system-specific defines seems to have failed, we need to force the
issue on SunOS. OmniOS-CE has a patch to revert the change, but that
doesn't quite sit right with me either.
Changes:
abi
* RAnalOp.srcs,dsts are not pointers
anal
* Working apt and add apl to list function preludes
* Rename axj to axlj, because axj is for jmp refs
* Introduce anal.tailcall.delta and use flags for better metrics
* Improve the tailcall detection logic
* Improve warning that only seems to happen when anal.nopskip is set
* Always show all the archinfo, even when not provided by the plug
* Dont show analysis progress on non-interactive shells
* Add esil.dfg.mapinfo and esil.dfg.maps config vars
* Some more improvements to esil_dfg
* Fix size returned from r_anal_op
* Fix warning in aflj when parsing vargarg signatures
* Add register computed const pointer support for esil dfg
* Add memory computed const pointer support for esil dfg
* Introduce R_ANAL_ESIL_DFG_TAG_{REG,MEM}
* Use treebuf io plugin as memory access backed for esil_dfg
* Fix pickle asm rejecting empty strings
* Do not recurse noreturn inspection when !addr or -1
* Generalize vector instruction types instead of following intel-specific
* Add /au to search for unknown destination jmp/call
* Add anal.noret and refactor anal.noret.refs
* Fix#20827 - Show srcs/dsts in aoj
* Fix aae argument parsing regression in and improve help
* Add support for stack-computed const pointers in esil_dfg
* Fix anal.a2f in aac
* Increase default anal.depth from 64 to 128
* Clarify which commands are used on each aaaa line
* Fix anal.depth usage when analyzing one basic block
* Loongarch analysis bug fixes (bl, race condition)
* Implement aflxv and aflx? commands
* Run /azq in aaaa
* Fix long1,long4 pickle opcodes
* Fix#20798 - Fix bx after add lr,pc,0 in arm32
* Fix null pointer in aflxj
* Implement aflxj
* Add noreturn column in afll
* Use RPVector in RAnalOp src/dst to support ldm/stm/simd
* Fix pickle arch thinking 0 is 64 bit
* Don't show the linearsize in the afl output
* Add anal.vars.newstack - configurable improved stack-relative var
analysis
* Working apt and add apl to list function preludes
* Rename axj to axlj, because axj is for jmp refs
* Introduce anal.tailcall.delta and use flags for better metrics
* Improve the tailcall detection logic
* Improve warning that only seems to happen when anal.nopskip is set
* Always show all the archinfo, even when not provided by the plug
* Dont show analysis progress on non-interactive shells
* Do not recurse noreturn inspection when !addr or -1
* Generalize vector instruction types instead of following intel-specific
* Add /au to search for unknown destination jmp/call
* Add anal.noret and refactor anal.noret.refs
* Fix#20827 - Show srcs/dsts in aoj
* Fix aae argument parsing regression in and improve help
* Fix anal.a2f in aac
* Increase default anal.depth from 64 to 128
* Clarify which commands are used on each aaaa line
* Loongarch analysis bug fixes (bl, race condition)
* Implement aflxv and aflx? commands
* Run /azq in aaaa
* Add noreturn column in afll
* Add anal.vars.newstack - configurable improved stack-relative var
api
* Make RReg refcounted
* Implement {ctz|clz}{32|64} RNum
* Define RPluginMeta and RPluginStatus
* Add new RCore.cmdCallf() helper function
* Merge RParse into RAsm
* Refactor RLang api to use the new design
* Fix null deref on wrong api usage for RCore.cmdStr
* Moving more logic between asm, arch, parse and anal
* RAnalEsil -> REsil api refactor
* Deprecate reil and sysarch defines
* More refactorings and api redesigns in r_arch
* Remove eprintf calls in favor of R_LOG
* Implement RReg.clone()
* Deprecate r_str_dup() - related to #20959
* Rename RVector.len to RVector.length for consistency
* Remove the unnecessary RThread.CpuAffinity()
* Add portable NaN and INF defines for different float sizes
* Deprecate r_cons_eprintf and use R_LOG instead
* Rename RStr.home() to RFile.home() as part of the Plan
* Rename r_mem_memzero to r_mem_zero
* Prefer _tostring() instead of _to_string()
* Improve r_ref implementation with debugging support
* R_BIN_NM -> R_BIN_LANG
* Implement thread-safe refcounting - but disabled by default
* Deprecate the unused RFList
* Implement r_str_ntrim() and speedup r_str_trim() with it
* Initial implementation of RString (30% faster than RStrBuf)
* Implement r_sys_getenv_asint
* Add r_cons_is_initialized
* Boolify r_core_yank_file_all() and fix shadow var bug
* Add r_file_is_executable and r_file_extension apis
* Fix UB bug when using r_vector random access
* Change R_LOG_INFO to R_LOG_TODO where suitable
* Merge rhash into rcrypto and improve apis
* Fix memory leak in r_str_list_join()
* Boolify and rename some methods and fields from RFS
* Add .author field in all the RLang plugins
* Add a public api for the yank-unset action
* Constify the help
arch
* Add the arch.preludes() callback and new RSearchKeyword constructor
* Move anal.v850 to arch
* Fix counted string bug in pickle
* Fix negative unsigned cast in the xtensa disassembler
* Add RAnalOp.weakbytes() and move more analop apis to arch
* Move anal.xap into the arch
* Update tests and better arch.patch/modify callback
* Move anal.{6502,snes} into arch
* Kill RAsmOp, we can reuse RAnalOp in here
* Improve pickle disasm on invalid instructions
* Remove RAsmPlugin struct and add the 'aia' command to show archinfo
* Move the remaining asm plugins into the arch
* Minor plugin selection improvements
* Move asm.nasm into the arch
* Move asm.vasm into arch.any_vasm
* Assemble large pickle instructions
* Fix and move failing tests, reorder lib build
* Move the arm assembler plugin from asm to arch
* Temporary add RAnal as dependency for REgg
* Improve x86.nz assembler parsing and other bugs in rnum
* Initial implementation of the arch.any.as plugin
* Better handle of RNum errors for egg and arch.x86.nz
* Support reg+idx and idx+reg in x86.nz assembler
* Move the x86.nz plugin
* Fix asm.acur supporting arch, anal and asm plugins
* Fix asm.acur supporting arch, anal and asm plugins
* Move anal_riscv to arch_riscv
* Fix rasm2 -LLL using the new multi-bits macros
* Introduce RSysBits and its packing/checking macros
* Implement archinfo() in RAnal.Plugin.tms320
* Deprecate the unused RArchPlugin.esil field
* Use PJ to return the list of mnemonics aoml in arm.v35
* Move anal.rsp to the new home
* Move anal.v810 into arch.v810
* Move pickle from anal to arch and add it to meson
* Remove anal.malbolge and fix CI r_esil issues
* Move the 'sh' plugin to the new home
* Honor plugin name in rate matching for RArch.use
* Move jdh8 from asm/anal to arch
* Unify RArchOp into RAnalOp using common include files
* Fix RArchOp.refptr from bool to int
* Bump cs5 to support FNOP on m68k
* Wire-up RArch into RAnalOp
* Fix arm64 plugin to work well with latest arm64 changes in capstone
* Use the latest capstone5-next with updated aarch64 support
* Copy anal_amd29k.c to rarch
* Change arch plugin definition
* Add some more arch config vars
* Introduce arch.endian config var
* Instantiate RArch in anal
* Introduce RArchConfig->decoder
* Add R_LIB_TYPE_ARCH and i4004 arch-plugin
* First arch plugin (arch.null), implement basic lib api
* Start moving EVM analysis from extras to core
* First implementation of r_arch decoder api
* Introduce the new r_arch library, just the skeleton
* Add some r_arch api declarations
* Initial commit on RArch structs
asm
* Deprecate more unused fields from RAsmPlugin
* Fix the parse.z80.pseudo plugin and add a test
* Remove the unused RAsm.binb
* Internal cleanup of asm.c, deprecate the disassembly callback
* Load cpu descriptions for multiarch plugins
* Fix rasm2 x86.nz for "xchg eax,eax" and add tests
bin
* Fix JSON encoding of section addresses
* Add test for cwd source listing, CLL and list
* Add warning when loading DWARF5 files (not supported)
* Add test for the obm with CL, support noncwd paths
* Use obm when spotting a companion dwarf file on macOS
* Implement RBinFile.merge() and obm command to use it
* Initial implementation of the ob-- command to close the last binobj
* Autoload the dwarf companion file on macOS systems if available
* Use rabin2 -rO for raw dump operations
* Use glob expressions to specify section name to dump
* Fix#14540 - klass->super must be an RList instead of char*
* Support for Xbox 360 PE32 architecture (PPC BE)
* Fix wrong detection of main in elf-arm32
* Fix rabin2 -gj and add tests
* Add bin.types and disable by default for CI reasons
* Support loading Plan 9 kernels
* Use API instead of commands to autoload a pdb
* Fix#21020 - fix json format for rabin2 -jM when no main is found
* Expose section type for coff, elf and macho formats
* Fix#18375 - Only patch arm64 relocs when not initialized
* Add experimental bin.str.nofp config for less false positives
* Parse the PT_DYNAMIC elf section for the preinit pointers
* Fix validation check in xnu
* Initialize macho header pf definitions
* Speedup class bin loading with bin.filter=false
* Implement 'ic.' command
* Fix initial seek for Rosetta2 aot binaries
* Silent noisy warning in dwarfprocess
* Expose klass->super details for objc categories
* Include fields in the ic output
* Enable the swift metadata parser by default and import classinfo
* Add lang field for classes, symbols and methods, expose it via ic
* Warn about unpatched relocs when no bin.cache is set for macho fixups
* Add support for 32bit Mach-O fixups
* Expose the macho reloc fixups and use internal buffer for parsing
* Fix obf and add tests for it
* Fix Cd4[ invalid syntax used in macho _const section
* Expose the id_dylib macho command info into the bin kv
* Enlarge the c++ demangler stack limit to solve a warning
* Implement Dwarf.register identifier mapping for v850
* Infuse asm.cpu from the elf flags for v850 ELFs
* Remove asm.features, improve RBinInfo with flags and abi details
* Improve brainfuck detection to reduce false positives
* Implement dwarf.regName() for arm64
* Import the free pascal symbol demangler from rizin
build
* Double lowerdash defines should be defined only by the compiler
* Fix and improve static build, faster libr.a with libtool if available
* Improve libr.a creation with ar -rcT instead of ar -x
* Speedup tcc builds by not using -g
* Move esil one level up in libs.mk
* Update sdb to remove double include paths
* Fix compilation with -lcrypto
* Do not use macos-latest (macos-11 is fine for LTS)
* Update the capstone4 support to 4.0.2
* Fix zig's @Cinclude of r_th.h
* Add crosscompiling support with sys/zig.sh
* Do not install the v35 archives (-50MB) in make install
* Fix capstone dynamic memory allocation setup issue
* Support sys/debian.sh crossbuilds
* Remove unused lc-printscan-long-double wasi flag
* Check if CWD contain spaces in sys/install.sh
* Add the acr --enable-threadsafety flag and the same for meson
* Fix meson infinite loop
* Disable libuv by default on meson and acr
* Change build order as long as now bin depends on fs
ci
* Test r2 build with all sysdependencies enabled
* Upgrade al-cheb/configure-pagefile-action to the v1.3
* Run unit tests in parallel
* Upgrade github actions/checkout from v2 to v3
* Upgrade actions/upload-artifact from v2 to v3
cmd
* Fix bugs in aeg command parser
config
* Deprecate the use of comma in e: as stated in the r2580 prophecy
* Deprecate graph.web eval config var
* Remove the file.offset unused config variable
cons
* Fix static themes listing
* Add support for statically compiled themes
* Fix Ctrl+Arrow dietline shortcut for word cursor
* Fix console history log path regression
core
* Enable cmd.undo by default
* Rename many bin.str evars into bin.str.
