Releases · JuliaLang/julia

08 Nov 21:03

ararslan

v1.0.2

d789231

v1.0.2

This is the second patch release for Julia version 1.0. It contains bugfixes only, no new features or changes.

An automatically generated set of patch notes is shown below:

Bug fixes

#29205 - fix digits(n::Unsigned) with neg base for n > typemax(n)÷2
#29192 - fix reinterpret(Char, ::UInt32) for "unnatural" values (fix #29181)
#29322 - fix #29306, teach effect_free that GotoIfNot with non-Bool can throw
#29314 - fix bug in parsing extended colon-like operators
#29246 - Fix: Complex SubArray times real Matrix
#29217 - fix crash when interpreting optimized top-level expressions
#29416 - doc: fix not-true-anymore example with promote_rule
#29380 - fix #29208, incorrect Union{} result from type intersection
#29414 - fix sizeof(str::AbstractString) docstring
#29401 - Change from parametric T signature to Int32 in _ldexp_exp
#29419 - fix #20872, handle UnionAll types in which
#29325 - check return from jl_restore_incremental correctly
#29487 - fix #29468, bug in intersection of different-length vararg tuples
#29465 - Fix #29451: parse(Int, s::AbstractString) when ncodeunits(s) > lastindex(s)
#29412 - REPL: really fix #29347, searching a string not in history
#29373 - REPL: make paste work in search mode
#29561 - Fixing abs for Rational{<:Unsigned}
#29646 - fix #29644: getproperty of Tridiagonal{SparseVector}
#29663 - fix #29662: missing import from Base.Printf to Printf stdlib.
#29667 - Use a stable inverse for Tridiagonal and SymTridiagonal
#29712 - fix #29711: triu! for m > n
#29457 - Uniform scaling cat with zero dimensions
#29725 - Fix bug in nullspace when the matrix is empty
#29732 - convert Integer in Cholesky constructors to BlasInt
#29742 - Fix hashing of Dates.Time. Fixes #29480
#29722 - fix #29718, union field alignment
#29780 - adding _small_enough for diagonal so that fill! works
#29760 - Fix out-of-tree build of docs, fix #29757.
#29539 - fix #29528, specificity transitivity error in DiffEqBase
#29765 - fix eval world updates
#29878 - Reinstate statically computed log event ids
#29908 - Fix isvalid for 3-byte overlong encoded UTF-8 sequences

Performance improvements

#29352 - avoid conversion from Char to String to Symbol in some LinearAlgebra routines
#29346 - fix #29326, return type lowering interfering with closure optimization
#29258 - always obey inline declarations if the calling signature is concrete
#29548 - flatten empty tuple - fix #29112
#29609 - improve performance for ascii(::String)
#29557 - propgate inbounds to substring and use it in split
#29677 - Don't splat a number in ReshapedArray
#29680 - add inbounds to some more substrings
#29682 - improve performance for sorting columns in sparse matrix (thereby improving sparse matmul)
#29696 - improve performance for sparse matrix vector indexing
#29746 - speed up logical indexing by bitarray
#29633 - help type inference for logical indexing
#29786 - Reduce allocation in flatten

Documentation

#29210 - correct the docs of Nothing
#29188 - Make using LinearAlgebra explicit
#29233 - fix awkward phrasing in parallel-computing.md
#29230 - Base.TCPServer --> Sockets.TCPServer
#29279 - Changed one identifier in one example
#29245 - Minor cleanup of noteworthy differences
#29229 - Improve description of Julia keywords
#29232 - fix minor typo in parallel-computing.md
#29299 - Fix a typo in docs
#29319 - typo fix in parallel-computing.md
#29332 - Improve noteworthy differences from R on ismissing
#29301 - Minor docs correction for maxintfloat.
#29351 - Some misisng paren and xrefs
#29349 - Backticks and xrefs
#29338 - Minor English fixes
#29391 - Minor English fixes/improvements
#29398 - Remove docstring relicts of findlast/findprev with a regex
#29415 - improve findnext/findprev docstring for strings
#29431 - Fix sentence fragment
#29376 - Some improvements to variables and scoping manual
#29281 - Improved English and egs
#29475 - fix #29454: better example for mutable struct
#29463 - tiny formatting fix on docs
#29502 - Correct language in documentation
#29497 - Add --project option to man page
#29336 - English fixes
#29337 - Minor English fixes
#29482 - initdefs: Typo
#29519 - [ci skip] Remove mention of Julia version
#29521 - [ci skip] Fix a typo in doc
#29543 - Update performance-tips.md
#29485 - Escape jl_atexit_hook first _
#29570 - basedocs: add 'See also' for getproperty and setproperty!
#29573 - Fix quantile doctest formatting
#29627 - Fix a typo: replacing 'temporarly' for 'temporarily'
#29591 - Document Dates.value
#29641 - fix an indent of Mmap.mmap
#29648 - Fix typo in arrays.md
#29689 - Adjust doc test for \ to avoid rounding
#29604 - Fix jldoctest example in mapreduce(...)
#29731 - Mention that the = part of --project is optional. Add to regular Julia docs as well.
#29763 - punctuation error
#29797 - Fix for the markdown table in getting-started.md
#29793 - clarify pinv documentation
#29789 - Created an actual table [Manual,Scope of Variables]
#29703 - make tables on Interfaces pages more consistent
#29834 - Fix CInt to Cint in docs

