OCaml Changelog

RSS

Read the latest releases and updates from the OCaml ecosystem.

In the last three months after the release of OCaml 5.1.0, three significant regressions have been discovered in OCaml 5.1.0. Those regressions concern the packaging of executables, the typechecking of OCaml programs, and the performance of numerical codes.

Since those regressions affect many users and could have lasting effects, we have published patch release OCaml 5.1.1 fixing those issues.

As a major exception to our policy for patch releases, OCaml 5.1.1 will contain one breaking change in the standard library: the Compression flag has been removed from the Marshal module.

This drastic measure was taken because supporting ZSTD compression in the standard library made ZSTD a dependency of all OCaml executables. Since the compiler should not impose its dependency on end users, the support for compressed marshaling has been moved to a compiler internal library in 5.1.1.

Accounting for the still-experimental nature of the Multicore runtime, this patch release 5.1.1 also contains safe fixes for subtle concurrency issues in the OCaml runtime.

The full list of changes is available below for more details.


Installation Instructions

The base compiler can be installed as an opam switch with the following commands:

opam update
opam switch create 5.1.1

The source code for the release is also directly available on:

See full changelog

Standard library:

  • (breaking change) #12562, #12734, #12783: Remove the Marshal.Compression flag to the Marshal.to_* functions introduced in 5.1 by #12006, as it cannot be implemented without risking to link -lzstd with all ocamlopt-generated executables. The compilers are still able to use ZSTD compression for compilation artefacts. (Xavier Leroy and David Allsopp, report by Kate Deplaix, review by Nicolás Ojeda Bär, Kate Deplaix, and Damien Doligez).

Runtime Bug Fixes:

  • #11800, #12707: Fix an assertion race condition in install_backup_thread (Jan Midtgaard, review by Gabriel Scherer)

  • #12486, #12535: Fix delivery of unhandled effect exceptions on AMD64 with --enable-frame-pointers (Miod Vallat, report by Jan Midtgaard, review by Gabriel Scherer)

  • #12712, #12742: Fix an assertion boundary case in caml_reset_young_limit (Jan Midtgaard, review by Guillaume Munch-Maccagnoni)

  • #12713, #12715: Disable common subexpression elimination for atomic loads (Gabriel Scherer and Vincent Laviron, review by Vincent Laviron, KC Sivaramakrishnan and Xavier Leroy, report by Vesa Karvonen and Carine Morel)

GC Performance Regression Bug Fixes:

  • #12318: GC: simplify the meaning of custom_minor_max_size: blocks with out-of-heap memory above this limit are now allocated directly in the major heap. (Damien Doligez, report by Stephen Dolan, review by Gabriel Scherer)

  • #12439: Finalise and collect dead custom blocks during minor collection (Damien Doligez, review by Xavier Leroy, Gabriel Scherer and KC Sivaramakrishnan)

  • #12590, #12595: Move caml_collect_gc_stats_sample in caml_empty_minor_heap_promote before barrier arrival. (B. Szilvasy, review by Gabriel Scherer)

  • #12491, #12493, #12500, #12754: Do not change GC pace when creating subarrays of bigarrays (Xavier Leroy, report by Ido Yariv, analysis by Gabriel Scherer, review by Gabriel Scherer and Fabrice Buoro)

Bug Fixes

  • #12581, #12609: Fix error on uses of packed modules outside their pack to correctly handle nested packs (Vincent Laviron, report by Javier Chávarri, review by Pierre Chambart)

  • #12623: Fix the computation of variance composition (Florian Angeletti, report by Vesa Karvonen, review by Gabriel Scherer)

  • #12645, #12649: Fix error messages for cyclic type definitions in presence of the -short-paths flag. (Florian Angeletti, report by Vesa Karvonen, review by Gabriel Scherer)

  • #12757: Fix ocamlnat (native toplevel) by registering frametables correctly (Stephen Dolan, Nick Barnes and Mark Shinwell, review by Vincent Laviron and Sébastien Hinderer)

  • #12791: extern is applied to definitions of caml_builtin_cprim and caml_names_of_builtin_cprim when linking bytecode -custom executables with a C++ linker. (Shayne Fletcher, review by Antonin Décimo and Xavier Leroy)

We are pleased to announce the release of OCaml LSP 1.17.0. This version comes packed with fixes and new features.

Notable features that come in this release include:

  • Compatibility with Odoc 2.3.0: This version is fully compatible with Odoc 2.3.0, introducing support for the latest syntax features like tables and "codeblock output."
  • New Actions for Code Cleanup: You can now mark or remove unused elements such as 'open', types, for loop indexes, modules, match cases, 'rec', and constructors, making your code cleaner and more efficient.
  • Auto-completion for 'in' Keyword: To address the long standing issue of the in keyword completing to other terms in the editor, we've added auto-completion for this keyword in OCaml LSP. Typing in will now only suggest auto completion for the in keyword.
See full changelog

Fixes

  • Fix missing super & subscripts in markdown documentation. (#1170)
  • Do not invoke dune at all if --fallback-read-dot-merlin flag is on. (#1173)
  • Fix semantic highlighting of infix operators that contain '.'. (#1186)
  • Disable highlighting unit as an enum member to fix comment highlighting bug. (#1185)
  • Improve type-on-hover and type-annotate efficiency by only formatting the type of the first enclosing. (#1191, #1196)
  • Fix the encoding of URI's to match how vscode does it (#1197)
  • Fix parsing of completion prefixes (#1181)

Features

  • Compatibility with Odoc 2.3.0, with support for the introduced syntax: tables, and "codeblock output" (#1184)
  • Display text of references in doc strings (#1166)
  • Add mark/remove unused actions for open, types, for loop indexes, modules, match cases, rec, and constructors (#1141)
  • Offer auto-completion for the keyword in (#1217)

Odoc 2.4.0

The odoc team is delighted to announce the release of odoc 2.4.0. It mainly contains support for search engines. There are of course bugfixes and smaller new features.

🌟 Spotlight Feature of Odoc 2.4.0 : Search

Odoc now support searching in the documentation ! The search is made to run in the browser, so that you do not need a server to enable search: you can have search on your documentation hosted on github pages or even locally on your machine.

No search engine is shipped with, you need to provide one, but all the facilities to make use of one are present. We adapted @art-w 's sherlodoc for seamless integration with odoc, alongside with new features. It is not yet released on opam, but we hope it will be soon.

You can already test sherlodoc and play with it on your own projects, there are instructions in its readme. Sherlodoc has fuzzy typed-based search like hoogle in the haskell world, and is made to work best for OCaml (unlike a general purpose search engine like elastic search).

Check the results on odoc's own online documentation : ocaml.github.io/odoc.

🤝 Join The Mission

While we are dedicated to developing the best tooling to generate and serve documentation on OCaml.org, creating a well-documented library ecosystem can only be a collective effort. Package authors: we’re working hard to give you great tools, but we’ll need all your help to create an ecosystem of well-documented libraries for OCaml!

If you find that writing documentation for your library isn’t as straightforward as you would like, please do share your feedback with us.

See full changelog

Added

  • Add support for external search engines (@panglesd, @EmileTrotignon, #972) This includes the generation of an index and the display of the results in the UI (HTML only).
  • Display 'private' keyword for private type extensions (@gpetiot, #1019)
  • Allow to omit parent type in constructor reference (@panglesd, @EmileTrotignon, #933)

Fixed

  • Warn and exit when table(s) is not closed (@lubegasimon, #1050)
  • Hint when list(s) is not closed (@lubegasimon, #1050)
  • Fix crash on functors returning an alias (@Julow, #1046)
  • Fix rendering of polymorphic variants (@wikku, @panglesd, #971)
  • Add references to extension declarations (@gpetiot, @panglesd, #949)

Changed

  • Style: Adjusted line height in the TOC to improve readability (@sorawee, #1045)
  • Style: Remove font fallback to Helvetica, Arial (@Julow, #1028)
  • Style: Preformatted elements fallback to UA monospace (@toastal, #967)
  • Style: Sidebar is now stuck to the left of the content instead of the left of the viewport (@EmileTrotignon, #999)

Two months (and half) after the release of OCaml 5.1.0, we have discovered three significant regressions in OCaml 5.1.0, affecting the packaging of executables, the typechecking of OCaml programs, and the performance of numerical codes.

Since those regressions affect many users and could have lasting effects, we have decided to cut a patch release of OCaml 5.1 with fixes for those issues next week.

To give time for opam packages to test this patch version, we have just published a first release candidate for OCaml 5.1.1 . The full list of changes in this release candidate is available below.

As a major exception to our policy for patch releases, OCaml 5.1.1 will contain one breaking change in the standard library: the Compression flag has been removed from the Marshal module. This drastic measure was taken because supporting zstd compression in the standard library made zstd a dependency of all OCaml executables. Since the compiler should not impose its dependency on end-users, the support for compressed marshaling has been moved to a compiler internal library in 5.1.1. This internally library might be released as an independent library in later releases of OCaml but not in OCaml 5.1.1 to give us time to polish the library integration and packaging.

Installation Instructions

The base compiler can be installed as an opam switch with the following commands on opam 2.1 and later:

opam update
opam switch create 5.1.1~rc1

The source code for the release candidate is also directly available on:

Fine-Tuned Compiler Configuration

If you want to tweak the configuration of the compiler, you can switch to the option variant with:

opam update
opam switch create <switch_name> ocaml-variants.5.1.1~rc1+options <option_list>

where option_list is a space-separated list of ocaml-option-* packages. For instance, for a flambda and no-flat-float-array switch:

opam switch create 5.1.1~rc1+flambda+nffa ocaml-variants.5.1.1~rc1+options ocaml-option-flambda ocaml-option-no-flat-float-array

All available options can be listed with opam search ocaml-option.

See full changelog

Changes in OCaml 5.1.1

Standard Library Dependency Fix:

  • (breaking change) #12562, #12734, #12783: Remove the Marshal.Compression flag to the Marshal.to_* functions introduced in 5.1 by #12006, as it cannot be implemented without risking to link -lzstd with all ocamlopt-generated executables. The compilers are still able to use ZSTD compression for compilation artefacts. (Xavier Leroy and David Allsopp, report by Kate Deplaix, review by Nicolás Ojeda Bär, Kate Deplaix, and Damien Doligez).

Type System Bug Fix:

  • #12623, fix the computation of variance composition (Florian Angeletti, report by Vesa Karvonen, review by Gabriel Scherer)

GC Performance Regression Fixes

  • #12590, #12595: Move caml_collect_gc_stats_sample in caml_empty_minor_heap_promote before barrier arrival. (B. Szilvasy, review by Gabriel Scherer)

  • #12318: GC: simplify the meaning of custom_minor_max_size: blocks with out-of-heap memory above this limit are now allocated directly in the major heap. (Damien Doligez, report by Stephen Dolan, review by Gabriel Scherer)

  • #12439: Finalize and collect dead custom blocks during minor collection (Damien Doligez, review by Xavier Leroy, Gabriel Scherer and KC Sivaramakrishnan)

  • #12491, #12493, #12500, #12754: Do not change GC pace when creating sub-arrays of bigarrays (Xavier Leroy, report by Ido Yariv, analysis by Gabriel Scherer,

Miscellaneous Bug Fixes

  • #12581, #12609: Fix error on uses of packed modules outside their pack to correctly handle nested packs (Vincent Laviron, report by Javier Chávarri, review by Pierre Chambart)

  • #12757: Fix ocamlnat (native toplevel) by registering frametables correctly (Stephen Dolan, Nick Barnes and Mark Shinwell, review by Vincent Laviron and Sébastien Hinderer)

  • #12645, #12649 fix error messages for cyclic type definitions in presence of the -short-paths flag. (Florian Angeletti, report by Vesa Karvonen, review by Gabriel Scherer)

We're happy to announce the release of Merlin 4.13. This release comes with a number of bug fixes to the Merlin binary.