* Use XDG instead of R2_HOME_CACHEDIR and R2_HOME_HISTORY
* Initial support for XDG env vars and paths
* Fix RConfig.setB when the key doesnt exist
* Deprecate scr.seek configuration variable
* Initial implementation of R_LOG_TODO
crash
* Fix segfault in poa 1
* Fix integer overflow in fuzzed dwarf rendering in graphs
* Fix use-after-free after @@@e spotted by meme
* Fix UAF in aaft when the BB is removed during the loop
* Fix UAF on quit exposed by r2frida
crypto
* Initial abi breaking changes in RCrypto/RHash
* Fix rahash2 -L listing full hash
* Initial work on the way RCrypto handles plugins
* Separate SM4 algorithm from plugin
* Implement the SIP hash algorithm
debug
* Bring back the 'dms' command
* Fixed incorrect thread arena output
* IO uses PID to read from child, tid is just for regs
* Fix r2 -d foo\bar.exe and r2 -d bar.exe on windows
* Make RDebug.regRead() and regWrite() return bool
decompiler
* Improve the outpuf of pdc by trimming the addresses of inline nops and colorize numbers
* Add colorization support to pdc output
disasm
* Implement asm.bytes.align to justify them to the right
* Fix char auto-comment in cmp instructions
* Improve pseudodisasm for arm64
* Improve arm.pseudo when no function information is available
* Fix mips.pseudo shortpath when function is null
* Fix x86.pseudo shortpath when function is null
* Fix NULL function xrefs in pd
* Implement asm.flags.right option
* Add dummy parse.evm plugin to fix portability of test
* Add a dummy bpf pseudo plugin
* Add RParse.justify() to easily fix commas and spaces
* Better spacing in arm.pseudo parse plugin
esil
* Initial support for threads in esil
* Fix#21052 - wrong emulation for pop rsp
* Move anal.esil into the new esil
* Add aegb command as an alias for 'aeg pieq $Fi'
* Add aegn command to combine N esil instructions into one dfg
* Use a function instead for the spaguetti code in all cmp esil opcodes -30LOC
globals
* Remove one global variable in RCore.cmdEval()
* Remove two globals from RCons.cpipe
* Remove the 3 globals in anal.xtensa
* Remove 3 globals from anal.tricore
* Remove the 3 globals from anal.vax
* Remove 3 globals from anal.nios2
* Remove 3 globals from anal.arc
* Remove 3 globals from anal.sparc.gnu
* Remove 3 globals from anal.sh
* Remove 3 globals from anal.alpha
* Remove 3 globals from anal.lanai.gnu
* Remove 3 more globals from anal.pdp11
* Remove 3 globals from anal.hppa
* Remove 3 globals from anal.m68k.gnu
* Remove the 3 globals in anal.ppc.gnu
* Remove 3 globals from anal.cris
* Remove 4 globals from anal.mips.gnu
* Deglob 4 vars in the arm.gnu disassembler
* Remove 3 globals in s390.gnu
* Remove 10 more globals from analysis and capstone
* Remove 3 globals in ccarg analysis
* Remove all global variables from RCrypto
* Move colortable global into the RConsContext
* Remove global from utf8
* Remove the last global variable in libmagic
* Remove in_log_process global
* Remove global in bin.obj.reloc_patch
* Remove global variable in esil loop
graph
* Initial work in graph.bubble for custom bg color in nodes
* Honor graph.layout in aegv too, instead of harcoding horizontal one
* Deprecate aegi and aggi, those were dupes for aegv and aggv
* Fix null deref in agg and avoid destructive manners of 'V ,'
help
* Make the anal.depth warning more useful
indent
* Balance spacings in braces
io
* Initial implementation of the generic io-stream api
* Fix UAF in streaming io plugins when used with io.va=1
* Remove invocation of v layer cache in r_io_desc_read
* Start rewriting io_cache.c
* Kill r_io_read_at_mapped
* Initial import of the serial plugin
* Kill io->buffer
* Remove unused fcn declarations
* omfg runs omm if no map is set
* New o++ command to create and open a new file
* Add "reset" system command to treebuf io plugin
* Minor bugfix in treebuf io plugin
* Add treebuf io plugin
* Fix free-before-use on r_io_reopen of a rbuf:// fd/desc
* Handle reloc maps properly in r_io_map_remap and r_io_map_resize
* Fix reloc map memleak
* Add rio reloc maps
* [5.8.0] Disable the default io.basemap
json
* Fix tj ttj tfj outputs
* Fix invalid json in tj command
* Fix#20772 - ihj rendering an invalid json because of pfj
lang
* Make r2 -j work as a hashbang handler for qjs
* Enable BigNum in qjs
* Use r2papi 0.0.4 with base64 and R2Api is now known as R2Papi
* Add requirejs, simplify compilation and add js_ prefix to all the c files
* Add typescript support
* Minor improvements for js: with r2.call() and r2.cmdj
* Enable Bignum support to the QJS interpreter
* Import the alpha r2papi 0.0.2 api for qjs
* Integrate the qjs repl into the js: command
* Initial import of the interactive QJS repl
* Import the quickjs rlang plugin
* Implement py command and add stdin slurp support for js- too
* Add 'js' and 'js:' commands, as well as improve help for #!?
* Fix null deref in rlang
* Add "lua" as an alias for "#!lua"
leaks
* Fix leaks in pdc
* Fix more memory leaks in rbin and ranal for arm64
* Patch more leaks in the analysis and rbin
* Some safe memleaks related to analysis and registers refcounting
* Fix memory leak in dietline
lint
* Enable the leading spaces linter rule and fix them all
* Use more tabs and add a (disabled for now) linter for it
logs
* Redirect RLog messages into the Corelog
* Add base64 support to the T and T* commands
panels
* Add ve command to set fg/bg colors for current panel
print
* New RPrint.spinBar() API used from scr.demo for now
* Fix pcc trifids confussion issue
* New command CLL (aka list) show function source using addrline (dwarf) info
* Fix#21080 - Add cfg.codevar to change the buffer varname from pc
* ASN.1 and x509: correct OCTET_STRING and Public key info parsing
* Add ASN1 Algorithm Identifiers for Edwards curves
* Fix#20993 - Correct ASN.1 BIT_STRING parsing
* Implement pcn command to print bytes as space separated numbers
* Add support for the swatch dot-beat internet time
* Workaround for "too large buffer" in formats
* px* is an alias for pc*
* Implement pFoj command
* Implement pFaj for asn1 json decoding
* Initial refactoring/cleanup of ASN1 parser api
* Add pFxj command to print x509 certificates in JSON format
* Add pFpj command to print PKCS7 files as JSON
* Fix issue in 'pdc' that was showing empty orphan nodes
* Implement new 'pcq' command, like pc, but inline-include-friendly
* Implement the new pieb command as an alias for pie $Fi
* Fix read buffer overflow in pxq -272
* Implement pFbJ command with quiet and verbose json formats
* Implemen pFbj for json printing of protobuf
* Set hex.hdroff=true by default
* Implement pFAj to render android xml in JSON (abi break)
* Implement TSV output format for RTable
projects
* Fix some problems when renaming projects
* Fix some bugs in projects
* Quote commit message to avoid git error when saving project
* @radare Do changes in Px->Pc, Pc->PS* as planned
r2pipe
* Check magic header before assuming an interpreted file is executable
r2pm
* Fix assert in Str.Trim() when r2 is not installed
* Add support for tarball and zip packages
* Honor EDITOR in r2pm -e
* Remove all the references to the old r2pm.sh
* Implement R2PM_FAIL and mark it as deprecation for r2-5.9.x
* Implement r2pm -cp like it was in r2pm.sh
* Fix clean installations with r2pm -c
* Expose R2PM_SUDO and list R2PM_PREFIX in -H
* Handle -HH in r2pm for verbose env listing and remove unused R2PM_GITSKIP
* Fixes pull/install/uninstall on windows
* r2pm -Ui can be combined now and fix extras package building
* Honor R2PM_DBDIR env var
* Honor -f in r2pm -U to force clean the r2pm db
* Support XDG on r2pm and expose the PKG_CONFIG_PATH
* Expose R2_LIBEXT for r2pm packages
* Update r2pm manpage and add -q and -a flags
* Show package source with r2pm -d
* Implement r2pm -H to make more packages build
* Fixes parsing the GIT URL on some packages
* Test the new default native r2pm fix flushing and using RLOG
* Make R2PM_NATIVE the default and provide R2PM_LEGACY
r2r
* Fix rvector assert when indexing empty ones
refactor
* Move the RParse.cparse into RAnal.cparse
* Stop aeg from abusing agg
* Rename R_ANAL_ESIL_DFG_BLOCK_ to R_ANAL_ESIL_DFG_TAG_
* Rename EsilDFGRegVar to EsilDFGVar and introduce EsilDFGVarType
* Avoid using RArchConfig->big_endian
* Add addr_bits to RArchPlugins and make info and decode cbs cfg aware
* Add archcond api to rarch, some small cleanup in anal
* Make bitness, endianess and esil-support fields of RArchPlugin again
* Copy value.c and op.c from anal to arch
* Copy switch.c from anal to arch
* Add some more typedefs and enums to r_arch
* Rename r_arch_set_ to r_arch_config_set_
* Rename R_ASM_SYNTAX to R_ARCH_SYNTAX
* Simplify x86_cs BSR and BSF esil
* Use r_strbuf_replacef in anal_mips_gnu esil generation
* Use r_strbuf_replacef in anal_mips_cs esil generation
rvc
* Initial refactoring of the version control api
* Move rvc from core to util
* Fix rvc.commit when non-interactive with a default message
scan
* Fix crash in the swift metadata parser spotted by coverity
search
* Fix JSON encoding of unsigned search values
* Fix /au after aeim
* Fix /w and /wi, add tests, minor code cleanup
* Add r_anal_optype_index to make /atl and /at use full listings of optypes
* Support space separated instruction types and family in /at and /af
* Improve json output for /asj and /atj
* Fix calling /re twice after ^C
* /az uses anal.in instead of search.in to improve scan results
* Make /az faster after aeim, skipping unrelated regions
* Fix last char bug in swift strings found with /az
* Add x86-64 support to /az
* Add flags under the asm.str flag prefix when doing /az
* Honor bin.minsz in /az is no argument is provided
* Implement /azq to search for assembly strings and add tests
* Implement the new /az command to find assembly constructed strings
* Superseed #20447 - remove some magic globals
shell
* Fix ?vi:123 and ?v:123 commands
* Print whatever is taken from io_system to rcons
* Fix runtime warning after leaving an rlang session
* Add "" command to run RCore.cmdCall()
* Fix#21136 - o <tab> autocompletion not working
* Show help when using invalid subcommand of afi
* Implement abo and afbo commands to list opcode offsets in function or bb
* Implement o-. command, add help for future o-$
* Expose RCore.cmdCall() and fix b64: command + add tests
* Show number conversion error messages in ?v command
* Fixes for the line editor using live save/load with new RFile apis
* Implement oe command to open a file using cfg.editor
* Add quiet and table listing for lang plugins
* Implement Ll, Llq and #!?q commands for better rlang listing
* Fix r2 /directory behaviour
* Support $r:REGNAME syntax and document it
* Implement cmd.usr1 and cmd.usr2 to handle signals on unix
* Add #!qjs and #!tiny for autocompletion
* Fix behaviour of -a and -b flags (no arg= show current, append? for help)
* Rename the drm command to drv for consistency
* Add -s -i -f r2 commands
* Implement -a, -b, -c and -e commands in r2
* Add tabhelp exception for pf.