Assets 6

29 Sep 19:48

ararslan

v1.0.1

0d71392

v1.0.1

This is the first patch release for Julia version 1.0. It contains bugfixes only, no new features or changes.

A set of automatically generated patch notes based on the backported commits is shown below:

Bug fixes

#28568 - Fix vec of adjoint vector
#28592 - fix #28576, parsing ? at end of input in conditional expression
#28620 - Fix #28619, logabsdet for singular matrices
#28642 - Keep parametric information when constructing a SparseMatrixCSC from …
#28655 - fix #28641, passing typevars to <: in typejoin and tuplemerge
#28656 - fix #28499, calling cluster_cookie before cluster starts
#28659 - fix #28597, error for empty arrays with some negative dim sizes
#28684 - LinearAlgebra remove unused export
#28677 - smarter order of checks in jl_types_equal
#28693 - fix some exported undefined variables and test this does not happen
#28694 - fix REPL completions not always being unique
#28732 - Carry LLVM patch to fix incorrect codegen
#28742 - Fix missing GC root in interpreter.c
#28747 - Fix missing gc root in jl_cglobal
#28751 - Fix three missing gc roots in subtyping
#28727 - fix shuffle! on empty arrays
#28749 - Resolve all method ambiguities in LinearAlgebra
#28832 - Added seekstart to getpass in base/util.jl (#28828)
#28780 - make cmp with BigInt return in [-1, 0, 1]
#28706 - fix #28593, macro hygiene problems in type definitions
#28625 - absolutify --project path
#28651 - Support parallelizing over collections that have non-Int lengths
#28875 - fix doccing call overload without named argument
#28871 - issubset aces Discrete Math but flunks Topology
#28910 - doc: fix accumulate examples not using the init keyword
#28682 - Fix download agent search relying on throwing of Sys.which(). Update t #28157
#28951 - fix #28915, identity conversion to union of tuple types
#29043 - fix #28900, macro hygiene bug inside keyword arg values
#29044 - fix #28991, error when adding default defs to constructed module Exprs
#28882 - fix #28833, crash in macros returning certain top-level forms
#29070 - fix #29064, bug in field name same as struct type name
#29072 - fix #25955, hygiene of arg name of function defined by type
#29139 - fix #22592, specificity of union compared to a subtype of it
#29148 - ndigits with negative base: error out instead of giving incorrect result
#29159 - fix #25474, support more forms in edit, which, etc. macros
#29154 - fix #29145, error for new{} with incomplete type
#29194 - fix #29175, invalid lowered IR from repeating code for declared types
#29056 - Improve support for constructing zero-step float ranges

Performance improvements

#28661 - improve performance for number to string conversion functions
#28670 - fix performance in printf
#28707 - Fix reinterpret performance
#28787 - improve performance of parse
#28847 - Remove bounds checks in iterate(::Tuple)
#28876 - improve performance for string(...)
#28873 - Compare Symbols and Chars more efficiently in Cholesky
#28954 - fixed regression in sparse of adjoint of sparse (#28948)
#29045 - Fix dispatch of SparseMatrixCSC*Diagonal multiplication
#29060 - work around a splatting penalty in twiceprecision
#29086 - fix #29083, not eliminating kwfunc due to logic error in builtin_nothrow
#29133 - fix perf regression from not specializing on iterate on tuples
#29116 - fix perf in exp(::Matrix) (for smallish matrices)
#29186 - fix #28481, faster reading of primitive types from IOStream and IOBuffer
#29075 - fix #29036, large slowdown in DelimitedFiles