It also introduces a new -cache-lifespan flag that can be used to set cache invalidation periods. This flag allows for fine-tuning of cache invalidation periods, enabling users to customize Merlin's internal configurations for optimal performance in various project contexts.

Additionally, this release includes several improvements in editor modes. For more details, we encourage you to delve into the full changelog.

See full changelog
  • merlin binary
    • Fix a follow-up issue to the preference of non-ghost nodes introduced in #1660 (#1690, fixes #1689)
    • Add -cache-lifespan flag, that sets cache invalidation period. (#1698, #1705)
    • Fix Merlin locate not fallbacking on the correct file in case of ambiguity (@goldfirere, #1699)
    • Fix Merlin reporting errors provoked by the recovery itself (#1709, fixes #1704)
  • editor modes
    • vim: load merlin when Vim is compiled with +python3/dyn (e.g. MacVim)
    • emacs: highlight only first error line by default (#1693, fixes #1663)

About six months ago, opam-repo-ci (opam.ci.ocaml.org) suffered from a lack of system memory (issue 220) which caused it to be moved to the machine hosting ocaml-ci (ocaml.ci.dev).

Subsequently, that machine suffered from BTRFS volume corruption (issue 51). Therefore, we moved both services to a larger new server. The data was efficiently migrated using BTRFS tools: btrfs send | btrfs receive.

Since the move, we have seen issues with BTRFS metadata. Plus, we have suffered from a build-up of subvolumes, as reported by other users: Docker gradually exhausts disk space on BTRFS.

Unfortunately, both services went down on Friday evening (issue 85). Analysis showed over 500 BTRFS subvolumes, a shortage of metadata space, and insufficient space to perform a BTRFS rebalance.

Returning to the original configuration of splitting the ci.dev and OCaml.org services, they have been moved onto new and separate hardware. The underlying filesystem is now a RAID1-backed ext4, formatted with -i 8192 in order to ensure the availability of sufficient inodes. Docker uses Overlayfs. RSYNC was used to copy the databases and logs from the old server. This change should add resilience and has doubled the capacity for storing history logs.

Dune 3.12.1

We're pleased to announce the release of Dune 3.12!

Note that due to a mistake during the release process, version 3.12.0 was not published to opam and should not be used.

This version contains many fixes but we would like to highlight the following features:

  • dune ocaml doc is a new command that will build the docs of your package, and open them in a web browser directly.
  • a new set of odoc rules will build documentation for the whole switch. It can be invoked by dune build @doc-new. Note that this is still experimental.
  • Dune can now be built and installed on Haiku.
See full changelog

Added

  • Introduce $ dune ocaml doc to open and browse documentation. (#7262, fixes #6831, @EmileTrotignon)

  • dune cache trim now accepts binary byte units: KiB, MiB, etc. (#8618, @Alizter)

  • Introduce the runtest_alias field to the cram stanza. This allows removing default runtest alias from tests. (@rgrinberg, #8887)

  • Display cache location in Dune log (#8974, @nojb)

  • Dune can now be built and installed on Haiku (#8795, fix #8551, @Alizter)

  • Mark installed directories in dune-package files. This fixes (package) dependencies against packages that contain such directories. (#8953, fixes #8915, @emillon)

  • Introduce new experimental odoc rules (#8803, @jonjudlam)

Changed

  • dune-build-info: when version="" is found in a META file, we now return None as a version string (#9177, @emillon)

  • No longer force colors for OCaml 4.03 and 4.04 (#8778, @rgrinberg)

  • Dependencies in the copying sandbox are now writeable (#8920, @rgrinberg)

  • Rules that only use internal dune actions (write-file, echo, etc.) can now be sandboxed. (#9041, fixes #8854, @rgrinberg)

  • Add test_ prefix to default test name in dune init project (#9257, fixes #9131, @9sako6)

Fixed

  • Do not ignore libraries named bigarray when they are defined in conjunction with OCaml 5.0 (#8902, fixes #8901, @rgrinberg)

  • Correctly ignore bigarray on recent versions of OCaml (#9076, @rgrinberg)

  • Absent packages shouldn't prevent all rules from being loaded (#8948, fixes #8630, @rgrinberg)

  • Correctly determine the stanza of menhir modules when (include_subdirs qualified) is enabled (@rgrinberg, #8949, fixes #7610)

  • Re-run actions whenever (expand_aliases_in_sandbox) changes (#8990, @rgrinberg)

  • Do not re-run rules when their location changes (#9052, @rgrinberg)

  • [coq rules] Be more tolerant when coqc --print-version / --config don't work properly, and fallback to a reasonable default. This fixes problems when building Coq projects with (stdlib no) and likely other cases. (#8966, fix #8958, @Alizter, reported by Lasse Blaauwbroek)

  • Dune will now run at a lower framerate of 15 fps rather than 60 when INSIDE_EMACS. (#8812, @Alizter)

Feedback on this post is welcomed on Discuss!

We are happy to announce the third alpha release of opam 2.2.0. It contains some fixes and a new feature for Windows. You can view the full list in the release note.

This version is an alpha, we invite users to test it to spot previously unnoticed bugs as we head towards the stable release.

Major change: Environment variables handling on Windows

opam files now support a new x-env-path-rewrite field which specifies rewriting rules for the environment variable updates defined in the setenv and build-env fields. This field allows greater control over the separator character used for PATH-like fields (i.e. ; vs :), conversion of slashes to backslashes, and even conversion from Windows native path format (C:\cygwin64\home\dra\.opam) to Cygwin format (/home/dra/.opam).

The rewriting rules allow opam directory-like variables (e.g. the %{lib}% directory of a switch) to be used in setenv and build-env fields in a portable way.

For example, given:

setenv: [
  [PKG_CONFIG_PATH += "%{lib}%/pkgconfig"]
  [PATH += "%{share}%/bin"]
]
x-env-path-rewrite: [
  [ PKG_CONFIG_PATH ":" "host" ]
  [ PATH (":" {os != "win32"} | ";" {os = "win32"}) ("target" {os != "win32"} | "target-quoted" {os = "win32"}) ]
]

the environment variable changes given by opam env on Windows would be:

PKG_CONFIG_PATH='/cygdrive/c/Users/DRA/AppData/Local/opam/default/lib/pkgconfig[:<rest-of-PKG_CONFIG_PATH, if given>]'
PATH='C:\Users\DRA\AppData\Local\opam\default\share\bin;C:\Users\DRA\AppData\Local\opam\default\bin;<rest-of-PATH>'

with the following interesting parts for Windows users:

  • PKG_CONFIG_PATH, which is consumed by a Cygwin-tool, has the directory given in Unix-like syntax, and opam's share variable was automatically converted
  • The correct separator is used for each (: for PKG_CONFIG_PATH, ; for PATH is used when adding entries)
  • In the PATH update, /bin was converted to \bin

Note that the specification for PATH is opam's default behaviour, so it's not actually necessary to have this formula for PATH in the x-env-path-rewrite field.

The full syntax is described in full in the manual.

Opam files carrying this new field are still compatible with the opam 2.0 syntax as it is an extension field, however its effect is only available with opam 2.2.0~alpha3 and above. If you want to make sure users of the package containing it have a compatible opam, you can use the available field to that end:

available: opam-version >= "2.2.0~alpha3"

or, if the change is Windows-specific:

available: opam-version >= "2.2.0~alpha3" | os != "win32"

Other noteworthy changes

  • Sandbox: /tmp is now writable again, restoring POSIX compliance
  • opam tree: opam tree package.version is now supported, displaying the dependency tree of a specific version of a package
  • opam tree: --recurse and --subpath are supported for directory arguments
  • opam admin: new add-extrafiles command to add/check/update the extra-files: field according to the files present in the files/ directory
  • opam lint: new syntax allow marking a set of warnings as errors e.g. -W @1..9
  • Releases: Pre-built binaries now include ppc64le and s390x

Miscellaneous changes

  • A handful of issues related to the compilation of opam on Windows were fixed
  • Bugs in the handling of the OPAMCURL, OPAMFETCH and OPAMVERBOSE environment variables were fixed
  • Bugs in the handling of the --assume-built argument were fixed
  • Sporadic crashes and segfaults during shell detection on Windows were fixed

Various other improvements were made and bugs were fixed. API changes are denoted in the release note linked above. This release also includes a handful of PRs improving the documentation and more than a dozen PRs improving and extending the tests.

Windows Support

The main opam-repository Windows compliance is still a work in progress, we recommend to use existing compatible repository (originally from @fdopen) and 32/64 bit mingw-w64 packages (by @dra27).

How to Test opam on Windows

If you're feeling adventurous, you can use the experimental pre-built binary for Windows available here. It should work but since it's our first public pre-built binary for Windows please use at your own risk.

Otherwise you can compile opam yourself using the following steps:

This alpha requires a preexisting Cygwin installation for compiling opam.

  1. Check that you have all dependencies installed:
  • autoconf, make, patch, curl
  • MinGW compilers: mingw64-x86_64-gcc-g++, mingw64-i686-gcc-g++
  • Or if you want to use the MSVC port of OCaml, you'll need to install Visual Studio or Visual Studio Build Tools
  1. Download & extract the opam archive
  2. In the directory:
  • if you are using MSVC: launch make cold
  • if you are using MinGW: make sure the path to the libc dlls are in your PATH and launch make cold. For instance: export PATH='C:\cygwin64\usr\x86_64-w64-mingw32\sys-root\mingw\bin':$PATH && make cold. Don’t forget to update PATH accordingly or place the opam binary in the same directory as the libc dlls if you want to move the resulting opam binary.
  • alternatively, if you're using MinGW: make cold CONFIGURE_ARGS=--with-private-runtime. If you change the location of the resulting opam binary, don't forget to copy Opam.Runtime.amd64 directory (or Opam.Runtime.i386) in the same directory.
  1. A coffee later, you now have an opam executable!
  2. Start your preferred Windows terminal (cmd or PowerShell), and initialise opam with the Windows sunset repository:
  • opam init git+https://github.com/ocaml-opam/opam-repository-mingw

From here, you can try to install the sunset repository packages. If you find any bugs, please submit an issue. It will help opam-repository maintainers to add Windows repository packages into the main repository.

Try it!

In case you plan a possible rollback, you may want to first backup your ~/.opam directory.

The upgrade instructions are unchanged:

  1. Either from binaries: run

    bash -c "sh <(curl -fsSL https://raw.githubusercontent.com/ocaml/opam/master/shell/install.sh) --version 2.2.0~alpha3"
    

    or download manually from the Github "Releases" page to your PATH.