* Add r2 -LL to list core plugins
* Improve help message for f subcommands
* Take into account static themes when listing
* Add ot command as an alias for touch
* Implement the ji: command as an alternative to ~{} without cons filtering
* Also handle (j) and (*), more syntax-consistent and add a test
* Implement (j for json output of macro commands
* Initial implementation of ?ie
* Add missing help for the '?i?' command
* Use RCoreHelp for /ca?
* Fix#20760 - Implement native gron via ~{=}
* Remove RPrintRowlog and use R_LOG isntead
* Allow changing number of saved input lines
* Implement log.source and log.origin
* Add the new ucu and ucd commands using the new core-undo apis
* Rename asm.{off} variables to asm.offset
syntax
* Move the preincrement and void arg from tests to lint.sh
threads
* Initial ref-counted RRegItems, needed for threadsafety
* Move the readahead logic to a local variable
* Guard more critical sections in cons and core
* Analysis now waits in background for the bin parsing to finish
* Add some RThreadLocks and start to use the critical sections
thready
* Dont call RCore.seek() and read a new buffer in disasm
tools
* rax2: corrects base64 encoding for null bytes
* Fix r2 -2
* Fix broken tests for long number conversion
* Fix base64 null byte decoding bug in rax2
* Implement rasm2 -LLL to list arch plugins
* Add Ls to list assemblers, and LA to list analysis plugins
* Bring back the r2 -t for parsing bin and analysing in background
* rarun2 supports multiple preload directives
util
* Fix the XML parser
* shlr/yxml -> libr/util/rxml - fork the abandoned yxml parser and expose it
* Add RStr.ansiStrip() and RStr.insert()
* Improve internal RBuffer API checks
* Add log error when pj depth limit reached
* Add R_SYS_BITS_12
* Introduce R_SYS_BITS_4
* Add :header and :noheader in RTable
* Fix crash in r_vector_shrink
* Minor COV fix in r_str_char_count
* Fix return type of r_str_char_count
* Add new RStr.replaceAll() api
* Use R_PRINTF_CHECK for r_strbuf_replacef
* Add r_strbuf_replace{f}
vc
* Initial rvc refactoring and cleanup of the api
* Move rvc apis into callbacks
* Make ravc2 accessible via blob and r2 shell
visual
* Use RAnalOp instead of RAsmOp in r_core_visual_bit_editor
* Fix issue with cursor disappearing towards the bottom of the screen
* Don't draw two cursors when too many bytes are on disasm panels
* Make j/k movement in panels' cursor mode more consistent with it's visual counterpart
* Don't skip byte when moving left/right in cursor mode (disassembly panel)
* Fix pdc glitching in panels
* Handle arrow keys in VT
* Handle JK0 keys in VT
* Honor cmd.vprompt and scr.notch in VT
* RStr.wrap() supports ansi and use it in VT
* Implement Tv command and use it from VT
* Fix the cache and other bugs in panels
* Override scr.maxpage in panels to avoid undesired prompts
wasm
* Upgrade to the latest wasi16 sdk
Changes in 229:
[ Chris Lamb ]
* Skip test_html.py::test_diff if html2text is not installed.
(Closes: #1026034)
[ Holger Levsen ]
* Bump standards version to 4.6.2, no changes needed.
Changes in 228:
[ FC Stegerman ]
* As an optimisation, don't run apktool if no differences are detected before
the signing block. (Closes: reproducible-builds/diffoscope!105)
[ Chris Lamb ]
* Support both the python3-progressbar and python3-progressbar2 Debian
packages, two modules providing the "progressbar" Python module.
(Closes: reproducible-builds/diffoscope#323)
* Ensure we recommend apksigcopier. (Re: reproducible-builds/diffoscope!105)
* Make the code clearer around generating the Debian substvars and tidy
generation of os_list.
* Update copyright years.
Changes in 227:
[ Chris Lamb ]
* Don't attempt to attach text-only differences notice if there are no
differences to begin with. (Closes: #1024171, #1024349)
* Don't run Python decompiling tests on Python bytecode that both file(1)
cannot yet detect and Python 3.11 cannot demarshall. (Closes: #1024335)
3.11.5:
## Changes Worth Mentioning
### Core Server
#### Enhancements
* Backported a number of free disk space monitor resiliency improvements.
GitHub issue: #5831
* `raft.adaptive_failure_detector.poll_interval` exposes [`aten`]()'s
`poll_interval` setting to RabbitMQ users. Increasing it can reduce
the probability of false positives in clusters where inter-node
communication links are used at close to maximum capacity. The
default is `5000` (5 seconds). GitHub issue: #6632
* When both `disk_free_limit.relative` and `disk_free_limit.absolute`,
or both `vm_memory_high_watermark.relative` and
`vm_memory_high_watermark.absolute` are set, the absolute settings
will now take precedence. GitHub issue: #4980
* Closing channels will now log a warning if they had any messages
pending a confirmation from the server. GitHub issue: #1399
* New quorum queue option for in-memory table (MemTable) compression.
GitHub issue: #6590
* Default queue type (a virtual host setting) is now applied when
importing definitions into a single virtual host. GitHub issue: #6599
#### Bug Fixes
* Feature flags provided by plugins were mistakingly disabled after
node restart. GitHub issue: #6500
* Classic queues with Single Active Consumer enabled could run into an
exception. GitHub issue: #6502
* When a global parameter was cleared, nodes emitted an internal event
of the wrong type. GitHub issue: #6538
### CLI Tools
#### Bug Fixes
* `rabbitmq-queues grow` and `rabbitmq-queues shrink` misformatted the
errors they could encounter. GitHub issue: #6601
#### Enhancements
* Implicit `help` command (when CLI tools were invoked without a
command name) now respects all global flags (such as `--node`). For
example, previously the `--node` flag in
rabbitmqctl --node rabbit@ns1.rabbitmq.cluster.local
was ignored but now CLI tools would discover what plugins are enabled
on node `rabbit@ns1.rabbitmq.cluster.local` and include them into
`help` output. GitHub issue: #6598
* New key supported by `rabbitmqctl list_queues`:
`effective_policy_definition` that returns merged definitions of
regular and operator policies effective for the queue. GitHub
issue: #6556
### Management Plugin
#### Enhancements
* It is now possible to omit explicitly specifying queue type when
declaring a queue (or stream) in the management UI, and rely on the
default queue type configured for the selected virtual host. GitHub
issue: #6600
* New HTTP API endpoint, `GET /api/config/effective`, returns effective
node configuration. This is an HTTP API counterpart of
`rabbitmq-diagnostics environment`. GitHub issue: #6016
### Shovel Plugin
#### Enhancements
* Flow control state for Shovels is now reported with higher fidelity
(of 1 second vs. several seconds previously). This means it should be
easier to spot Shovels that run into flow control using management
UI. GitHub issue: #6615
### Sharding Plugin
#### Bug Fixes
* Plugin could fail to boot and halt node boot due to incorrect boot
step metadata. GitHub issue: #6583
## Dependency Upgrades
* `ra` was upgraded from `2.4.1` to `2.4.5`.
3.11.4:
## Changes Worth Mentioning
### Core Server
#### Enhancements
* Import of definition files with many streams is now more efficient.
GitHub issue: #6436
* Lower CPU throughput in clusters with many mostly idle streams.
GitHub issue: #6436
* Streams with `max_age` retention now attempt to reclaim disk space
every hour. This is relevant in environments with a lot of mostly
inactive streams that set `max_age`. GitHub issue: #6436
* Quorum queues are now more resilient to WAL log growth with workloads
that involve clients that register a consumer and then close the
channel or connection without ever consuming any deliveries or
cancelling the consumer. GitHub issue: #6447
#### Bug Fixes
* When a node encouters an invalid `definitions.local.file` on boot,
it will refuse to start instead of ignoring the file. GitHub
issue: #2610
* Fixed a type analyzer definition. GitHub issue: #6401
### CLI Tools
#### Enhancements
* `rabbitmq-diagnostics check_if_node_is_quorum_critical` and
`rabbitmq-upgrade await_online_quorum_plus_one` now consider stream
(not just quorum queues) replica placement when determining if target
node is quorum-critical. GitHub issue: #6448
* Queue info keys now support more inclusive property names related to
(deprecated) classic mirrored queues. For example, `mirror_pids` can
now be used instead of `slave_pids`. GitHub issue: #2635
* `rabbitmq-diagnostics memory_breakdown` now executes significantly
faster in environments with a large number (say, tens or hundreds of
thousands) of quorum queues. Two orders of magnitude faster, in fact.
GitHub issue: #6388
#### Bug Fixes
* Definition export in JSON failed on nodes that used
`definitions.skip_if_unchanged`. GitHub issue: #6424
* Using quorum queue-specific commands on streams now results in
clearer error messages. GitHub issue: #6488
### LDAP Plugin
#### Bug Fixes
* LDAP server password could end up in the logs in certain types of
exceptions. GitHub issue: #4842
### STOMP Plugin
#### Enhancements
* `x-max-age` stream setting now can be set by STOMP clients via a
header. GitHub issue: #5003
## Dependency Upgrades
* `osiris` was upgraded from `1.3.3` to `1.4.0`
Previously max version was 3.9.
