- Fix qname-minimisation documentation (A QTYPE, not NS)
- Fix that DS queries with referral replies are answered straight
away, without a repeat query picking the DS from cache.
The correct reply should have been an answer, the reply is fixed
by the scrubber to have the answer in the answer section.
- Fix that expiration date checks don't fail with clang -O2.
- Fix queries being leaked above stub when refetching glue.
- Copy query and correctly set flags on REFUSED answers when cache
snooping is not allowed.
- make depend: code dependencies updated in Makefile.
- Fix#3397: Fix that cachedb could return a partial CNAME chain.
- Fix#3397: Fix that when the cache contains an unsigned DNAME in
the middle of a cname chain, a result without the DNAME could
be returned.
- Fix that unbound-checkconf -f flag works with auto-trust-anchor-file
for startup scripts to get the full pathname(s) of anchor file(s).
- Print fatal errors about remote control setup before log init,
so that it is printed to console.
- Use NSEC with longest ce to prove wildcard absence.
- Only use *.ce to prove wildcard absence, no longer names.
- Fix unfreed locks in log and arc4random at exit of unbound.
- Fix lock race condition in dns cache dname synthesis.
- Fix#3451: dnstap not building when you have a separate build dir.
And removed protoc warning, set dnstap.proto syntax to proto2.
- Added tests with wildcard expanded NSEC records (CVE-2017-15105 test)
- Unit test for auth zone https url download.
- tls-cert-bundle option in unbound.conf enables TLS authentication.
- Fixes for clang static analyzer, the missing ; in
edns-subnet/addrtree.c after the assert made clang analyzer
produce a failure to analyze it.
- Fix#3505: Documentation for default local zones references
wrong RFC.
- Fix#3494: local-zone noview can be used to break out of the view
to the global local zone contents, for queries for that zone.
- Fix for more maintainable code in localzone.
- more robust cachedump rrset routine.
- Save wildcard RRset from answer with original owner for use in
aggressive NSEC.
- Fixup contrib/fastrpz.patch so that it applies.
- Fix compile without threads, and remove unused variable.
- Fix compile with staticexe and python module.
- Fix nettle compile.
- Fix to check define of DSA for when openssl is without deprecated.
- iana port update.
- Fix#3582: Squelch address already in use log when reuseaddr option
causes same port to be used twice for tcp connections.
- Reverted fix for #3512, this may not be the best way forward;
although it could be changed at a later time, to stay similar to
other implementations.
- Fix for windows compile.
- Fixed contrib/fastrpz.patch, even though this already applied
cleanly for me, now also for others.
- patch to log creates keytag queries, from A. Schulze.
- patch suggested by Debian lintian: allow to -> allow one to, from
A. Schulze.
- Attempt to remove warning about trailing whitespace.
- Added documentation for aggressive-nsec: yes.
0.28:
Cdef classes can now multiply inherit from ordinary Python classes.
Type inference is now supported for Pythran compiled NumPy expressions.
The const modifier can be applied to memoryview declarations to allow read-only buffers as input.
C code in the docstring of a cdef extern block is copied verbatimly into the generated file. Patch by Jeroen Demeyer.
When compiling with gcc, the module init function is now tuned for small code size instead of whatever compile flags were provided externally. Cython now also disables some code intensive optimisations in that function to further reduce the code size.
Decorating an async coroutine with @cython.iterable_coroutine changes its type at compile time to make it iterable. While this is not strictly in line with PEP-492, it improves the interoperability with old-style coroutines that use yield from instead of await.
The IPython magic has preliminary support for JupyterLab.
The new TSS C-API in CPython 3.7 is supported and has been backported. Patch by Naotoshi Seo.
Cython knows the new Py_tss_t type defined in PEP-539 and automatically initialises variables declared with that type to Py_tss_NEEDS_INIT, a value which cannot be used outside of static assignments.
The set methods .remove() and .discard() are optimised. Patch by Antoine Pitrou.
dict.pop() is optimised. Original patch by Antoine Pitrou.
Iteration over sets and frozensets is optimised.
Safe integer loops (< range(2^30)) are automatically optimised into C loops.
alist.extend([a,b,c]) is optimised into sequential list.append() calls for short literal sequences.
Calls to builtin methods that are not specifically optimised into C-API calls now use a cache that avoids repeated lookups of the underlying C function.
Single argument function calls can avoid the argument tuple creation in some cases.
Some redundant extension type checks are avoided.
Formatting C enum values in f-strings is faster, as well as some other special cases.
String formatting with the ‘%’ operator is optimised into f-strings in simple cases.
Subscripting (item access) is faster in some cases.
Some bytearray operations have been optimised similar to bytes.
Some PEP-484/526 container type declarations are now considered for loop optimisations.
Indexing into memoryview slices with view[i][j] is now optimised into view[i, j].
Python compatible cython.* types can now be mixed with type declarations in Cython syntax.
Name lookups in the module and in classes are faster.
Python attribute lookups on extension types without instance dict are faster.
Some missing signals were added to libc/signal.pxd. Patch by Jeroen Demeyer.
The warning about repeated extern declarations is now visible by default.
The exception handling of the function types used by CPython’s type slot functions was corrected to match the de-facto standard behaviour, so that code that uses them directly benefits from automatic and correct exception propagation. Patch by Jeroen Demeyer.
Defining the macro CYTHON_NO_PYINIT_EXPORT will prevent the module init function from being exported as symbol, e.g. when linking modules statically in an embedding setup. Patch by AraHaan.
3.49.1:
This patch fixes our tests for Numpy dtype strategies on big-endian platforms, where the strategy behaved correctly but the test assumed that the native byte order was little-endian.
There is no user impact unless you are running our test suite on big-endian platforms. Thanks to Graham Inggs for reporting :issue:1164.
3.49.0:
This release deprecates passing elements=None to collection strategies, such as :func:~hypothesis.strategies.lists.
Requiring lists(nothing()) or builds(list) instead of lists() means slightly more typing, but also improves the consistency and discoverability of our API - as well as showing how to compose or construct strategies in ways that still work in more complex situations.
Passing a nonzero max_size to a collection strategy where the elements strategy contains no values is now deprecated, and will be an error in a future version. The equivalent with elements=None is already an error.
4.2.0:
Features added
* SelectElement.value returns more standard-compliant and browser-like defaults for non-multi-selects. If no option is selected, the value of the first option is returned (instead of None). If multiple options are selected, the value of the last one is returned (instead of that of the first one). If no options are present (not standard-compliant) SelectElement.value still returns None.
* The HTMLParser() now supports the huge_tree option. Patch by stranac.
Bugs fixed
* Some XSLT messages were not captured by the transform error log.
* Crash at shutdown after an interrupted iterparse run with XMLSchema validation.