  2. Or from source, manually: see the instructions in the README.

You should then run:

opam init --reinit -ni

Please report any issues to the bug-tracker.

Thanks for trying this new release out, and we hope you will enjoy the new features!

We’re pleased to announce the release of opam-publish 2.3.0.

This release, apart from a couple of light improvements, mainly consists of the following new option:

  • You can now use opam-publish with the --no-confirmation argument for use in automated pipeline. Use this option with extreme caution if you do use it.

The full changelog is available below.

Enjoy, The opam team

See full changelog
  • Add a new --no-confirmation argument for use in automated pipeline [#158 @AltGr - fix #132]
  • Improve the error message when a file expected to be an opam file is given as argument [#150 @kit-ty-kate - partially fix #149]
  • Adopt the OCaml Code of Conduct [#151 @rikusilvola]
  • Changes the makefile to make sure the standard "make && make install" works [#157 @kit-ty-kate]

Odoc 2.3.1

Following the release of odoc 2.3.0, we're releasing a patch release with odoc 2.3.1 that fixes support for OCaml 5.1.

See full changelog

After giving space for feedback and objections by the community, we have deprecated ocaml-migrate-parsetree (aka OMP). It is superseded by Ppxlib.

There are four major differences between OMP and Ppxlib, which all go hand in hand.

The first major difference is in the library and therefore impacts how to write PPXs. With OMP, each PPX author had to choose a parsetree version to define their PPX against. There was no version agreement between different PPXs. With Ppxlib, each PPX author uses the same parsetree version.

The second major difference is about compatibility with new compiler syntax. While with OMP, each PPX was on its own parsetree version, Ppxlib keeps them all on the version of the latest stably released compiler. That makes using any PPX compatible with using the latest compiler syntax features!

The third major difference is in the philosophy of PPXs. With OMP, all PPX transformations were global transformations, i.e., transformations of the whole parsetree. Ppxlib has introduced the concept of "context-free" transformations, i.e., transformations that transform only one parsetree node. By restricting their scope of action, context-free PPXs are a lot more predictable and less dangerous! Also, Ppxlib merges all context-free PPXs into one parsetree pass, defining clear semantics of PPX composition.

The fourth major difference is in the driver, i.e., the binary that drives the application of all used PPXs in a project. The Ppxlib driver is significantly more performant than the OMP driver used to be. That's partly because it does a lot fewer parsetree migrations and partly thanks to merging all context-free PPXs into one parsetree pass.

As a consequence of the deprecation, OMP will be incompatible with any new compiler version. The first incompatible compiler version is OCaml 5.1.

Thanks a lot to everyone involved in OCaml's transition from OMP to Ppxlib, for example by porting their PPX!

Dune 3.11.1

Following the release of Dune 3.11.0 a few days ago, we've just released version 3.11.1 with a couple of bug fixes:

  • Fix the dune rpc status command that was incorrectly showing that there were no RPC servers running on Windows.
  • Fix a bogus error when no inline test was declared in a library using (inline_tests).
See full changelog

Fixed

  • Fix dune rpc commands on Windows (#8806, fixes #8799, @nojb)

  • Fix inline_tests when the partition list is empty (#8849, fixes #8848, @hhugo)

We're happy to announce the release of Ppxlib 0.31.0. Shortly after OCaml 5.1.0 has been released, this Ppxlib release fixes a bug in the support of OCaml 5.1.0. Before that bug fix, the warnings about a generative/applicative mismatch between a functor creation and its application introduced by OCaml 5.1.0 were also triggered when that mismatch didn't exist.

Furthermore, the release contains a couple of bug fixes in the context of attributes.

We're also excited about two main enhancements. One allows authors of extension node rewriters to add a path argument to the extension node. That's excellent for hygiene since it allows the PPX to be explicit about modules rather than depending on its scope.

The other main enhancement allows an opt-in for compiler warnings about unused code generated by derivers (warnings w32 and w60). That opting in needs to happen on both sides of the deriver, the writer side and the user side. Opting in to those code warnings will help to clean up unused code, leading to performance improvements in compilation and editor support.

See full changelog
  • Fix support for OCaml 5.1: migrated code preserves generative functor warnings, without creating more. Locations are better preserved. (#432, @pitag-ha, @panglesd)

  • Driver: Add -unused-code-warnings command-line flag. (#444, @ceastlund)

  • Add ?warning flag to Deriving.Generator.make. (#440, @jacksonzou123 via @ceastlund)

  • Restore the path_arg functionality in the V3 API. (#431, @ELLIOTTCABLE)

  • Expose migration/copying/etc. functions for all AST types needed by Pprintast. (#454, @antalsz)

  • Preserve quoted attributes on antiquotes in metaquot. (#441, @ncik-roberts)

  • Attribute namespaces: Fix semantics of reserving multi-component namespaces. (#443, @ncik-roberts)

Dune 3.11.0

We're pleased to announce the release of Dune 3.11!

Here are some highlights from this release.

Dune Terminal User Interface (TUI)

The biggest highlight in this release is the introduction of the new Terminal User Interface (TUI) display mode.

You can now run dune build --display tui to open a TUI. We recommend using it in conjunction with Dune watch mode (dune build -w --display tui).

From the TUI, you will be able to navigate messages from Dune build, scroll through them, or click to minimize them. Press ? to show the help screen.

Installing Source Directories

Another exciting feature in this release is the new (source_trees ..) field added to the (install ..) stanzas to add every files in a directory to the installation. For instance, to add your manual to the installation, you can use:

(install
 (section doc)
 (source_trees manual))

Let us know what you think about these features, and don't hesistate to open issues on Dune's bug tracker if you encounter any issue.

See full changelog

Added

  • enabled_if now supports arch_sixtyfour variable (#8023, fixes #7997, @Alizter)

  • Experimental: Added a $ dune monitor command that can connect to a running dune build in watch mode and display the errors and progress. (#8152, @Alizter)

  • The progress RPC procedure now has an extra field for the In_progress constructor for the number of failed jobs. (#8212, @Alizter)

  • Add a --preview flag to dune fmt which causes it to print out the changes it would make without applying them (#8289, @gridbugs)

  • Introduce (source_trees ..) to the install stanza to allow installing entire source trees. (#8349, @rgrinberg)

  • Add --stop-on-first-error option to dune build which will terminate the build when the first error is encountered. (#8400, @pmwhite and @Alizter)

  • Dune now displays the number of errors when waiting for changes in watch mode. (#8408, fixes #6889, @Alizter)

  • Add with_prefix keyword for changing the prefix of the destination of installed files matched by globs. (#8416, @gridbugs)

  • Added experimental --display tui option for Dune that opens an interactive Terminal User Interface (TUI) when Dune is running. Press '?' to open up a help screen when running for more information. (#8429, @Alizter and @rgrinberg)

  • Add a warnings field to dune-project files as a unified mechanism to enable or disable dune warnings (@rgrinberg, 8448)

  • dune exec: support syntax like %{bin:program}. This can appear anywhere in the command line, so things like dune exec time %{bin:program} now work. (#6035, #8474, fixes #2691, @emillon, @Leonidas-from-XIV)

  • Add a new alias @doc-json to build odoc documentation in JSON format. This output can be consumed by external tools. (#8178, @emillon)

Changed and Fixed

  • Use posix_spawn instead of fork on MacOS. This gives us a performance boost and allows us to re-enable thread. (#8090, @rgrinberg)

  • Modules that were declared in (modules_without_implementation), (private_modules) or (virtual_modules) but not declared in (modules) will raise an error. (#7674, @Alizter)

  • No longer emit linkopts(javascript) in META files (#8168, @hhugo)

  • RPC message styles are now serialised meaning that RPC diagnostics keep their ANSI styling. (#8516, fixes #6921, @Alizter)

  • Truncate output from actions that produce too much output (@tov, #8351)

  • Allow libraries to shadow OCaml built-in libraries. Previously, built-in libraries would always take precedence. (@rgrinberg, #8558)

  • dune utop no longer links utop in "custom" mode, which should make this command considerably faster. (#8631, fixes #6894, @nojb)

  • Ensure that package names in dune-project are valid opam package names. (#8331, @emillon)

  • init: check that module names are valid (#8644, fixes #8252, @emillon)

  • dune init: parse --public as a public name (#8603, fixes #7108, @emillon)

  • Stop signing source files with substitutions. Sign only binaries instead (#8361, fixes #8360, @anmonteiro)

  • Make copy sandbox support directory targets. (#8705, fixes #7724, @emillon)

Deprecated and Removed

  • Deprecate install destination paths beginning with ".." to prevent packages escaping their designated installation directories. (#8350, @gridbugs)

  • Remove warning against .dune files generated by pre Dune 2.0 (#8611, @rgrinberg)

  • Remove versions 0.1 and 0.2 of the experimental ctypes extension. (#8293, @emillon)

Odoc 2.3.0

We are thrilled to announce the release of odoc 2.3.0! 🎉 This release is the result of almost a year of diligent work from the odoc team since the last major release of odoc 2.2.0, it comes packed with significant new features and improvements!

🌟 Spotlight Features of Odoc 2.3.0

Here are a couple of the new features introduced in Odoc 2.3.0 that we'd like to highlight.

Table Support

Table support is the last addition to the odoc language, and comes with two syntax flavours: a light one, and a heavy one. The light markup is similar to markdown's markup for table, producing tables that are readable in the source file as well.

However, this markup has some limitation, since it only allows inline content in cells. It can also be difficult to read and mantain for big tables, without a proper editor support. For this reason, Odoc also provides a "heavy" markup, closer to the html one, with fewer limitations!

Here is a table in heavy, light, and rendered form:

{t
Table | support
------|--------
is    | cool!
}
{table
{tr {th Table} {th support}}
{tr {td is} {td cool!}}
}

Table | support ------|-------- is | cool!

Source Code Rendering

Source code rendering is an extremely exciting new feature. Not only odoc is now able to generate a rendering of the source files (and source hierarchy) of a project, but it is also able to create direct links from the documentation to the implementation!

This puts the documentation browsing to a new level, by helping to quickly answer any implementation-related question!

The source code rendering is also tailored to OCaml, for instance with links from variables to their definition, something missing from traditional html-based source viewing such as github!

Using this features in odoc’s driver will require some work, but you can already have a preview of the feature by going to the odoc API website, which was built with the feature enabled. For instance, the Odoc_html 19 module is now populated with many Source links, jumping right into the implementation file 5! Directory pages to browser the implementation are also included 8 :smiley:

🗺️ Background & Roadmap

Some background on our roadmap and what comes next.

The lack of access to comprehensive documentation for OCaml libraries is one of the biggest pain points reported by the OCaml community, as highlighted in the 2022 OCaml survey (c.f. Q50).

This motivated the odoc and OCaml.org teams to jointly work on a centralised package documentation, that went live in April 2022, as part of the new version of OCaml.org.

With documentation for OCaml libraries readily available on OCaml.org, we now turn our focus on making sure that library authors have the tooling they need to create high-quality documentation.

Our roadmap highlights some features we believe will make the generated documentation significantly better for readers, and documentation-writing much more pleasant and rewarding.

This release is a significant milestone in implementing the features on our roadmap and is the precursor to a series of upcoming releases. Odoc 2.4.0 will follow shortly and will bring support for search. Stay tuned and follow our progress through the OCaml Platform newsletter!