Hopefully at some point someone will have the time to upgrade pkgsrc
to a newer version of thunderbird, but in the meantime...
Merge pull requests:
pdewacht#107 Added DCP-7070DW; Thanks Jan Musinsky
pdewacht#133 Fixed MFC-7460DN; Thanks Peter Ye
pdewacht#169 Changed Code; Thanks Thomas Nixon
Added additional Brother printer entries:
HL-L5000D series
HL-L2370DN series
DCP-7070DW
DCP-8065DN
Errors Fixed:
Fix the ppd file name for the MFC-7460DN, changed from br7365dn.ppd to
br7460dn.ppd
Code Changes:
Explicitly disable duplex in PCL \033&l0S.
Issues Resolved:
The HL-L2350DW printer continued to duplex even when disabled, but every
other page was garbled. Thanks to Thomas Nixon for the fix.
## v1.14.2 (2022-11-11)
### 1. Enhancements
#### Elixir
* [Code] Add `Code.eval_quoted_with_env/4` with support for the
`:prune_binding` option
#### ExUnit
* [ExUnit.Case] Allow test cases to not be registered on use
* [ExUnit.DocTest] Include `:doctest` and `:doctest_line` as meta tags
* [ExUnit.Formatter] Expose `ExUnit.Formatter.format_assertion_diff/4`
#### Mix
* [Mix] `Mix.install/2` accepts atoms as paths
### 2. Bug fixes
#### Elixir
* [Code.Formatter] Fix `size*unit` shortcut in bitstring
* [Kernel] Generate unique variables for macro expansion of `defguard`
* [Protocol] Expand `:for` in protocols with the appropriate env
#### ExUnit
* [ExUnit] Do not run duplicate cases on `ExUnit.run/1`
#### Mix
* [mix test] Ensure proper error message when there is no test directory
## v1.14.1 (2022-10-10)
### 1. Enhancements
#### Elixir
* [Kernel] Perform partial expansion of literals in module attributes
* [Kernel] Do not add compile-time dependencies for literals as
defaults in `Application.compile_env/3` inside module attributes
* [Macro] Add `Macro.expand_literals/2` and `Macro.expand_literals/3`
* [System] Add `:close_stdin` to `System.shell/2`
#### Mix
* [mix test] Accept `--all-warnings` option
## 2. Bug fixes
#### Elixir
* [Kernel] Fix misleading warning when `:uniq` is given in `for`
comprehensions and the result is unused
* [Kernel] Improve error message for when there is a conflicting
struct and ignoring module conflict
* [Kernel] Do not delete `@enforce_keys` attribute after `defstruct`
declaration
* [Kernel] Do not crash the checker on modules with missing
`:debug_info` chunk
* [Macro] Fix error in `Macro.to_string/2` when converting an AST with
`:erlang.binary_to_atom/2`
* [String] Fix `String.split/3` and `String.next_grapheme/1` returning
invalid results on invalid UTF-8 encoding
* [System] Do not close stdin by default in `System.shell/2`
* [URI] Do not return `uri.port` as `:undefined` in certain cases in
`URI.new/1`
#### ExUnit
* [ExUnit.DocTest] Do not crash when both `:moduledoc` and functions
are specified in `:only`
#### IEx
* [CLI] Fix invalid argument handling when `--no-pry` is given
#### Mix
* [mix format] Do not cache inputs from `.formatter.exs` so they are
properly re-evaluted on every call
Potential incompatibilities:
- The inet:setopts/2 {reuseaddr, true} option will now be ignored on
Windows unless the socket is an UDP socket. For more information see
the documentation of the reuseaddr option part of the documentation of
inet:setopts/2. Prior to OTP 25 the {reuseaddr, true} option was
ignored for all sockets on Windows, but as of OTP 25.0 this was
changed so that it was not ignored for any sockets.
an OpenAPI spec.
swagger-codegen contains a template-driven engine to generate
documentation, API clients and server stubs in different languages by
parsing your OpenAPI / Swagger definition.
GNG is a script that automatically search your gradlew when you are
inside your Gradle project and execute it. It also contains an official
Gradle wrapper. You can create gradle projects from scratch without
installing Gradle.
This is originally inspired by gdub and gradlew-bootstrap.
asdf is a CLI tool that can manage multiple language runtime versions on
a per-project basis. It is like `gvm`, `nvm`, `rbenv` & `pyenv` (and
more) all in one! Simply install your language's plugin!
Why use asdf?
- single CLI for multiple languages
- consistent commands to manage all your languages
- single global config keeping defaults in one place
- single `.tool-versions` config file per project
- support for existing config files `.node-version`, `.nvmrc`,
`.ruby-version` for easy migration
- automatically switches runtime versions as you traverse your directories
- simple plugin system to add support for your language of choice
- shell completion available for common shells (Bash, Zsh, Fish, Elvish)
- Modernize patch filenames
- Take MAINTAINER
From the upstream changelog:
2.24.0:
Note: LilyPond releases can contain syntax changes, which may
require modifications in your existing files written for older
versions so that they work in the new version. To upgrade
files, it is *strongly recommended* to use the 'convert-ly'
tool distributed with LilyPond, which is described in *note
(lilypond-usage)Updating files with convert-ly::.
'convert-ly' can perform almost all syntax updates
automatically. Frescobaldi users can run 'convert-ly'
directly from Frescobaldi using "Tools > Update with
convert-ly...". Other editing environments with LilyPond
support may provide a way to run 'convert-ly' graphically.
Major changes in LilyPond
*************************
. LilyPond now requires Guile 2.2. Even if you are not writing
Scheme code yourself, you may be using libraries that contain a
non-trivial amount of customizations. If they do not work with
LilyPond version 2.24.0, please report this to the library's
developers. If you _are_ a library developer, *note Notes on Guile
2.2:: below.
. The infrastructure for creating the official binaries has been
completely rewritten, incident with the switch to Guile 2.2. As of
this release, we provide 64-bit binaries for macOS and Windows.
Also, all packages are made available as simple archives that can
be extracted to any "installation" location. To uninstall, simply
delete that directory. We also discontinued the limited editor
that was installed on macOS and Windows, LilyPad, and recommend
switching to an external solution instead, such as the popular
editor Frescobaldi (https://frescobaldi.org), or one of the others
listed in *note (lilypond-web)Easier editing::. For more
information, please refer to the detailed instructions in *note
(lilypond-learning)Installing::.
The version of Ghostscript that ships in the official binaries no
longer includes its own fonts, to reduce download size. LilyPond
does not need them (it has its own set of default fonts). This
does mean, however, that code such as '/Arial findfont' in '\markup
\postscript' no longer works. We recommend using normal '\markup'
syntax to print glyphs.
Known issues in LilyPond 2.24.0
===============================
On Windows, compiling very large scores (several hundreds of pages) can
result in crashes.(1) We hope to address this in a future 2.24.X
release.
---------- Footnotes ----------
(1) Note that very large scores could already crash in LilyPond 2.22
due to restrictions imposed by the 32-bit architecture.
Notes for source compilation and packagers
==========================================
This section is aimed at enthusiasts compiling LilyPond from source and
packagers preparing LilyPond for distribution. If you are not part of
either group, you can skip over this section.
. As mentioned above, LilyPond now requires Guile 2.2. If needed for
distribution reasons, it can also be compiled with Guile 3.0 by
passing 'GUILE_FLAVOR=guile-3.0' to the 'configure' script.
However, this is not yet recommended nor officially supported.
. The Scheme code evaluator in Guile 2.2 is slower than in Guile 1.X.
To offset most of the performance penalty, we recommend compiling
the '.scm' files into bytecode by first running 'make bytecode'
during compilation and then 'make install-bytecode' in addition to
'make install'.
. Starting with this stable release, LilyPond's build system does not
install text fonts anymore. Please provide them as separate
packages while paying attention to the fonts' license and notice
files.
New for musical notation
************************
Pitches improvements
====================
. Support for alternate accidentals was improved. Through the
'alterationGlyphs' property of staff-like contexts, accidental
glyphs may be set for all grobs at once (refer to *note
(lilypond-notation)Alternate accidental glyphs::).
. Ottava brackets may apply to a single voice instead of the entire
staff. This used to require workarounds.
Rhythm improvements
===================
. The new '\section' command inserts a double bar line that interacts
gracefully with repeat bar lines. A passage can be named with the
new '\sectionLabel' command.
. '\numericTimeSignature' and '\defaultTimeSignature' now act on all
staves at once (more precisely, on all staves in the same 'Timing'
context), thus matching the behavior of '\time'.
. The '\rhythm' markup command has been added. It is a simple way to
enter rhythms mixed with text, such as in "swing" indications.
. The '\enablePolymeter' command is now provided as an input
shorthand for moving engravers as is necessary to allow different
time signatures in parallel.
Independent of this, 'Default_bar_line_engraver' has been removed.
. The new option 'visible-over-note-heads' can be used to make tuplet
brackets always appear when their direction is set to be over the
note heads. It can be used with the default tuplet bracket
visibility style or with '#'if-no-beam'.
. Measure counts now take compressed multi-measure rests and
alternatives into account.
. Bar numbers may be centered in their measure, as is common in film
scores.
. The alignment of bar numbers appearing in the middle or end of a
system has been changed to align them on their left edge. This is
in keeping with the advice of Elaine Gould ('Behind Bars', p. 237),
and was mostly the consensus reached in a discussion of the issue
by developers. The alignment of bars at the beginning of a system
remains unchanged.
. '\bar ","' creates a short bar line.
. The following predefined bar types no longer appear as a single bar
line at the end of a line. Annotated bar types (e.g., '\bar
"S-|"') have been added for that purpose.
. '\bar ""' is no longer necessary to print the first bar number. It
now suffices to set 'barNumberVisibility' to
'all-bar-numbers-visible', or one of the other visibility settings
where the first bar number is visible.
Note that this is a change in behavior for scores that set
'barNumberVisibility' to 'all-bar-numbers-visible' or such and
'BarNumber.break-visibility' to '#t' without having '\bar ""'.
Now, a bar number is printed at the beginning. This is just the
expected behavior (_all_ bar numbers should be visible), but due to
slightly unclear documentation, users may have used these settings
to print bar numbers in the middle of systems except for the first
bar number. In such cases, simply remove '\set
Score.barNumberVisibility = #all-bar-numbers-visible' since
'\override BarNumber.break-visibility = ##t' does the relevant
setting alone.
. The '\break' command now always inserts a break, bypassing all
default decisions about break points. For example, it is no longer
necessary to insert '\bar ""' to obtain a mid-measure break.
The new '\allowBreak' command inserts a possible break point,
without forcing it, but bypassing default decisions like '\break'
does.
. The bar line type '"-"' has been removed. 'convert-ly' converts it
to '""'. There is a slight difference in horizontal spacing at
line breaks.
. 'automaticBars' has been removed. 'convert-ly' converts
'automaticBars = ##f' to 'measureBarType = #'()'.
. '\defineBarLine' now accepts '#t' in lieu of repeating the mid-line
glyph name.
. 'Bar_engraver' used to forbid line breaks between bar lines in all
cases, but now it only does so when the
'forbidBreakBetweenBarLines' context property is set to '#t', which
is the default. The 'barAlways' context property, which previously
worked around the lack of 'forbidBreakBetweenBarLines', has been
removed.
. Due to changes in the internals of '\bar', it is no longer
supported to use it before creating lower contexts with '\new'.
Such uses will now create an extra staff. This is similar to what
happens with commands such as '\override Staff...' (see *note
(lilypond-usage)An extra staff appears::).
The solution is to place '\bar' inside the music for each staff, as
is usual with most commands.
. The bar type '"-span|"' creates a mensurstrich.
. 'Staff' contexts use the new 'Caesura_engraver' to notate the
'\caesura' command.
Expressive mark improvements
============================
. Events attached to notes (e.g., dynamics or articulations) can be
delayed by an arbitrary duration using '\after'. This simplifies
many situations that previously required the use of explicit
polyphony and spacer rests.
. Broken hairpins now have some left padding by default. This is in
line with published scores and it fixes some cases where broken
hairpins were vertically displaced by the key signature.
. The ends of hairpins may now be aligned to the 'LEFT', 'CENTER' or
'RIGHT' of 'NoteColumn' grobs by overriding the property
'endpoint-alignments'.
. The direction of a trill spanner can now be set with direction
indicators like other articulations, i.e. with '_\startTrillSpan'
or '^\startTrillSpan'.
. The default appearance of trill spanners has changed to better
match classical engraving conventions. They now end before the
next note, not over it. If the next note has an accidental, they
stop before it. If the next note is the first note of a measure,
they stop over the bar line instead.
. The padding by default for fermatas is now larger. This avoids
some cases where the fermata was placed too close to dots and other
objects.
. The flageolet symbol is now smaller and slightly thicker. This is
in line with published scores and makes the recommended workaround
to make it smaller ('\tweak font-size -3 \flageolet') unnecessary.
. The accent glyph is now a bit smaller. This fixes some cases where
a natural sign would vertically displace accents.
. The comma glyph shape, as used in the '\breathe' command, has been
changed to a more common form.
The old glyph remains available under the name 'raltcomma'.
. The new context property 'breathMarkType' selects the mark that
'\breathe' produces from several predefined types.
Repeat improvements
===================
. Repeat alternatives may appear within the repeated section.
. The volta numbers for repeat alternatives may be set with the
'\volta' command.
. The new '\repeat segno' command automatically notates a variety of
_da-capo_ and _dal-segno_ forms.
. The new '\fine' command inserts a final bar line that interacts
gracefully with repeat bar lines. Used inside '\repeat', it also
prints _Fine_ and ends the music after unfolding.
. The '\volta' command removes music when a repeat is unfolded.
. The '\unfolded' command adds music when a repeat is unfolded.
Editorial annotation improvements
=================================
. The new '\staffHighlight' and '\stopStaffHighlight' commands can be
used to color a musical passage.
. A new grob 'FingerGlideSpanner' is now available, indicating a
finger gliding on a string from one to another position. Several
appearances are possible, depending on the setting of 'style'.
Shown in the image are 'line', 'stub-left', 'stub-right' and
'stub-both'.
Also possible is 'dashed-line', 'dotted-line', 'zigzag', 'trill',
'bow' and 'none'.
. Balloons now have changeable formatting.
. Parenthesizing chords is supported. Currently, the font size of
the parentheses has to be adjusted manually.
. Parenthesizing spanners is supported.
. A "time-based" version of the '\parenthesize' command was added.
It takes a grob path: '\parenthesize GROBNAME' or '\parenthesize
CONTEXTNAME.GROBNAME'. It acts like a '\once \override'. This
interface complements the already existing form '\parenthesize
EVENT', in a fashion similar to '\footnote'.
. Adding the 'Melody_engraver' to the 'Voice' context now works out
of the box to change the stem direction of the middle note
according to the melody. Previously, this required an additional
override to 'Stem.neutral-direction'.
The 'suspendMelodyDecisions' context property may be used to turn
off this behavior temporarily, as '\override Stem.neutral-direction
= #DOWN' used to do.
. The new 'Mark_tracking_translator' takes over from 'Mark_engraver'
the decision of when to create a mark. 'Mark_engraver' continues
to control formatting and vertical placement.
By default, 'Mark_engravers' in multiple contexts create a common
sequence of marks. If independent sequences are desired, multiple
'Mark_tracking_translators' must be used.
Text improvements
=================
. New commands '\textMark' and '\textEndMark' are available to add an
arbitrary piece of text between notes, called a text mark. These
commands improve over the previously existing syntax with the
'\mark' command called as '\mark MARKUP' (i.e., '\mark "..."' or
'\mark \markup ...').
'\textMark' and '\textEndMark' are now the recommended way to
create textual marks. The use of '\mark' for this purpose is still
supported, but discouraged (note that the '\mark' command itself is
not discouraged, only calling it on a markup argument; '\mark
\default' or '\mark NUMBER' is still the recommended and only way
to create a rehearsal mark).
The new commands have several differences to '\mark MARKUP'. There
can be an arbitrary number of them at a given moment, while there
can only be one use of '\mark'. They output grobs of the dedicated
'TextMark' type, whereas '\mark' creates a 'RehearsalMark' grob
regardless of whether it is called for a rehearsal mark or a
textual mark; introducing this distinction allows stylesheets to
set different layout settings for rehearsal marks and text marks.
The alignment set by the new commands is different: '\textMark'
always creates a left-aligned mark, and '\textEndMark' creates a
right-aligned mark; in contrast, the alignment of a 'RehearsalMark'
depends on the anchor point of the object it aligns to.
See *note (lilypond-notation)Text marks:: for full details.
. Text variant glyphs for sharp, flat, natural, double sharp, and
double flat are now available in the Emmentaler fonts. In markup,
they can be easily accessed with standard Unicode values.
. It is now possible to control the width and the shape of (some)
Emmentaler digits using OpenType features.
. '\smallCaps' now works on any markup, not just on a bare string.
. The syntax for conditions in markups was made more flexible and
user-friendly. It uses the new markup commands '\if' and
'\unless'. Here are example replacements:
2.22 syntax 2.24 syntax
--------------------------------------------------------------------------------
'\on-the-fly #first-page ...' '\if \on-first-page ...'
'\on-the-fly #not-part-first-page ...' '\unless \on-first-page-of-part ...'
'\on-the-fly #(on-page N) ...' '\if \on-page #N ...'
. With the new markup list command 'string-lines' it is now possible
to split a string at a given character. The default is to split at
line break. Surrounding white space gets dropped. The resulting
list of markups may be further formatted. This is a very
convenient way of entering additional stanzas for songs.
. The new markup command '\align-on-other' translates a markup as if
it was aligned to another markup.
. Two new markup functions '\with-dimension' and
'\with-dimension-from' are available. They are similar to
'\with-dimensions' and '\with-dimensions-from', respectively,
modifying only a single dimension (instead of both).
. New markup commands '\with-true-dimension' and
'\with-true-dimensions' are available. They give the markup the
actual extent(s) of its printed ink, which may differ from the
default extents for some font glyphs due to text regularity
constraints.
. Text replacements can now replace strings with any markup, not just
with a string.
. A new markup command '\with-string-transformer' is available. It
interprets a markup with a "string transformer" installed; the
transformer is called when the interpretation of the markup
requires interpreting a string, and allows to perform modifications
on this string, such as changing case.
. The 'markup->string' function converts a markup into an approximate
string representation; it is used for outputting PDF metadata as
well as MIDI lyrics and markers. Markup commands can now define a
custom method to convert markups created using them into strings,
for use by 'markup->string'. For example:
#(define-markup-command (upcase layout props arg) (string?)
#:as-string (string-upcase arg)
(interpret-markup layout props (string-upcase arg)))
New for specialist notation
***************************
Fretted string instrument improvements
======================================
. The string tunings 'banjo-double-c' and 'banjo-double-d' were
added.
. A new grob 'BendSpanner' is now available for 'TabStaff',
indicating a bent string. Apart from the default three styles are
possible: ''hold', ''pre-bend' and ''pre-bend-hold'.
Percussion improvements
=======================
. The drum notation style 'weinberg-drums-style' was added. It is
based on Norman Weinberg's standardization work.
Wind instrument improvements
============================
. Additional display details of a '\woodwind-diagram' can now be
specified including the angle of partially-covered-keys and the
display of non-graphical trill keys.
Chord notation improvements
===========================
. Support for chord grids has been added.
. In 'ChordNames', multi-measure rests now also cause the "N.C."
symbol to be printed, just like normal rests.
. In figured bass, '_' now creates an empty figure that still takes
up space.
. Formatting of figured bass has been improved. In particular, the
default size is reduced to a value used by many Urtext editions of
Baroque music.
. In figured bass, specially designed glyphs for '6\\', '7\\', and
'9\\' are now used by default. Similarly, specially designed
glyphs for symbols '2\+', '4\+', and '5\+' are used by default if
plus signs appear after the number.
Use the new command '\figured-bass' to access these glyphs in
markup.
. In figured bass, brackets can now also be added around accidentals.
Ancient notation improvements
=============================
. A new context 'VaticanaLyrics' is available. It is similar to
'Lyrics', providing a hyphenation style (a single, flush-left
hyphen between two syllables) as used in the notational style of
Editio Vaticana.
. The predefined commands for Gregorian divisiones are no longer
variations on '\breathe'. '\divisioMinima', '\divisioMaior',
'\divisioMaxima', and '\virgula' are variations on the basic
'\caesura'. '\finalis' is equivalent to '\section'.
'MensuralStaff' and 'VaticanaStaff' use 'Divisio_engraver' to
interpret the above commands as well as '\repeat volta' and
'\fine'.
. 'KievanStaff', 'MensuralStaff', 'PetrucciStaff', and
'VaticanaStaff' now allow line breaks anywhere, and they no longer
create '""' measure bar lines.
. In 'GregorianTranscriptionStaff', divisiones are now engraved as
'BarLine' grobs by default. To change them to 'Divisio' grobs, use
'\EnableGregorianDivisiones'.
. 'GregorianTranscriptionStaff' allows a line break after any note
and no longer uses 'Time_signature_engraver'.
. 'GregorianTranscriptionVoice' no longer uses 'Stem_engraver'.
World music improvements
========================
. Support for Persian classical music is now available. For this,
two accidental glyphs, _sori_ and _koron_, have been added to
LilyPond.