🤝 Join Our Mission

While we are dedicated to developing the best tooling to generate and serve documentation on OCaml.org, creating a well-documented library ecosystem can only be a collective effort. Package authors: we're working hard to give you great tools, but we'll need all your help to create an ecosystem of well-documented libraries for OCaml!

If you find that writing documentation for your library isn't as straightforward as you would like, please do share your feedback with us.

See full changelog

Additions

  • Source code rendering (@Julow, @panglesd, @jonludlam #909, #996, #993, #982)
  • Handle tables markup (@panglesd, @gpetiot, #893)
  • Initial support for assets (@trefis, #975)
  • odoc-parser remerged (@jonludlam, #973) This includes table support (@gpetiot, @panglesd, ocaml-doc/odoc-parser#11 ocaml-doc/odoc-parser#14) and delimited code blocks with optional output (@jonludlam, ocaml-doc/odoc-parser#17)
  • Add a tooltip to references with text (@Julow, #945)
  • Add emoji to alerts in CSS (@yawaramin, #928)
  • Add common language in shipped highlightjs (@Julow, #953)

Bugfixes

  • Fix --hidden not always taken into account (@panglesd, #940)
  • Syntax highlight labels in function arguments (@panglesd, #990)
  • Ensure generated html ends with a newline (@3Rafal, #954)
  • Warn against tags in pages (@Julow, #948)
  • Remove unhelpful 'Unresolved_apply' errors (@gpetiot, #946)
  • Allow links and references in headings (@EmileTrotignon, @panglesd, #942)
  • Fix rendering of method types (@zoggy, #935)
  • Fix section labelling with submodules (@EmileTrotignon, @panglesd, #931)
  • LaTeX backend fixes (@Octachron, #921 #920)
  • html: Remove extra space in class declarations (@Julow, #936)
  • Fix rendering of unresolved references (@Julow, #957)

MDX 2.3.1

We are happy to announce the release of MDX 2.3.1! This is the first release of MDX to be compatible with OCaml 5.1.

We've also vendored the odoc-parser library, eliminating the need to have it as a dependency. MDX can now be installed independently of the odoc version you're using.

See full changelog

Added

  • Add os_type label to enable/disable based on Sys.os_type (#433, @polytypic)

  • Make MDX compatible with OCaml 5.1 (#435, @polytypic and @kit-ty-kate)

Changed

  • Vendored the odoc-parser library, removing the need to have it as a dependency. (#430, @jonludlam)

Merlin 4.12 is a small patch release fixing regressions introduced in previous versions.

See full changelog
  • merlin binary
    • Fix issue with let operators and tuples (ocaml/merlin#1684, fixes ocaml/merlin#1683, fixes ocaml/ocaml-lsp#1182)
    • Fix an issue causing Merlin locate queries to hang (ocaml/merlin#1686, fixes ocaml/ocaml-lsp#1192)

We are pleased to announce the release of Merlin 4.11 for OCaml 4.14.1, 5.0, and 5.1.

This release brings official support for OCaml 5.1 and introduces some changes to Emacs' default bindings proposed by the community:

  • Unbind C-c C-r (to avoid shadowing tuareg-eval-region) and bind C-c C-v instead to merlin-error-check;
  • Rebind C-c C-d to merlin-document and bind C-c M-d and C-c | instead to merlin-destruct;
  • Bind C-u C-c C-t to merlin-type-expr.

Thanks to everyone involved!

See full changelog
  • Merlin binary
    • Improve error messages for missing configuration reader (ocaml/merlin#1669)
    • Fix regression causing crash when using PPXs under Windows (ocaml/merlin#1673)
    • Fix confusion between aliased modules and module types (ocaml/merlin#1676, fixes ocaml/merlin#1667)
    • Ignore hidden branches when listing occurrences (ocaml/merlin#1677, fixes ocaml/merlin#1671)
  • Editor modes
    • Emacs: fix/improve keybindings (ocaml/merlin#1668, fixes ocaml/merlin#1386): Unbind C-c C-r (to avoid shadowing tuareg-eval-region) and bind C-c C-v instead to merlin-error-check; rebind C-c C-d to merlin-document and bind C-c M-d and C-c | instead to merlin-destruct; bind C-u C-c C-t to merlin-type-expr.
    • Emacs: remove use of obsolete defadvice macro (ocaml/merlin#1675)

Previously the infrastructure team had made FreeBSD available for opam-repo-ci. Now we can announce that the same support has been added to ocaml-ci, giving coverage for both OCaml 4.14 and the new OCaml 5.1 release. opam-repo-ci has also been upgraded to support OCaml 5.1. We aim to support both 4.14 as the Long Term Support release and the latest 5.* release.

Additionally an opam-health-check instance has been setup to provide continuous checking of opam repository packages against FreeBSD 13.2 x86_64 for both the 4.14 and 5.1 releases of OCaml. This will allow the community to check whether packages work on FreeBSD and provide fixes to opam-repository that will then get tested on FreeBSD. Closing the loop and giving the community the tools to support OCaml on FreeBSD effectively.

We hope the community finds the FreeBSD support useful.

We are pleased to announce the release of OCamlFormat 0.26.1!

This is the first OCamlFormat release to be compatible with OCaml 5.1.

We highlight notable formatting improvements below:

  1. Removal of extra breaks in constructor declarations
 type t =                               
   | Foo
   | (* Redirect (None, lib) looks up lib in the same database *)
-    Redirect of
-      db option * (Loc.t * Lib_name.t)
+    Redirect of db option * (Loc.t * Lib_name.t)
+    
  1. Consistent formatting for arrow class types, and consistent indentation of the object keyword
module type S = sig                    
-  class tttttttttttt : aaaaaaaaaaaaaaaaaa:int -> bbbbbbbbbbbbbbbbbbbbb:float ->
+  class tttttttttttt :
+    aaaaaaaaaaaaaaaaaa:int ->
+    bbbbbbbbbbbbbbbbbbbbb:float ->
    cccccccccccccccccccc
end

class type ct =
  let open M in
-  object
+object
  val x : t
end

We've also fixed a few bugs. Attributes that were previously skipped are not preserved, and we fixed a crash that occured in the presence of nested modules.

Have a look at the full changelog to see the list of improvements, and don’t hesitate to share your feedback on this release on OCaml Discuss.

See full changelog

Changed

  • Compatible with OCaml 5.1.0 (#2412, @Julow) The syntax of let-bindings changed sligthly in this version.
  • Improved ocp-indent compatibility (#2428, @Julow)
  • * Removed extra break in constructor declaration with comment (#2429, @Julow)
  • * De-indent the object keyword in class types (#2425, @Julow)
  • * Consistent formatting of arrows in class types (#2422, @Julow)

Fixed

  • Fix dropped attributes on a begin-end in a match case (#2421, @Julow)
  • Fix dropped attributes on begin-end in an if-then-else branch (#2436, @gpetiot)
  • Fix non-stabilizing comments before a functor type argument (#2420, @Julow)
  • Fix crash caused by module types with nested with module (#2419, @Julow)
  • Fix ';;' formatting between doc-comments and toplevel directives (#2432, @gpetiot)

We have the pleasure of celebrating the anniversary of Olympe de Gouges' "Declaration of the Rights of Woman and of the Female Citizen" by announcing the release of OCaml version 5.1.0.

Some of the highlights in OCaml 5.1.0 are:

  • Many runtime performance regression and memory-leaks fixes (dynlinking, weak array, weak hash sets, GC with idle domains, GC prefetching)
  • Restored support for native code generation on RISC-V and s390x architectures
  • Restored Cygwin port
  • Reduced installation size (50% reduction)
  • Compressed compilation artefacts (.cmi, .cmt, .cmti, .cmo, .cma files)
  • 19 error message improvements
  • 14 standard library functions made tail-recursive with Tail-Recursion-Modulo-Cons (TRMC), such as List.append and List.map
  • 57 new standard library functions
  • More examples in the standard library documentation
  • 42 bug fixes

OCaml 5.1.0 is still a relatively experimental release compared to the OCaml 4.14 branch. In particular:

  • The POWER port is being tested in the dev version of the compiler.
  • The Windows MSVC port is still unavailable.
  • Ephemeron performances need to be investigated.
  • GC compaction is a work in progress.
  • statmemprof is a work in progress.
  • There are a number of known runtime concurrency bugs (that trigger under rare circumstances).

We are planning to address those regressions, hopefully in time for the OCaml 5.2.0 release for some of them. Meanwhile, the OCaml 4.14 branch will be maintained, and the next release on the OCaml 4.14 branch, OCaml 4.14.2, should follow this release in the upcoming months.

Please report any unexpected behaviours on the OCaml issue tracker and post any questions or comments you might have on our discussion forums.

The full list of changes can be found in the changelog below.


Installation Instructions

The base compiler can be installed as an opam switch with the following commands:

opam update
opam switch create 5.1.0

The source code for the release candidate is also directly available on:

Fine-Tuned Compiler Configuration

If you want to tweak the configuration of the compiler, you can switch to the option variant with:

opam update
opam switch create <switch_name> ocaml-variants.5.1.0+options <option_list>

where <option_list> is a comma-separated list of ocaml-option-* packages. For instance, for a flambda and no-flat-float-array switch:

opam switch create 5.1.0+flambda+nffa ocaml-variants.5.0.0+options ocaml-option-flambda ocaml-option-no-flat-float-array
See full changelog

Restored backends

  • #11418, #11708: RISC-V multicore support. (Nicolás Ojeda Bär, review by KC Sivaramakrishnan)

  • #11712, #12258, #12261: s390x / IBM Z multicore support: OCaml & C stack separation; dynamic stack size checks; fiber and effects support. (Aleksei Nikiforov, with help from Vincent Laviron and Xavier Leroy, additional suggestions by Luc Maranget, review by the same and KC Sivaramakrishnan)

  • #11642: Restore Cygwin port. Add GC messages for address space reservations when OCAMLRUNPARAM option v includes 0x1000. (David Allsopp, review by Xavier Leroy, Guillaume Munch-Maccagnoni and Gabriel Scherer)

Standard library:

  • #12006, #12064: Add Marshal.Compression flag to Marshal.to_* functions. When this flag is explicitly set, marshaled data is compressed using ZSTD. On some practical examples, the marshalled output became three times smaller at no noticeable cost on the marshalling time. (Xavier Leroy, review by Edwin Török and Gabriel Scherer, fix by Damien Doligez)

  • #10464: Add List.is_empty. (Craig Ferguson, review by David Allsopp)

  • #11848: Add List.find_mapi, List.find_index: ('a -> bool) -> 'a list -> int option, Seq.find_mapi, Seq.find_index, Array.find_mapi, Array.find_index, Float.Array.find_opt, Float.Array.find_index, Float.Array.find_map, Float.Array.find_mapi. (Sima Kinsart, review by Daniel Bünzli and Nicolás Ojeda Bär)

  • #11410: Add Set.to_list, Map.to_list, Map.of_list, Map.add_to_list: key -> 'a -> 'a list t -> 'a list t. (Daniel Bünzli, review by Nicolás Ojeda Bär and Gabriel Scherer)

  • #11836, #11837: Add Array.map_inplace, Array.mapi_inplace, Float.Array.mapi_inplace and Float.Array.mapi_inplace. (Léo Andrès, review by Gabriel Scherer, KC Sivaramakrishnan and Nicolás Ojeda Bär)

  • #10967: Add Filename.temp_dir. (David Turner, review by Anil Madhavapeddy, Valentin Gatien-Baron, Nicolás Ojeda Bär, Gabriel Scherer, and Daniel Bünzli)

  • #11246: Add "hash" and "seeded_hash" functions to Bool, Int, Char, Float, Int32, Int64, and Nativeint. (Nicolás Ojeda Bär, review by Xavier Leroy and Gabriel Scherer)

  • #11488: Add Mutex.protect: Mutex.t -> (unit -> 'a) -> 'a for resource-safe critical sections protected by a mutex. (Simon Cruanes, review by Gabriel Scherer, Xavier Leroy, Guillaume Munch-Maccagnoni)

  • #11581: Add type equality witness type (_, _) eq = Equal: ('a, 'a) eq in a new module Stdlib.Type. (Nicolás Ojeda Bär, review by Daniel Bünzli, Jacques Garrigue, Florian Angeletti, Alain Frisch, Gabriel Scherer, Jeremy Yallop and Xavier Leroy)

  • #11843: Add In_channel.input_lines and In_channel.fold_lines. (Xavier Leroy, review by Nicolás Ojeda Bär and Wiktor Kuchta).