Miscellaneous improvements
**************************
. In the Emmentaler font, identical-looking noteheads whose only
difference was stem direction have been consolidated into a single
glyph. For instance, the glyphs 'noteheads.u2triangle' and
'noteheads.d2triangle' have been replaced by a single glyph,
'noteheads.s2triangle'. Notehead pairs that look visually
different depending on the direction remain distinct.
In addition, the 'stem-attachment' property of 'NoteHead' grobs now
returns its actual, direction-dependent stem attachment point
instead of a hypothetical upwards-stem attachment point.
. Two redundant glyphs in the Emmentaler font have been removed:
'scripts.trillelement' (use 'scripts.trill_element' instead) and
'scripts.augmentum' (use 'dots.dotvaticana' instead).
. Using '\paper { bookpart-level-page-numbering = ##t}', it is now
possible to make bookparts independent with respect to page
numbering. If this is used for all bookparts, each bookpart has
its own numbering sequence, starting at 1 by default. It can also
be used in an individual bookpart, which is useful to achieve the
standard practice of numbers pages in an analytical introduction
independently and in roman numerals (the latter is achieved using
'page-number-type = #'roman-lower').
. A new grob callback function 'break-alignment-list' is now
available for returning different values depending on a grob's
break direction. As an example, use it to provide different
alignments of a grob depending on whether it is positioned at the
beginning, the middle, or the end of a line.
. The new 'Mark_performer' creates MIDI Marker events like
'Mark_engraver' creates printed marks.
. Properties of 'PaperColumn' and 'NonMusicalPaperColumn' (such as
'NonMusicalPaperColumn.line-break-system-details') can now be
overridden mid-music with the usual command '\once \override'.
They used to be a special case requiring the '\overrideProperty'
command.
. The new 'show-horizontal-skylines' and 'show-vertical-skylines'
properties allow to display an object's skylines. This is more
flexible than the already existing 'debug-skylines' option because
it works for all grobs. While primarily meant for debugging
LilyPond, this can be useful when trying to understand spacing
decisions or overriding stencils in Scheme.
. The new command '\vshape' is like '\shape', but also shows the
control points and polygon for easier tweaking.
. '\markup \path' now also works in SVG output even if the path does
not begin with a 'moveto' or 'rmoveto' command. Also, it now
accepts single-letter SVG equivalents ('moveto' = 'M', etc.).
. 'set-default-paper-size' and 'set-paper-size' now accept a custom
paper size.
. 'lilypond-book' supports two new music fragment options
'paper-width' and 'paper-height' to set a custom paper size.
. 'lilypond-book' supports a new snippet option 'inline' for inline
music, that is, music snippets like [[image of music]] that appear within a paragraph
of text.
. The 'lilypond-book' script now allows braces in the argument of the
commands '\lilypond' (for LaTeX) and '@lilypond' (for Texinfo).
. 'lilypond-book' now appends the current directory as the last entry
to search for included files, instead of prepending it to the list
of specified include paths. This allows include directories to
shadow files from the current directory, and will only be noticed
if there are files with the same name in both.
. The new Scheme function 'universal-color' provides an eight-element
color palette designed to be unambiguous to people with
dichromatism.
. The '-dembed-source-code' option now also embeds images added with
'\epsfile' and files included with '\verbatim-file'.
. The default of the 'aux-files' program option changed to '#f'. If
you are calling LilyPond with the '-dbackend=eps' argument and need
the auxiliary '.tex' and '.texi' files, you now have to specify
'-daux-files' explicitly. The formats for 'lilypond-book' images
can be set separately for the tall page image (typically PNG for
HTML output) and per-system images (typically, EPS or PDF for
printed output) with the '-d' sub-options '-dtall-page-formats' and
'-dseparate-page-formats' respectively.
. The 'big point' unit (1bp = 1/72in) is now available by appending
'\bp' to length values.
. Scheme-defined translators usable in both '\layout' and '\midi' can
now be created with 'make-translator'. Scheme-defined performers
usable only in '\midi' can now be created with 'make-performer'.
Those macros work strictly like the previously existing macro
'make-engraver' for creating engravers only usable in '\layout'.
. Scheme translators can now define a new slot called
'pre-process-music'. It is called on all translators, after all
listeners but before all 'process-music' slots. This can be used
for processing that depends on all events heard but needs to set
context properties before other translators read them.
. Scheme translators can now contain listeners written as
(listeners
((event-class engraver event #:once)
...))
These are never triggered more than once per time step. They emit
a warning if they receive two events in the same time step, except
if the events are equal.
. The same grob definition can now be used to create grobs of
different classes ('Item', 'Spanner', 'Paper_column', 'System').
As part of this change, the grob types 'FootnoteItem' and
'FootnoteSpanner' were consolidated into a single type 'Footnote'.
Similarly, 'BalloonTextSpanner' and 'BalloonTextItem' are unified
into 'BalloonText'.
When the grob definition does not mandate a class, engravers should
choose what class to create a grob with. For authors of Scheme
engravers, this means using either 'ly:engraver-make-item' or
'ly:engraver-make-spanner'. The utility function
'ly:engraver-make-sticky' is provided to support the frequent case
of _sticky_ grobs, such as footnotes and balloons. It creates a
grob with the same class as another grob and administrates parents
and bounds.
. The new command-line option '-dcompile-scheme-code', also settable
in the LilyPond input with '#(ly:set-option 'compile-scheme-code)',
provides with better diagnostics when running Scheme code leads to
an error. Internally, this uses the byte-compiler provided by
Guile, instead of the interpreter.
However, due to a limitation in Guile, this currently has the
disadvantage of making it impossible to run more than a few
thousand Scheme expressions. Also, be aware that the Guile
compiler has a few differences to the interpreter. For example,
constant parts of quasiquotes are made actual constants more
aggressively, making code such as '(let ((x 4)) (sort! `(,x 3 2
1)))' produce an error because the "cdr" of the quasiquoted list is
constant, and it is an error in Scheme to mutate literal data. (In
this specific case, the code could avoid the issue by using the
non-destructive 'sort', or by creating a fresh list each time with
'(list x 3 2 1)'.)
Furthermore, this option does not currently work on Windows.
Notes on Guile 2.2
******************
This version of LilyPond switches from Guile 1.8 to Guile 2.2. This
section lists some of the most common incompatibilities that you could
have to deal with in order to upgrade your Scheme code.
A full, detailed log of changes in Guile can be found in the 'NEWS'
file (https://git.savannah.gnu.org/cgit/guile.git/tree/NEWS) of the
Guile source.
. The 'format' function now requires a boolean or port as the first
argument. This argument was optional in Guile 1.8. In order to
make the function return the formatted output as a string, like
'format' does without this argument in Guile 1.8, pass '#f' for
this argument, i.e., '(format #f "STRING" ARGUMENTS ...)' instead
of '(format "STRING" ARGUMENTS ...)'.
. The rules for internal (i.e., non-toplevel) definitions have become
stricter. Definitions are no longer allowed in various expression
contexts. This is no longer valid, for example:
(if (not (defined? 'variable))
(define variable 'value))
The solution in this particular example is:
(define variable
(if (not (defined? 'variable))
'value
variable))
. Strings now support Unicode characters. Previously, a Unicode
character was represented by several characters, and various
functions were not tailored for Unicode support.
. Some numeric functions now return exact results in more cases. For
instance, '(sqrt 4)' returns '2.0' in Guile 1.8, but '2' (an
integer) in Guile 2.2.
2.22.2:
New for musical notation
------------------------
Displaying pitch improvements
.............................
. An accidental glyph corresponding to SMuFL code U+E2E3 has been
added (Extended Helmholtz-Ellis accidentals: Raise by one undecimal
quartertone).
. Quarter-tone note names are now available in all the input
languages.
. Setting 'suggestAccidentals' to ''cautionary' turns only cautionary
accidentals into 'AccidentalSuggestion's. This can be used to
distinguish between facsimile and editorial accidentals.
. The command '\ambitusAfter' has been added. It is used to move the
ambitus to a different position.
. Ottava brackets are now introduced with a single number by default
(e.g., '8' or '15'), and printed in bold. A new
'ottavationMarkups' property has been introduced to modify that
behavior, with several predefined lists of markups available.
Rhythm improvements
...................
. Tuplets can now be printed with slurs instead of brackets.
. French beaming ('\override Stem.french-beaming = ##t') now
_exactly_ behaves like standard (i.e., default) beaming in every
respect (beam positioning and placement of any articulation,
fingering, etc.). The only remaining difference are inner stems
not passing through beams.
. Swing and irregular rhythmic patterns may now be applied to music
expressions made of regular durations, which may be used to render
inequal rhythmic interpretation in MIDI.
. Font glyphs for 256th, 512th, and 1024th flags and rests have been
added.
. The new 'Merge_mmrest_numbers_engraver' hides duplicate numbers on
multi-measure rests.
Expressive mark improvements
............................
. A dynamic command '\n' for _niente_ has been added.
. Two new ornaments have been added.
. A very short fermata and Henze variants of long and short fermatas
have been added.
. Fermatas and other articulations can now be added directly to
multi-measure rests. Therefore the command '\fermataMarkup' is
deprecated.
Editorial annotation improvements
.................................
. The 'New_fingering_engraver' will now consider the
'X-align-on-main-noteheads' property of the
'self-alignment-interface'. If set to true ('##t'), all fingerings
oriented 'up' or 'down' will be arranged in a straight column and
aligned on the noteheads on the correct side of the stem.
Text formatting improvements
............................
. The default fonts are now those from the URW Core 35 set, version
2.0, replacing the TeX Gyre fonts. Note that the fonts come with a
ligature for 'Nr.'; see *note (lilypond-notation)Fonts explained::
for solutions to avoid it locally and globally.
. The '\note' markup command now takes as its first argument a
duration instead of a string.
New for specialist notation
---------------------------
Vocal music improvements
........................
. A lyric hyphen may now be repeated at the start of a system
beginning with a new syllable.
. A gradual change of vowel (or sustained consonant) may be indicated
by adding a vowel transition between lyric syllables with the
command '\vowelTransition'.
Unfretted and fretted string instrument improvements
....................................................
. Fret-diagrams may now be printed left-handed, setting 'handedness'
to 'LEFT' (as a subproperty of 'fret-diagram-details').
. Some ukulele fretboard-diagrams have been fixed, and new diagrams
have also been added.
Chord notation improvements
...........................
. Chords may now be automatically inverted or voiced with dropped
notes.
. Legacy chord naming functions 'banter-chord-names' and
'jazz-chord-names', have been removed from the main codebase, as
have been the properties 'chordNamesExceptionsFull' and
'chordNamesExceptionsPartial'. As part of that rewrite, power
chords are now included in the default exceptions; they will be
printed correctly (with a superscript like all other chords) with
no additional tweaks required, thereby removing the need for the
'\powerChords' predefined command. Anyone interested in the legacy
functions may find a (somewhat working) copy of them in the
'chord-names-alternative.ly' snippet.
Contemporary music improvements
...............................