  • #11856, #11859: Using TRMC, the following Stdlib functions are now tail-recursive: Stdlib.(@), List.append, List.concat_map. (Jeremy Yallop, review by Daniel Bünzli, Anil Madhavapeddy, Nicolás Ojeda Bär, Gabriel Scherer, and Bannerets)

  • #11362, #11402: Using TRMC, the following Stdlib functions are now tail-recursive: List.map, List.mapi, List.map2, List.filter, List.filteri, List.filter_map, List.init, List.of_seq. (Nicolás Ojeda Bär, review by Xavier Leroy and Gabriel Scherer)

  • #11878, #11965: Prevent seek_in from marking buffer data as valid after closing the channel. This could lead to inputting uninitialized bytes. (Samuel Hym, review by Xavier Leroy and Olivier Nicole)

  • #11128: Add In_channel.isatty, Out_channel.isatty. (Nicolás Ojeda Bär, review by Gabriel Scherer and Florian Angeletti)

  • #10859: Add Format.pp_print_iter and Format.pp_print_array. (Léo Andrès and Daniel Bünzli, review by David Allsopp and Hugo Heuzard)

  • #10789: Add Stack.drop (Léo Andrès, review by Gabriel Scherer)

  • (breaking change) #10899: Change Stdlib.nan from signaling NaN to quiet NaN. (Greta Yorsh, review by Xavier Leroy, Guillaume Melquiond and Gabriel Scherer)
  • #11026, #11667, #11858: Rename the type of the accumulator of fold functions to 'acc: fold_left : ('acc -> 'a -> 'acc) -> 'acc -> 'a list -> 'acc fold_right : ('a -> 'acc -> 'acc) -> 'a list -> 'acc -> 'acc fold_left_map : ('acc -> 'a -> 'acc * 'b) -> 'acc -> 'a list -> 'acc * 'b list ... (Valentin Gatien-Baron and Francois Berenger, review by Gabriel Scherer and Nicolás Ojeda Bär)

  • #11354: Hashtbl.find_all is now tail-recursive. (Fermín Reig, review by Gabriel Scherer)

  • #11500: Make Hashtbl.mem non-allocating. (Simmo Saan, review by Nicolás Ojeda Bär)

  • #11412: Add Sys.is_regular_file (Xavier Leroy, review by Anil Madhavapeddy, Nicolás Ojeda Bär, David Allsopp)

  • #11322, #11329: serialization functions Random.State.{of,to}_binary_string between Random.State.t and string (Gabriel Scherer, report by Yotam Barnoy, review by Daniel Bünzli, Damien Doligez, Hugo Heuzard and Xavier Leroy)

  • #11830: Add Type.Id with val provably_equal : 'a Type.Id.t -> 'b Type.Id.t -> ('a, 'b) Type.eq option (Daniel Bünzli, review by Jeremy Yallop, Gabriel Scherer, Wiktor Kuchta, Nicolás Ojeda Bär)

  • #12184, #12320: Sys.rename Windows fixes on directory corner cases. (Jan Midtgaard, review by Anil Madhavapeddy)

  • (breaking change) #11565: Enable -strict-formats by default. Some incorrect format specifications (for printf) where silently ignored and now fail. Those new failures occur at compile-time, except if you use advanced format features like %(...%) that parse format strings dynamically. Pass -no-strict-formats to revert to the previous lenient behavior. (Nicolás Ojeda Bär, review by David Allsopp)

Installation size

Specific efforts have been made during this release to reduce the filesystem size of installed artifacts of the compiler distribution. The installation size of 5.1 is 272 MiB compared to 521 MiB for 5.0. Some of those changes will benefit all OCaml packages.

  • ocaml/RFCs#23, #12006: use compressed marshaled format from #12006 for .cmi, .cmt, .cmti files, and for debug info in .cmo and .cma files, resulting in major reduction in size. (Xavier Leroy, review by Edwin Török and Gabriel Scherer, RFC by Simon Cruanes)

  • #11981: Reduce size of OCaml installations by removing debugging information from installed bytecode executables. It is no longer possible to run ocamldebug over these installed bytecode executables, nor to get exception backtraces for them. (Xavier Leroy, review by David Allsopp, report by Fabrice Le Fessant)

  • (breaking change) #11993: install only bytecode executables for the ocamlmklib, ocamlcmt, ocamlprof, ocamlcp, ocamloptp, and ocamlmktop tools, but no native-code executables. A tool like ocamlmklib for example is now installed directly to $BINDIR/ocamlmklib; ocamlmklib.byte and ocamlmklib.opt are no longer installed to $BINDIR. (Xavier Leroy, review by Gabriel Scherer)

Runtime system:

  • #11589, #11903: Modify the GC pacing code to make sure the GC keeps up with allocations in the presence of idle domains. (Damien Doligez and Stephen Dolan, report by Florian Angeletti, review by KC Sivaramakrishnan and Sadiq Jaffer)

  • #11743: Speed up weak array operations (KC Sivaramakrishnan, review by François Bobot and Sadiq Jaffer)

  • #12131: Simplify implementation of weak hash sets, fixing a performance regression. (Nick Barnes, review by François Bobot, Alain Frisch and Damien Doligez).

  • #11474, #11998, #12065: Add support for user-defined events in the runtime event tracing system. (Lucas Pluvinage, review by Sadiq Jaffer, Guillaume Munch-Maccagnoni, Enguerrand Decorne, Gabriel Scherer and Anil Madhavapeddy)

  • #11827, #12249: Restore prefetching for GC marking (Fabrice Buoro and Stephen Dolan, review by Gabriel Scherer and Sadiq Jaffer)

  • #11144: Restore frame-pointers support for amd64 (Fabrice Buoro, review by Frederic Bour and KC Sivaramakrishnan)

  • #11935: Load frametables of dynlink'd modules in batch (Stephen Dolan, review by David Allsopp and Guillaume Munch-Maccagnoni)

  • #11284, #12525: Use compression of entries scheme when pruning mark stack. Can decrease memory usage for some workloads, otherwise should be unobservable. (Tom Kelly, review by Sabine Schmaltz, Sadiq Jaffer and Damien Doligez)

  • (breaking change) #11865, #11868, #11876: Clarify that the operations of a custom block must never access the OCaml runtime. The previous documentation only mentioned the main illicit usages. In particular, since OCaml 5.0, it is no longer safe to call caml_remove_global_root or caml_remove_generational_global_root from within the C finalizer of a custom block, or within the finalization function passed to caml_alloc_final. As a workaround, such a finalization operation can be registered with Gc.finalize instead, which guarantees to run the finalizer at a safe point. (Report by Timothy Bourke, discussion by Yotam Barnoy, Timothy Bourke, Sadiq Jaffer, Xavier Leroy, Guillaume Munch-Maccagnoni, and Gabriel Scherer)
  • #12130: Fix multicore crashes with weak hash sets. Fixes #11934. (Nick Barnes, review by François Bobot)

  • #12099: Add ocamlrund option, -events, to produce a trace of debug events during bytecode interpretation. Fixes #12098. (Richard L Ford, review by Gabriel Scherer)

  • #12001: Fix book keeping for last finalisers during the minor cycle (KC Sivaramakrishnan and Enguerrand Decorne, report by Guillaume Bury and Vincent Laviron, review by Sadiq Jaffer and KC Sivaramakrishnan)

  • #11919: New runtime events counters for major heap stats and minor heap resizing. (Sadiq Jaffer, review by Gabriel Scherer and David Allsopp)

  • #11287, #11872, #11955: Clean up reserved header bits (once used for Spacetime profiling). (Nick Barnes, review by Gabriel Scherer and Damien Doligez)

  • #11750: Decouple major slice from minor GC. (KC Sivaramakrishnan, review by Sadiq Jaffer, Guillaume Munch-Maccagnoni and Damien Doligez)

  • #11796: protect lazy computation of code fragment digest by a mutex. This makes the thread sanitizer happier, and avoids duplicating the hashing work. (Gabriel Scherer, review by Xavier Leroy, report by Olivier Nicole)

  • #11137: new Unsafe_store_tag(val, new_tag) macro to stop using Tag_val(val) as lvalue. (Gabriel Scherer, review by Xavier Leroy, Guillaume Munch-Maccagnoni and Nicolás Ojeda Bär)

  • #11880: Restore the correct sigmask in systhreads. (Christiano Haesbaert, review by Guillaume Munch-Maccagnoni and Sébastien Hinderer)

  • #11881: Fix thread-unsafety of registration of operations for "custom" values. (Guillaume Munch-Maccagnoni, review by Gabriel Scherer and KC Sivaramakrishnan)

  • #11980: fix quadratic behavior in natdynlink by using a STW section for frame-descriptor updates. (Gabriel Scherer, review by Sadiq Jaffer, report by André Maroneze for Frama-C and Guillaume Melquiond for Coq)

  • #12121: unrooted implementations of caml_callback*_exn (Gabriel Scherer, review by KC Sivaramakrishnan and Xavier Leroy)

  • #3921, #12039, #12128: poll for signals in long-running polymorphic comparisons. (B. Szilvasy, Gabriel Scherer and Xavier Leroy, review by Stefan Muenzel, Guillaume Munch-Maccagnoni and Damien Doligez)

  • #12231: Support MinGW-w64 11.0 winpthreads library, where the macro to set up to get flexdll working changed (David Allsopp and Samuel Hym, light review by Xavier Leroy)

Language features:

  • (breaking change) #11694: Add short syntax for generative functor types () -> ... (Jeremy Yallop, review by Gabriel Scherer, Nicolás Ojeda Bär, Jacques Garrigue)

  • (breaking change) #11457: Remove old polymorphic variant syntax. With type t = [ `A | `B ], one could use the syntax #t in types, where it means the same thing as [< t], and in patterns, where it means (`A | `B). The use of #t in types for polymorphic variants was deprecated since 2001, and is now removed. The syntax remains available in patterns, or for objects -- when t is a class type. (Stefan Muenzel, review by Gabriel Scherer and Jacques Garrigue)