. A new grob 'DurationLine' is now available. It continues a
rhythmic grob with a line, ending at the next rhythmic grob.
Possible styles are ''beam', ''line', ''dashed-line',
''dotted-line', ''zigzag', ''trill' and ''none'. The duration line
may end with a hook (beam-style only) or an arrow.
New for input and output
------------------------
Input file improvements
.......................
. As announced in version 2.17.3 nearly eight years ago, the
''relative-includes' option is now enabled by default; included
files that contain an '\include' command of their own must account
for their own path rather than the main file's directory. That
behavior may however be switched off by setting
''relative-includes' to '#f', either as a command line option or
using 'ly:set-option' in source files.
. '\compressFullBarRests' has been renamed to
'\compressEmptyMeasures', to avoid possible confusion with
'\compressMMRests'. Likewise, '\expandFullBarRests' has now become
'\expandEmptyMeasures'.
. The '\partcombine' command, as well as all 'partCombine'-prefixed
commands, subroutines and property names, are now written with a
capital C, such as '\partCombine', '\partCombineApart' etc. The
'\autochange' command is now also capitalized as '\autoChange'.
. All input languages ('\language' statement) can be entered using
their proper UTF-8 spelling (i.e., including special characters).
Output improvements
...................
. Skylines will now take account of the 'rotation' property of layout
objects. Rotating a crescendo hairpin by applying '\override
Hairpin.rotation = #'(15 0 0)', for instance, will now actually
have influence on the skylines and thus help to enhance spacing.
. Slight padding added between natural glyphs just touching at the
corners in key cancellations.
. Skylines of boxes now reflect the actual box outline including
rounded corners and rotation.
. CSS-style colors can now be used directly as text strings; either
with predefined color names (like with the already available
'css-color' function), or with hexadecimal color codes prefixed
with '#'. All stencils that support a 'color' property now accept
either a list or a string; in the latter case, that string is used
directly in the SVG output. This allows to use alpha transparency
(entered as "#RRGGBBAA" or "#RGBA") in SVG.
. PDF bookmarks are now supported and allow for '\tocItem' entries to
appear in the 'table of contents' panel of PDF viewers that support
it.
. '\table-of-contents' now accepts a hierarchical structure;
'\tocItem' entries may optionally take a symbol (like '\label') or
a dot-separated list of symbols, indicating their position in the
score's structure. A side-effect of that feature is that
'\tocItem' can no longer take a simple string as its argument; a
'\markup' command must be used.
. Using the new options '-dpng-width' and '-dpng-height' it is now
possible to specify the pixel size of PNG output images.
. SVG output is now available through the '--svg' command-line option
(or its canonical form '--format=svg', also shortened as '-fsvg').
Due to its backend's specificity, that option is not yet compatible
with other output formats; to get a file in both SVG and PDF, PNG
or EPS, a second LilyPond run is required.
2.20:
New for musical notation
------------------------
*Displaying pitch improvements*
. Pitches that have a sharp or flat in their name now need to be
hyphenated.
Pitches that contain _double_ sharps or flats in their name,
however, do not need a second hyphen.
. Accidental rules can now be defined _across_ 'ChoirStaff' contexts.
. Two new accidental rules have been added. Both combine the
characteristics of 'modern-voice', 'piano' and their equivalents:
'choral'
This is the now the default accidental style for 'ChoirStaff'.
'choral-cautionary'
The same as 'choral' but with the extra accidentals typeset as
cautionaries instead.
. Four new clef glyphs are now available; 'GG' (double-G), 'Tenor G',
'varC' plus related tessitura and 'Varpercussion'.
. French note names are now explicitly defined -- previously they were
aliased to Italian note names.
Double sharps are entered using an 'x' suffix.
*Rhythm improvements*
. Multi-measure rests have length according to their total duration,
under the control of 'MultiMeasureRest.space-increment'. Note the
default value is '2.0'.
. Improvements to the '\partial' command have been made when used
with parallel music and/or multiple contexts.
. It is now possible to change the time signature mid-measure by
using both the '\time' and '\partial' commands together.
. Isolated durations in music now stand for unpitched notes. Pitches
are taken from the preceding note or chord. This is especially
convenient for specifying rhythms in both music and scheme
functions and can help improve the readability of LilyPond source
files.
. Beaming exceptions can now be constructed using the simpler
'\beamExceptions' scheme function. Note that writing the
exception pattern without pitches is convenient but not
mandatory (also see the previous documented rhythm improvement --
_Isolated durations in music now stand for unpitched notes_.
. The positioning of tuplet numbers for kneed beams has been
improved. Previously, tuplet numbers were placed according to the
position of the tuplet bracket, even if the bracket was not
printed. This could lead to tuplet numbers being 'stranded'.
. Collision detection for the kneed beam tuplet numbers has also been
added, shifting the offset horizontally if the number is too close
to an adjoining note column (but still preserving the number's
vertical distance). In the event of a collision -- for example with
an accidental -- the tuplet number will be shifted vertically
instead. If the tuplet number is itself too large to fit within
the available space, the original, 'bracket-based', positioning
system will be used instead.
The original kneed-beam tuplet behavior is still available with a
new, 'knee-to-beam' property for the 'TupletNumber' layout object.
*Expressive mark improvements*
. The ends of hairpins may now be fine-tuned using the 'shorten-pair'
grob property. This previously only affected text-spanners (e.g.
'TupletBracket' and 'OttavaBracket').
Positive and negative values offset right and left respectively.
. Individual slurs and phrasing slurs may now be started from an
explicit note within a chord.
. A new command '\=X' has been added -- where 'X' can be any
non-negative integer or symbol -- so that a specific 'id' can be
assigned to the start and end of slurs and phrasing slurs.
This is useful when simultaneous slurs are required or if one slur
overlaps another or when nesting short slurs within a longer one.
*Repeat notation improvements*
. The visual style of tremolo slashes (shape, style and slope) is now
more finely controlled.
. The music function '\unfoldRepeats' can now take an optional
argument-list specifying which type(s) of repeated music should be
unfolded. Possible entries are 'percent', 'tremolo', 'volta'. If
the optional argument-list is unspecified, 'repeated-music' will be
used, unfolding all.
*Staff notation improvements*
. A new command '\magnifyStaff' has been added which scales staff
sizes, staff lines, bar lines, beamlets and horizontal spacing
generally at the 'Staff' context level. Staff lines are prevented
from being scaled smaller than the default since the thickness of
stems, slurs, and the like are all based on the staff line
thickness.
. A new command '\magnifyMusic' has been added, which allows the
notation size to be changed without changing the staff size, while
automatically scaling stems, beams, and horizontal spacing.
. A new command, '\RemoveAllEmptyStaves', has been made available,
which acts exactly like '\RemoveEmptyStaves', except for also
removing empty staves on the first system in a score.
. A new markup command '\justify-line' has been added. Similar to
the '\fill-line' markup command except that instead of setting
_words_ in columns, the '\justify-line' command balances the
whitespace between them ensuring that when there are three or more
words in a markup, the whitespace is always consistent.
*Editorial annotation improvements*
. It is now possible to add text to analysis brackets through the
'HorizontalBracketText' object.
*Text formatting improvements*
. Support for making it easier to use alternative 'music' fonts other
than the default Emmentaler in LilyPond has been added. See *note
(lilypond-notation)Replacing the notation font:: for more
information.
. Default text fonts have been changed from 'Century Schoolbook L',
'sans-serif', and 'monospace'.
For 'svg' backend:
Family Default font
-----------------------------------
_roman_ 'serif'
_sans_ 'sans-serif'
_typewriter_'monospace'
'serif', 'sans-serif', and 'monospace' are 'generic-family' in SVG
and CSS specifications.
For other backends:
Family Default font (alias) Alias definition lists
----------------------------------------------------------------------------
_roman_ 'LilyPond Serif' TeX Gyre Schola, C059, Century
SchoolBook URW, Century Schoolbook L,
DejaVu Serif, ..., serif
_sans_ 'LilyPond Sans TeX Gyre Heros, Nimbus Sans, Nimbus
Serif' Sans L, DejaVu Sans, ..., sans-serif
_typewriter_'LilyPond Monospace' TeX Gyre Cursor, Nimbus Mono PS,
Nimbus Mono, Nimbus Mono L, DejaVu
Sans Mono, ..., monospace
'LilyPond Serif', 'LilyPond Sans Serif', and 'LilyPond Monospace'
are font aliases defined in the LilyPond dedicated FontConfig
configuration file '00-lilypond-fonts.conf'. Where a character
dosen't exist in the first font listed, the next font listed will
be used instead for that character. For details of alias
definitions, please see to '00-lilypond-fonts.conf' under the
installed directory.
. When using OpenType fonts, font features can be used. Note: Not
all OpenType fonts have all functions.
. Two new styles of whiteout are now available. The 'outline' style
approximates the contours of a glyph's outline, and its shape is
produced from multiple displaced copies of the glyph. The
'rounded-box' style produces a rounded rectangle shape. For all
three styles, including the default 'box' style, the whiteout
shape's 'thickness', as a multiple of staff-line thickness, can be
customized.
. A new markup-command, '\with-dimensions-from', makes
'\with-dimensions' easier to use by taking the new dimensions from
a markup object, given as first argument.
. Markup-command '\draw-squiggle-line' is now available. Customizing
is possible with overrides of 'thickness', 'angularity', 'height'
and 'orientation'
. Markup-commands '\undertie' and '\overtie' are now available, as
well as the generic markup-command '\tie'.
New for specialist notation
---------------------------
*Vocal music improvements*
. A new flexible template suitable for a range of choral music, is
now provided. This may be used to create simple choral music, with
or without piano accompaniment, in two or four staves. Unlike
other templates, this template is 'built-in', which means it does
not need to be copied and edited: instead it is simply '\include''d
in the input file. For details, see *note
(lilypond-learning)Built-in templates::.
. The '\addlyrics' function now works with arbitrary contexts
incuding 'Staff'.
. '\lyricsto' and '\addLyrics' have been 'harmonized'. Both now
accept the same kind of delimited argument list that '\lyrics' and
'\chords' accept. Backward compatibility has been added so music
identifiers (i.e. '\mus') are permitted as arguments. A
'convert-ly' rule has been added that removes redundant uses of
'\lyricmode' and rearranges combinations with context starters such
that '\lyricsto' in general is applied last (i.e. like
'\lyricmode' would be).
*Unfretted and fretted string instrument improvements*
. A new note head style for Tabulature has been added --
'TabNoteHead.style = #'slash'.
. In fret-diagrams the distance between frets and the distance
between strings is now independently adjustable. Available are
'fret-distance' and 'string-distance' as subproperties of
'fret-diagram-details'.
. It is now possible to individually color both the dots and
parentheses in fret diagrams when using the '\fret-diagram-verbose'
markup command.
. Two new properties have been added for use in
'fret-diagram-details' when using the '\fret-diagram-verbose'
markup command; 'fret-label-horizontal-offset' which affects the
'fret-label-indication' and 'paren-padding' which controls the
space between the dot and the parentheses surrounding it.
. Additional bass strings (for lute tablature) are supported.
. String numbers can now also be used to print roman numerals (e.g.
for unfretted string instruments).
. 'TabStaff' is now able to print micro-tones for bendings etc.
*Chord notation improvements*
. '\chordmode' can now use '< >' and '<< >>' constructs.
. It is now possible to override the 'text' property of chord names.
New for input and output
------------------------
*Input structure improvements*
. Blocks introduced with '\header' can be stored in variables and
used as arguments to music and scheme functions and as the body of
'#{...#}' constructs. They are represented as a Guile module.
While '\book', '\bookpart', '\score', '\with', '\layout', '\midi',
'\paper' blocks can be passed around in similar manner, they are
represented by different data types.
*Titles and header improvements*
. Page numbers may now be printed in roman numerals, by setting the
'page-number-type' paper variable.
*Input file improvements*
. A new command '\tagGroup' has now been added. This complements the
existing '\keepWithTag' and '\removeWithTag' commands.
*Output improvements*
. LilyPond source files may now be embedded inside the generated PDF
files. This experimental feature is disabled by default and may be
regarded as unsafe, as PDF documents with hidden content tend to
present a security risk. Please note that not all PDF viewers have
the ability to handle embedded documents (if not, the PDF output
will appear normally and source files will remain invisible). This
feature only works with the PDF backend.
. The 'output-classic-framework' procedure and the '-dclip-systems'
are now available with the 'SVG' backend.
. An argument, '-dcrop', has been added, formatting 'SVG' and 'PDF'
output without margins or page-breaks.
. A new 'output-attributes' grob property is now used for svg output
instead of the 'id' grob property. It allows multiple attributes
to be defined as an association list. For example, '#'((id . 123)
(class . foo) (data-whatever . "bar"))' will produce the following
group tag in an SVG file: '<g id="123" class="foo"
data-whatever="bar"> ... </g>'.
. The PostScript functionality of stroke adjustment is no longer
applied automatically but left to the discretion of the PostScript
device (by default, Ghostscript uses it for resolutions up to
150dpi when generating raster images). When it is enabled, a more
complex drawing algorithm designed to benefit from stroke
adjustment is employed mostly for stems and bar lines.
Stroke adjustment can be forced by specifying the command line
option '-dstrokeadjust' to LilyPond. When generating 'PDF' files,
this will usually result in markedly better looking 'PDF' previews
but significantly larger file size. Print quality at high
resolutions will be unaffected.
. Added a new 'make-path-stencil' function that supports all 'path'
commands both relative and absolute:
'lineto', 'rlineto', 'curveto', 'rcurveto', 'moveto', 'rmoveto',
'closepath'. The function also supports 'single-letter' syntax
used in standard SVG path commands:
'L', 'l', 'C', 'c', 'M', 'm', 'Z' and 'z'. The new command is also
backward-compatible with the original 'make-connected-path-stencil'
function. Also see 'scm/stencil.scm'.
*MIDI improvements*
. The most common articulations are now reflected in MIDI output.
Accent and marcato make notes louder; staccato, staccatissimo and
portato make them shorter. Breath marks shorten the previous note.
This behavior is customizable through the 'midiLength' and
'midiExtraVelocity' properties on 'ArticulationEvent'. See
'script-init.ly' for examples.
. Improved MIDI output for breathe marks. After tied notes, breaths
take time _only_ from the last note of the tie; e.g. '{ c4~ c8
\breathe }' performs as '{ c4~ c16 r }' instead of '{ c4 r8 }'.
This is more consistent with articulations and how humans interpret
breaths after ties. It now also makes it easier to align
simultaneous breathe marks over multiple parts, all with different
note lengths.
. There is now support for controlling the 'expression level' of MIDI
channels using the 'Staff.midiExpression' context property. This
can be used to alter the perceived volume of even sustained notes
(albeit in a very 'low-level' way) and accepts a number value
between '0.0' and '1.0'.
. When outputting MIDI, LilyPond will now store the 'title' defined
in a score's '\header' block (or, if there is no such definition on
the '\score' level, the first such definition found in a '\header'
block of the score's enclosing '\bookpart', '\book', or top-level
scope) as the name of the MIDI sequence in the MIDI file.
Optionally, the name of the MIDI sequence can be overridden using
the new 'midititle' '\header' field independently of 'title' (for
example, in case 'title' contains markup code which does not render
as plain text in a satisfactory way automatically).
. Support for making it easier to use alternative 'music' fonts other
than the default Emmentaler in LilyPond has been added. See *note
(lilypond-notation)Replacing the notation font:: for more
information.
*Extracting music improvements*
. '\displayLilyMusic' and its underlying Scheme functions no longer
omit redundant note durations. This makes it easier to reliably
recognize and format standalone durations in expressions like
{ c4 d4 8 }
New for spacing issues
----------------------
*Page breaking improvements*
. There are two new page breaking functions. 'ly:one-page-breaking'
automatically adjusts the height of the page to fit the music, so
that everything fits on one page.
'ly:one-line-auto-height-breaking' is like 'ly:one-line-breaking',
placing the music on a single line and adjusting the page width
accordingly, however it also automatically adjusts the page height
to fit the music.
*Vertical and Horizontal spacing improvements*
. It is now possible to move systems with reference to their current
position using the 'extra-offset' subproperty of
'NonMusicalPaperColumn.line-break-system-details'. Both vertical
and horizontal changes are possible. This feature is especially
useful for making slight adjustments to the default vertical
position of individual systems. See *note
(lilypond-notation)Explicit staff and system positioning:: for more
information.
. Improved visual spacing of small and regular 'MI' Funk and Walker
noteheads so they are now the same width as other shaped notes in
their respective sets. 'SOL' noteheads are also now visually
improved when used with both the normal Aiken and Sacred Harp
heads, as well as with the thin variants.
. 'LeftEdge' now has a definable 'Y-extent' (i.e.vertical). See
*note (lilypond-internals)LeftEdge::.
. Grobs and their parents can now be aligned separately allowing more
flexibility for grob positions. For example the 'left' edge of a
grob can now be aligned on the 'center' of its parent.
. Improved horizontal alignment when using 'TextScript', with
'DynamicText' or 'LyricText'.
New for changing defaults
-------------------------
An optional argument for the '\afterGrace' command has been added.
'\afterGrace' now has an optional argument to specificy the spacing
fraction position of its notes.
. All of '\override', '\revert', '\set', and '\unset' now work with
the '\once' prefix for making one-time settings.
New for Internal interfaces and functions
-----------------------------------------
. The music and grob property 'spanner-id', used for distinguishing
simultaneous slurs and phrasing slurs, has been changed from a
string to a _key_ which can be either a non-negative integer or
symbol (also see the previous documented expressive mark
improvement -- _A new command \=X has been added_).
. Context properties named in the 'alternativeRestores' property are
restored to their value at the start of the _first_ alternative in
all subsequent alternatives.
. LilyPond functions defined with 'define-music-function',
'define-event-function', 'define-scheme-function' and
'define-void-function' can now be directly called from Scheme as if
they were genuine Scheme procedures. Argument checking and
matching will still be performed in the same manner as when calling
the function through LilyPond input. This includes the insertion
of defaults for optional arguments not matching their predicates.
Instead of using '\default' in the actual argument list for
explicitly skipping a sequence of optional arguments,
'*unspecified*' can be employed.
. Current input location and parser are now stored in GUILE fluids
and can be referenced via the function calls '(*location*)' and
'(*parser*)'. Consequently, a lot of functions previously taking
an explicit 'parser' argument no longer do so.
Functions defined with 'define-music-function',
'define-event-function', 'define-scheme-function' and
'define-void-function' no longer use 'parser' and 'location'
arguments.
With those particular definitions, LilyPond will try to recognize
legacy use of 'parser' and 'location' arguments, providing
backwards-compatible semantics for some time.
. Scheme functions and identifiers can now be used as output
definitions.
. Scheme expressions can now be used as chord constituents.
. Music (and scheme and void) functions and markup commands that just
supply the final parameters to a chain of overrides, music function
and markup command calls can now be defined in the form of just
writing the expression cut short with '\etc'.
. Dot-separated symbol lists like 'FretBoard.stencil' were already
supported as of version 2.18. They may now also contain unsigned
integers, and may alternatively be separated by commata.
. Such lists may also be used in expressions for assignments, sets,
and overrides.
. Association list elements could previously be assigned values
individually (for example, paper variables like
'system-system-spacing.basic-distance').
In combination with the previously mentioned changes, this allows
setting and referencing pseudovariables like 'violin.1'.
. The markup-list-command '\table' is now available. Each column may
be aligned differently.
. 'InstrumentName' now supports 'text-interface'.
. The 'thin-kern' property of the 'BarLine' grob has been renamed to
'segno-kern'.
. 'KeyCancellation' grobs now ignore cue clefs (like 'KeySignature'
grobs do).
. Add support for '\once \unset'
In IncrementalParser.cpp: NetBSD doesn't have ENOTRECOVERABLE, so
doesn't have std::errc::state_not_recoverable either. Use
std::errc::operation_not_supported instead.
Bump PKGREVISION.
This is a milestone release that marks version 1.0.0 of git-cliff.
It means that the API is somewhat stabilized and development will take a route
in improving the existing functionality and adding new features rather than
changing the internals. There are a couple of breaking changes in this release
(such as the removal of --date-order) so the major version is increased.
Bug Fixes
- Fix test fixture failures
Documentation
- Fix GitHub badges in README.md
Features
- [breaking] Replace --date-order by --topo-order
- Allow running with --prepend and --output (#120)
- [breaking] Use current time for --tag argument (#107)
- Include completions and mangen in binary releases (#115)
- Publish Debian package via release workflow (#113)
Miscellaneous Tasks
- Run all test fixtures
- Remove deprecated set-output usage
- Update actions/checkout to v3
- Comment out custom commit preprocessor (#112)
Refactor
- Apply clippy suggestions
Styling
- Update README.md about the styling of footer field
What's Changed
- build(deps): bump clap from 4.0.27 to 4.0.29 by @dependabot in #494
- chore: Update grammars by @afnanenayet in #501
- Update deps by @afnanenayet in #499
- Add related project by @afnanenayet in #500
- chore: Update dependencies by @afnanenayet in #508
- chore(grammars): Update grammars by @afnanenayet in #510
- docs(testing): Add note about why tests break by @afnanenayet in #511
- ci(windows): Add back windows target by @afnanenayet in #512
The class provides an environment for creating a fancily laid
out tabular curriculum vitae inspired by the european
curriculum vitae. The distribution comes with a German and an
English template.
nia
khorben
minskim
fcambus
schmonz
abs
he
pin
wiz