  • (breaking change) #11984: Add dedicated syntax for generative functor application. Previously, OCaml did not distinguish between F () and F (struct end), even though the latter looks applicative. Instead, the decision between generative and applicative functor application was made based on the type of F. With this patch, we now distinguish these two application forms; writing F (struct end) for a generative functor leads to new warning 73. (Frederic Bour and Richard Eisenberg, review by Florian Angeletti)

  • #9975, #11365: Make empty types (type t = |) immediate. (Antal Spector-Zabusky, review by Gabriel Scherer)

Type system:

  • (breaking change) #6941, #11187, #12483: prohibit using classes through recursive modules inheriting or including a class belonging to a mutually-recursive module would previous behave incorrectly, and now results in a clean error. (Leo White, review by Gabriel Scherer and Florian Angeletti)

  • (breaking change) #12189, #12211: anonymous row variables in explicitly polymorphic type annotation, e.g. 'a. [< X of 'a ] -> 'a, are now implicitly universally quantified (in other words, the example above is now read as 'a 'r. ([< X of 'a ] as 'r) -> 'a). (Florian Angeletti and Gabriel Scherer, review by Jacques Garrigue)

Code generation and optimizations:

  • #11967: Remove traces of Obj.truncate, which allows some mutable loads to become immutable. (Nick Barnes, review by Vincent Laviron and KC Sivaramakrishnan)

  • #9945, #10883: Turn boolean-result float comparisons into primitive operations Uses the architecture's elementary operations for float comparisons, when available, rather than branching and then setting the return value. (Stefan Muenzel, review by Stephen Dolan, Alain Frisch and Vincent Laviron)

  • #8998, #11321, #11430: change mangling of OCaml long identifiers from camlModule__name_NNN to camlModule.name_NNN. The previous mangling schema, using __, was ambiguous. (Xavier Leroy, report by sliquister and Michael Bacarella, review by Gabriel Scherer)

  • #10834: The -safer-matching option disables type-based optimizations of pattern-matching compilation. This allows to produce a match failure if a pattern-matching was wrongly assumed to be exhaustive. Since the exhaustiveness check for GADTs has had bugs in the past, it may be useful if you need extra security with GADTs. (Jacques Garrigue, review by Gabriel Scherer)

  • #11102: Speed up register allocation by permanently spilling registers (Stephen Dolan, review by Xavier Leroy)

  • #11383: Restrict the local function optimisation to forbid moving code inside a sub-function (Vincent Laviron, review by Gabriel Scherer)

  • #11686: Better spilling heuristic for the Linear Scan allocator for more efficient stack usage. (Nicolás Ojeda Bär, Gabriel Scherer, Alain Frisch, review by Gabriel Scherer, Alain Frisch and Nathanaëlle Courant)

  • #11904: Remove arm, i386 native-code backends that were already disabled at configuration time. (Nicolás Ojeda Bär, review by Stephen Dolan, Anil Madhavapeddy, and Xavier Leroy)

  • #11134: Optimise 'include struct' in more cases (Stephen Dolan, review by Leo White and Vincent Laviron)

Other libraries:

  • #11374: Remove pointer cast to a type with stricter alignment requirements in Windows implementation of Unix.gettimeofday. Windows implementations of caml_unix_map_file, caml_unix_lseek and caml_unix_lseek_64 now release the runtime lock. Windows implementation of caml_unix_lockf modernised and simplified. Where possible, 64 bit integers are used instead of LARGE_INTEGER structs. (David Allsopp, review by Jonah Beckford and Xavier Leroy)

  • #11475: Make Unix terminal interface bindings domain-safe (Olivier Nicole and Xavier Leroy, review by Xavier Leroy)

  • #11775: Unix.write on a non-blocking socket under Windows will return normally if the write blocks after some data has already been written (as otherwise there is no way of knowing how much data has been written before blocking). The same behaviour was already present under Unix. (Nicolás Ojeda Bär, review by David Allsopp)

  • (breaking change) #11991: Unix on Windows: map ERROR_TOO_MANY_LINKS to EMLINK. (Nicolás Ojeda Bär)
  • #12067: Document Windows specific meanings of Unix.process_status type (Samuel Hym, review by David Allsopp)

  • #12072: Document and test that Sys.rename works over directories too (Jan Midtgaard, review by Anil Madhavapeddy and Xavier Leroy)

Tools:

  • #11889, #11978: ocamldoc: handle injectivity annotations and wildcards in type parameters. (Florian Angeletti, report by Wiktor Kuchta, review by Jules Aguillon)

  • #11787: Fix GDB scripts to work with OCaml 5's heap layout. (Nick Barnes)

  • #11772: fix ocamlyacc's handling of raw string literals (Demi Marie Obenour)

  • #9290: Add a directive to switch off debugging in toplevel. This allows to see optimized bytecode with -dlambda. (Jacques Garrigue, review by Gabriel Scherer)

  • #11166: ocamllex: the union of two character sets "cset1 | cset2" can now be used in any context where a character set is expected. (Nicolás Ojeda Bär, Martin Jambon, review by Sébastien Hinderer)

  • #11718: ocamlyacc: OCaml-style comments are now supported, in addition to the C-style comments already supported. The syntax is the same as that used in OCaml code. (Demi Marie Obenour, review by Damien Doligez)

  • #11728: ocamlyacc: generate line directives for %type declarations (Demi Marie Obenour, review by Damien Doligez)

  • #11773: ocamlyacc: Do not allow quoted literals (such as 'a' or "bc") in a token name or %type declaration. Previously such literals were accepted by ocamlyacc, but produced malformed OCaml that was rejected by the compiler. (Demi Marie Obenour, review by Gabriel Scherer)

  • #11774: ocamlyacc: fail if there is an I/O error (Demi Marie Obenour, review by Gabriel Scherer)

  • #11973: Add support for postfixed mingw host triplets (Romain Beauxis)

  • #12165: ocamldoc, use standard doctype to avoid quirk mode. (Florian Angeletti, review by Gabriel Scherer)

Manual and documentation:

  • #11476: Add examples in documentation of Hashtbl, Queue, Atomic, Format (Simon Cruanes, review by Yotam Barnoy, Gabriel Scherer, Daniel Bünzli, Ulugbek Abdullaev, and Nicolás Ojeda Bär)

  • #11883, #11884: Update documentation for In_channel and Out_channel with examples and sections to group related functions. (Kiran Gopinathan, review by Daniel Bünzli and Xavier Leroy)

  • #12095, #12097: Put the sample code of the user's manual and reference documentation of the standard library under the CC0 1.0 Universal (CC0 1.0) Public Domain Dedication license.

  • #11892: Document the semantic differences of Unix.exec* between Unix and Windows. (Boris Yakobowski, review by Daniel Bünzli, Gabriel Scherer and Nicolás Ojeda Bär)

  • #9430, #11291: Document the general desugaring rules for binding operators. (Gabriel Scherer, review by Nicolás Ojeda Bär)

  • #11481: Fix the type of Unix.umask to Unix.file_perm -> Unix.file_perm (Favonia, review by Sébastien Hinderer)

  • #11514: Document ocamltest builtin variables and actions (Olivier Nicole, review by Sébastien Hinderer)

  • #11676: Fix missing since annotation in the Sys and Format modules (Github user Bukolab99, review by Florian Angeletti)

  • #12028: Update format documentation to make it clearer that pp_print_newline flushes its newline (Florian Angeletti, review by Gabriel Scherer)

  • #12201: in the tutorial on modules, replace priority queue example by a simpler example based on FIFO queues. (Xavier Leroy, review by Anil Madhavapeddy and Nicolás Ojeda Bär).

  • #12352: Fix a typo in the documentation of Arg.write_arg (Christophe Raffalli, review by Florian Angeletti)

  • #7179, #11894: correct the description of CAMLreturn and CAMLreturn0 in the Interfacing C page and memory.h file. (Dong An, review by Guillaume Munch-Maccagnoni and Olivier Nicole )

Compiler user-interface and warnings:

  • #10647: Show hints for the "undefined global" error in the toplevel (Wiktor Kuchta, review by Gabriel Scherer)

  • #12116: Don't suggest to insert a semicolon when the type is not unit (Jules Aguillon, review by Florian Angeletti)

  • #11679: Improve the error message about too many arguments to a function (Jules Aguillon, review by Gabriel Scherer and Florian Angeletti)

  • #10009: Improve the error reported by mismatched struct/sig and =/: in module and module type bindings. (Jules Aguillon, review by Gabriel Scherer)

  • #11530: Include kinds in kind mismatch error message. "Error: This variant or record definition does not match that of type M.t The original is abstract, but this is a record". (Leonhard Markert, review by Gabriel Scherer and Florian Angeletti)

  • #11646: Add colors to error message hints. (Christiana Anthony, review by Florian Angeletti)

  • #11235, #11864: usage warnings for constructors and fields can now be disabled on field-by-field or constructor-by-constructor basis (Florian Angeletti, review by Gabriel Scherer)

  • #11888: Improve the error message when type variables cannot be deduced from the type parameters: Before: "Error: In this definition, a type variable cannot be deduced from the type parameters." After: "Error: In the GADT constructor T : 'a -> 'a s t the type variable 'a cannot be deduced from the type parameters." (Stefan Muenzel, review by Florian Angeletti and Gabriel Scherer)

  • #10818: Preserve integer literal formatting in type hint. (Leonhard Markert, review by Gabriel Scherer and Florian Angeletti)

  • #11338: Turn some partial application warnings into hints. (Leo White, review by Stephen Dolan)

  • #10931: Improve warning 14 (illegal backslash) with a better explanation of the causes and how to fix it. (David Allsopp, Florian Angeletti, Lucas De Angelis, Gabriel Scherer, review by Nicolás Ojeda Bär, Florian Angeletti, David Allsopp and Gabriel Scherer)

  • #10911: Improve the location reported by parenthesized assert expressions (Fabian Hemmer, review by Gabriel Scherer)

  • #1391, #7645, #3922: Add an early error when compiling different modules with mismatching -for-pack (Pierre Chambart and Vincent Laviron, review by Mark Shinwell)

  • #11297: Report "unclosed" error when "done" is missing in a "do .. done" construct. (Nicolás Ojeda Bär, review by Gabriel Scherer)

  • #11635, #5461, #10564: turn warning 31 (Module_linked_twice) into a hard error for ocamlc — this was already an error with ocamlopt. (Hugo Heuzard, review by Valentin Gatien-Baron and Gabriel Scherer)

  • #11653: Add the -no-absname option to ocamlc, ocamlopt and ocamldep. (Abiola Abdulsalam, review by Sébastien Hinderer and Florian Angeletti)

  • #11696: Add the -no-g option to ocamlc and ocamlopt. (Abiola Abdulsalam, review by Sébastien Hinderer, Nicolás Ojeda Bär and Florian Angeletti)

  • #11722: clearer error messages on non-well-founded type definitions (Gabriel Scherer, review by Jacques Garrigue)

  • #11819: make the native_compiler and native_dynlink configuration variables available through ocamlc -config. (Sébastien Hinderer, review by Gabriel Scherer and David Allsopp)

  • #8602, #11863: Add -stop-after lambda flag option (Douglas Smith and Dmitrii Kosarev, review by Gabriel Scherer)

  • #11910: Simplify naming convention for shadowed or ephemeral identifiers in error messages (eg: Illegal shadowing of included type t/2 by t) (Florian Angeletti, review by Jules Aguillon)

  • #12024: insert a blank line between separate compiler messages (Gabriel Scherer, review by Florian Angeletti, report by David Wong)

  • #12088, #9265, #11949: ocamldebug: fix confusing repeating behavior on blank lines within source scripts (Damien Doligez, review by Gabriel Scherer, report by Gaëtan Gilbert)

  • #12107: use aliases to mark weak row variables: _[< ... ], < _..>, _#ct are now rendered as [< ...] as '_weak1, < .. > as '_weak1, and #ct as '_weak1. (Florian Angeletti, suggestion by Stefan Muenzel, review by Gabriel Scherer)

  • #12051: Improve the error messages when type variables cannot be generalized (Stefan Muenzel, review by Florian Angeletti)

  • (breaking change) #12094: Trigger warning 5 (ignored-partial-application) when the scrutinee of a pattern matching is of arrow type and all cases match wildcard or exception patterns. (Nicolás Ojeda Bär, review by Gabriel Scherer)

Internal/compiler-libs changes:

  • #11018, #11869: Clean up Types.Variance, adding a description of the lattice used, and defining explicitly composition. (Jacques Garrigue, review by Gabriel Scherer and Jeremy Yallop)

  • #11536: Introduce wrapper functions for level management ([Ctype.with_level], etc) and for type variable scoping ([Typetexp.with_local_type_variable_scope]). The older API ([Ctype.(begin_def,end_def)], [Typetexp.(narrow,widen)], etc.) is now removed. (Jacques Garrigue and Takafumi Saikawa, review by Gabriel Scherer)

  • #11601, #11612, #11628, #11613, #11623, #12120 : Clean up some global state handling in emitcode, bytepackager, bytegen, bytesections, spill. (Hugo Heuzard, Stefan Muenzel, review by Vincent Laviron, Gabriel Scherer and Nathanaëlle Courant)

  • #12119, #12188, #12191: mirror type constraints on value binding in the parsetree: the constraint typ in let pat : typ = exp is now directly stored in the value binding node in the parsetree. (Florian Angeletti, review by Richard Eisenberg)

  • #11912: Refactoring handling of scoped type variables (Richard Eisenberg, review by Gabriel Scherer and Florian Angeletti)

  • #11691, #11706: use asm instead of asm for strict ISO C conformance (Xavier Leroy, report by Gregg Reynolds , review by Sadiq Jaffer)

  • #11764: add prototypes to old-style C function definitions and declarations (Antonin Décimo, review by Xavier Leroy)

  • #11693: Remove use of C99 Variable Length Arrays (VLAs) in the runtime. (David Allsopp, review by Xavier Leroy, Guillaume Munch-Maccagnoni, Stefan Muenzel and Gabriel Scherer)

  • #12138: Generalise interface for BUILD_PATH_PREFIX_MAP mapping. Absolute paths are now rewritten too. (Richard L Ford, suggestions and review by Gabriel Scherer)

  • #10512: explain the compilation strategy for switches on constructors (Gabriel Scherer, review by Vincent Laviron)

  • #11990: Improve comments and macros around frame descriptors. (Nick Barnes, review by Gabriel Scherer)

  • #11847, #11849, #11851, #11898: small refactorings in the type checker (Gabriel Scherer, review by Nicolás Ojeda Bär)

  • #11027: Separate typing counter-examples from type_pat into retype_pat; type_pat is no longer in CPS. (Jacques Garrigue and Takafumi Saikawa, review by Gabriel Scherer)

  • #11286, #11515: disambiguate identifiers by using how recently they have been bound in the current environment (Florian Angeletti, review by Gabriel Scherer)

  • #11364: Allow make -C testsuite promote to take TEST and LIST variables (Antal Spector-Zabusky, review by Gabriel Scherer and David Allsopp)

  • #11446: document switch compilation (lambda/switch.ml) (Gabriel Scherer, review by Luc Maranget and Vincent Laviron)

  • #11568: Encode inline record types in Path.t (Leo White and Hyunggyu Jang, review by Gabriel Scherer)

  • #11569: Remove hash type encoding (Hyunggyu Jang, review by Gabriel Scherer and Florian Angeletti)

  • #11627: use return values instead of globals for linear scan intervals (Stefan Muenzel, review by Nicolás Ojeda Bär)

  • #11634: Dll.open_dll now properly handles opening for execution while already opened for checking (Hugo Heuzard, review by Nicolás Ojeda Bär)

  • (breaking change) #11745, #12358: Debugger and toplevels: embed printer types rather than reading their representations from topdirs.cmi at runtime. This change also removes the ocamlmktop initialization module introduced in #11382 which was no longer useful. This change breaks toplevel scripts relying on the visibility of Topdirs in the initial toplevel environment without loading topfind. Since the opam default .ocamlinit file loads topfind, it is expected that only scripts run with ocaml -noinit are affected. For those scripts, accessing Topdirs now requires the compiler-libs directory to be added to the toplevel search path with
      #directory "+compiler-libs";;
    
    as was already the case for the other modules in the toplevel interface library. (Sébastien Hinderer, review by Florian Angeletti, Nicolás Ojeda Bär and Gabriel Scherer)
  • #11615: remove global variables form asmcomp/linearize.ml (Stefan Muenzel, review by Nicolás Ojeda Bär

  • #10856: Add location, attribute(s) visitors to Tast_mapper/Tast_iterator (Yan Dong, review by Nicolás Ojeda Bär and Gabriel Scherer)

  • #11763, #11759, #11861: Enable stricter C compilation warnings, use strict prototypes on primitives. (Antonin Décimo, review by Xavier Leroy, David Allsopp and Sébastien Hinderer)

  • #11933: Use the correct machtype when reading the code pointer from closures (Nathanaëlle Courant, review by Gabriel Scherer and Vincent Laviron)

  • #11972: refactor runtime/frame_descriptors.c in preparation for quadratic-time fix (Gabriel Scherer, review by Enguerrand Decorne)

  • #11997: translate structured constants into their Obj.t representation at compile time rather than link time. Changes the way dumpobj prints these constants because their representation becomes untyped. (Sébastien Hinderer, review by Xavier Leroy, Nicolás Ojeda Bär and Hugo Heuzard)

  • #12011: remove Ctype.reified_var_counter (Takafumi Saikawa and Jacques Garrigue, review by Gabriel Scherer)

  • #12012: move calls to Typetexp.TyVarEnv.reset inside with_local_level etc. (Jacques Garrigue and Takafumi Saikawa, review by Gabriel Scherer)

  • #12034: a logarithmic algorithm to find the next free variable (Gabriel Scherer, review by Stefan Muenzel)

  • #12092: remove Lev_module_definition from lambda (Nick Roberts, review by Gabriel Scherer)

  • #12117: Remove arity-interrupting elaboration of module unpacks (Nick Roberts, review by Richard Eisenberg and Jacques Garrigue)

  • #12118: stop storing names of predefined exceptions in the cu_required_globals field of compilation unit descriptors. (Sébastien Hinderer, review by Vincent Laviron)

  • #12125: Add Misc.print_see_manual and modify [@manual_ref] to accept lists for simpler printing of manual references (Stefan Muenzel, review by Florian Angeletti)

  • #12509: Use strict prototypes on primitives when generating a standalone bytecode executable (ocamlc -custom). (Antonin Décimo, review by Xavier Leroy)

Build system:

  • #11844: Reduce verbosity of make logs by printing program invocations in shorthand (eg OCAMLC foo.cmo). Setting V=1 recovers the old style (with full command-lines). (Xavier Leroy, Nicolás Ojeda Bär, review by Sébastien Hinderer)

  • #11590: Allow installing to a destination path containing spaces. (Élie Brami, review by Sébastien Hinderer and David Allsopp)

  • #11243, #11248, #11268, #11420, #11675: merge the sub-makefiles into the root Makefile. (Sébastien Hinderer, review by David Allsopp and Florian Angeletti)

  • #11828: Compile otherlibs/ C stubs in two version for native and bytecode (Olivier Nicole, review by Sébastien Hinderer and Xavier Leroy)

  • #12265: Stop adding -lexecinfo to cclibs (leftover debugging code from the multicore project). Harden the feature probe for -lm in configure so -lm is only added if strictly necessary. configure.ac now correctly propagates library flags for the Windows ports, allowing Windows OCaml to be configured with ZSTD support. (David Allsopp, review by Sébastien Hinderer)

  • #12372: Pass option -no-execute-only to the linker for OpenBSD >= 7.3 so that code sections remain readable, as needed for closure marshaling. (Xavier Leroy and Anil Madhavapeddy, review by Anil Madhavapeddy and Sébastien Hinderer)

Bug fixes:

  • #12062: fix runtime events consumer: when events are dropped they shouldn't be parsed. (Lucas Pluvinage)

  • #12132: Fix overcounting of minor collections in GC stats. (Damien Doligez, review by Gabriel Scherer)

  • #12017: Re-register finaliser only after calling user alarm in Gc.create_alarm (Fabrice Buoro, report by Sam Goldman, review by Guillaume Munch-Maccagnoni)

  • #11887, #11893: Code duplication in pattern-matching compilation (Vincent Laviron, report par Greta Yorsh, review by Luc Maranget and Gabriel Scherer)

  • #10664, #11600: Unsoundness in the typing of polymorphic methods involving polymorphic variants (Jacques Garrigue, report by Mike Shulman, review by Gabriel Scherer)

  • #11302, #11412: ocamlc and ocamlopt should not remove generated files when they are not regular files. (Xavier Leroy, report by Thierry Martinez, review by Anil Madhavapeddy, Nicolás Ojeda Bär, David Allsopp)

  • #10348, #10560, #11561: Expand GADT equations lazily during unification to avoid ambiguity (Jacques Garrigue, review by Leo White)

  • #11436: Fix wrong stack backtrace for out-of-bound exceptions raised by leaf functions. (Tom Kelly and Xavier Leroy, review by Mark Shinwell)

  • #11450, #12018: Fix erroneous functor error messages that were too eager to cast struct end functor arguments as unit modules in F(struct end). (Florian Angetti, review by Gabriel Scherer)

  • #11643: Add missing test declaration to float_compare test, so that it will run. (Stefan Muenzel, review by David Allsopp)

  • #11630: Use correct location when reporting record labels with non-existent paths. (Nicolás Ojeda Bär, report by Jason Gross, review by Gabriel Scherer)

  • #11727: Ensure push_defaults can push past module patterns, fixing an currying optimisation accidentally disabled by #10340. (Stephen Dolan, review by Gabriel Scherer)

  • #11771: Use a more relaxed mode for unification in Ctype.subst (Leo White, review by Jacques Garrigue and Gabriel Scherer)

  • #11803, #11808: on x86, the destination of an integer comparison must be a register, it cannot be a stack slot. (Vincent Laviron, review by Xavier Leroy, report by Emilio Jesús Gallego Arias)

  • #11809: Protect Parmatch.pats_of_type from missing cmis (Jacques Garrigue, review by Stephen Dolan and Gabriel Scherer)

  • #11824: Fix a crash when calling ocamlrun -b (Florian Angeletti, review by Sébastien Hinderer)

  • #11815: Marshalling continuations raises invalid argument exception. (Jérôme Vouillon, review by Nicolás Ojeda Bär, Stephen Dolan and Hugo Heuzard)

  • #11846: Mark rbx as destroyed at C call for Win64 (mingw-w64 and Cygwin64). Reserve the shadow store for the ABI in the c_stack_link struct instead of explictly when calling C functions. This simultaneously reduces the number of stack pointer manipulations and also fixes a bug when calling noalloc functions where the shadow store was not being reserved. (David Allsopp, report by Vesa Karvonen, review by Xavier Leroy and KC Sivaramakrishnan)

  • #11850: When stopping before the emit phase (using -stop-after), an empty temporary assembly file is no longer left in the file system. (Nicolás Ojeda Bär, review by Gabriel Scherer and Xavier Leroy)

  • #11866: Fix the result of caml_read_directory() on non-existent paths. (Andrei Paskevich and Charlène Gros, review by David Allsopp and Nicolás Ojeda Bär)

  • #11879: Bugfix for Ctype.nondep_type (Stephen Dolan, review by Gabriel Scherer)

  • #12004: Don't ignore function attributes on lambdas with locally abstract types. (Chris Casinghino, review by Gabriel Scherer)

  • #12037: Fix some data races by using volatile when necessary (Fabrice Buoro and Olivier Nicole, review by Guillaume Munch-Maccagnoni, Gabriel Scherer and Luc Maranget)

  • #12046: Flush stderr when tracing the parser (Hugo Heuzard, review by David Allsopp and Nicolás Ojeda Bär)

  • #12061, #12063: don't add inconsistent equalities when computing high-level error messages for functor applications and inclusions. (Florian Angeletti, review by Gabriel Scherer)

  • #12075: auto-detect whether ar support @FILE arguments at configure-time to avoid using this feature with toolchains that do not support it (eg FreeBSD/Darwin). (Nicolás Ojeda Bär, review by Xavier Leroy, David Allsop, Javier Chávarri, Anil Madhavapeddy)

  • #12103, 12104: fix a concurrency memory-safety bug in Buffer (Gabriel Scherer, review by Florian Angeletti, report by Samuel Hym)

  • #12112: Fix caml_callback{2,3}_exn when used with effect handlers. (Lucas Pluvinage, review by Gabriel Scherer, David Allsopp and Xavier Leroy)

  • #12134: Use ghost location for nodes created when handling defaults in optional arguments. (Paul-Elliot Anglès d'Auriac, review by Gabriel Scherer)

  • #12153: Fix segfault in bytecode programs involving recursive value definitions of values of size 0 (Vincent Laviron, Xavier Leroy, Gabriel Scherer, review by Xavier Leroy, report by Nick Roberts)

  • #12162: Fix miscompilation on amd64 backends involving integer overflows (Vincent Laviron and Greta Yorsh, review by Stefan Muenzel)

  • #12170: fix pthread_geaffinity_np configure check for android (David Allsopp, review by Sébastien Hinderer)

  • #12178: Fix runtime events consumer poll function returning an invalid value instead of an OCaml integer value. (Lucas Pluvinage)

  • #12252: Fix shared library build error on RISC-V. (Edwin Török, review by Nicolás Ojeda Bär and Xavier Leroy)

  • #12255, #12256: Handle large signal numbers correctly (Nick Barnes, review by David Allsopp).

  • #12277: ARM64, fix a potential assembler error for very large functions by emitting stack reallocation code before the body of the function. (Xavier Leroy, review by KC Sivaramakrishnan)

  • #12253, #12342: Fix infinite loop in signal handling. (Guillaume Munch-Maccagnoni, report by Thomas Leonard, review by KC Sivaramakrishnan and Sadiq Jaffer)

  • #12445: missing GC root registrations in runtime/io.c (Gabriel Scherer, review by Xavier Leroy and Jeremy Yallop)

  • #12481, #12505: Fix incorrect initialization of array expressions [|e1;...;eN|] when N is large enough to require major heap allocation. (Xavier Leroy, report by Andrey Popp, analysis by KC Sivaramakrishnan and Vincent Laviron, review by Gabriel Scherer)

  • #11150, #11207, #11936: Avoid recomputation in Typedecl.check_wellfounded (Jacques Garrigue, report by Boris Yakobowski, review by Gabriel Scherer)

  • #11186, #11188: Fix composition of coercions with aliases (Vincent Laviron, report and review by Leo White)

  • #12486: Fix delivery of unhandled effect exceptions on s390x (Miod Vallat, report by Jan Midtgaard, review by Vincent Laviron and Xavier Leroy)

Since last week, there were two significant bugs fixed in the OCaml 5.1.0 runtime (one overflow bug and a stack corruption bug in the s390x port). Since those bug fixes are as small as they are subtle, they were deemed worthy of a release of a third release candidate for OCaml 5.1.0.

If there are no more surprises this week, the release of OCaml 5.1.0 shall happen next week.

If you find any bugs, please report them on OCaml's issue tracker.

The full changelog for OCaml 5.1.0 is available on GitHub

A short summary of the two fixed bugs in this release candidate is also available below.


Installation Instructions

The base compiler can be installed as an opam switch with the following commands on opam 2.1 and later:

opam update
opam switch create 5.1.0~rc3

The source code for the release candidate is also directly available on:

Fine-Tuned Compiler Configuration

If you want to tweak the configuration of the compiler, you can switch to the option variant with:

opam update
opam switch create <switch_name> ocaml-variants.5.1.0~rc3+options <option_list>

where <option_list> is a comma-separated list of ocaml-option-* packages. For instance, for a flambda and no-flat-float-array switch:

opam switch create 5.1.0~rc3+flambda+nffa ocaml-variants.5.1.0~rc3+options ocaml-option-flambda ocaml-option-no-flat-float-array

All available options can be listed with opam search ocaml-option.

See full changelog

Last Second Bug Fixes

  • #11284, +#12525: Use compression of entries scheme when pruning mark stack. Can decrease memory usage for some workloads, otherwise should be unobservable. (Tom Kelly, review by Sabine Schmaltz, Sadiq Jaffer, and Damien Doligez)

  • #12486: Fix delivery of unhandled effect exceptions on s390x (Miod Vallat, report by Jan Midtgaard, review by Vincent Laviron and Xavier Leroy)

In the last two weeks, two significant bugs have been discovered in the release candidate for OCaml 5.1.0 (one affecting the type system, another in the runtime).

Those bugs are now fixed and we are publishing a second release candidate to check that everything is in order before the release in the upcoming week.

If you find any bugs, please report them on OCaml's issue tracker.

The full changelog for OCaml 5.1.0 is available on GitHub. A short summary of the two fixed bugs in this release candidate is also available below.


Installation Instructions

The base compiler can be installed as an opam switch with the following commands on opam 2.1 and later:

opam update
opam switch create 5.1.0~rc2

The source code for the release candidate is also directly available on:

Fine-Tuned Compiler Configuration

If you want to tweak the configuration of the compiler, you can switch to the option variant with:

opam update
opam switch create <switch_name> ocaml-variants.5.1.0~rc2+options <option_list>

where <option_list> is a comma-separated list of ocaml-option-* packages. For instance, for a flambda and no-flat-float-array switch:

opam switch create 5.1.0~rc2+flambda+nffa ocaml-variants.5.1.0~rc2+options ocaml-option-flambda ocaml-option-no-flat-float-array

All available options can be listed with opam search ocaml-option.

See full changelog

Last Minute Bug Fixes

Type System:

  • (breaking change) #6941, #11187, +#12483: prohibit using classes through recursive modules inheriting or including a class belonging to a mutually-recursive module would previous behave incorrectly, and now results in a clean error. (Leo White, review by Gabriel Scherer and Florian Angeletti)

Runtime

  • #12481, #12505: Fix incorrect initialization of array expressions [|e1;...;eN|] when N is large enough to require major heap allocation. (Xavier Leroy, report by Andrey Popp, analysis by KC Sivaramakrishnan and Vincent Laviron, review by Gabriel Scherer)

We're thrilled to announce the release of Merlin 4.10, which comes with many bug fixes and improvements.

One of the standout features of this release is the significantly enhanced support for binding operators like let+ and and+. You'll find that the results from type-enclosing on expressions that contain let bindings are now more precise.

See full changelog
  • merlin binary
    • Constrain socket path buffer size to avoid build warnings (#1631)
    • Handle concurrent server start (#1622)
    • Omit module prefixes for constructors and record fields in the construct command (#1618). Prefixes are still produced when warning 42 (disambiguated name) is active.
    • Correctly invalidate PPX cache when pipeline ran partially (#1650, fixes #1647)
    • Prevent short-path from looping in some cases related to recursive type definitions (#1645)
    • Support parsing negative numbers in sexps (#1655)
    • Fix construct not working with inline records (#1658)
    • Improve behavior of type-enclosing on let/and operators (#1653)
    • Fix occurrences of extension constructors (#1662)
    • Improve node selection when ghosts are present (#1664, fixes #1660)
  • editor modes
    • emacs: call merlin-client-logger with "interrupted" if the merlin binary itself is interrupted, not just the parsing of the result (#1626).
    • emacs: merlin-construct, with a prefix argument, now includes local values in the completion options. Alternatively, this behavior can be enabled permanently by customizing merlin-construct-with-local-values (#1644)
    • emacs: add support for opam-switch-mode (#1654, fixes #1591). See https://github.com/ProofGeneral/opam-switch-mode

Odoc 2.2.1

As OCaml 5.1 is on the horizon, we're excited to announce the release of odoc 2.2.1. This latest version brings compatibility with the upcoming OCaml 5.1 release.

See full changelog

Additions

  • OCaml 5.1.0 compatibility (@Octachron, #956)

The release of OCaml 5.1.0 is imminent. As a final step, we are publishing a release candidate to check that everything is in order before the release in the upcoming week(s).

If you find any bugs, please report them on OCaml's issue tracker.

Compared to the beta release, this release contains one safe runtime fix and two configuration tweaks.

The full change log for OCaml 5.1.0 is available on GitHub A short summary of the changes since the beta release is also available below.


Installation Instructions

The base compiler can be installed as an opam switch with the following commands on opam 2.1 and later:

opam update
opam switch create 5.1.0~rc1

The source code for the release candidate is also directly available on:

Fine-Tuned Compiler Configuration

If you want to tweak the configuration of the compiler, you can switch to the option variant with:

opam update
opam switch create <switch_name> ocaml-variants.5.1.0~rc1+options <option_list>

where <option_list> is a comma-separated list of ocaml-option-* packages. For instance, for a flambda and no-flat-float-array switch:

opam switch create 5.1.0~rc1+flambda+nffa ocaml-variants.5.1.0~rc1+options ocaml-option-flambda ocaml-option-no-flat-float-array

All available options can be listed with opam search ocaml-option.

See full changelog

Changes Since the Beta

Bug Fix

  • #12445: missing GC root registrations in runtime/io.c (Gabriel Scherer, review by Xavier Leroy and Jeremy Yallop)

Configuration Fix (openBSD)

  • #12372: Pass option -no-execute-only to the linker for OpenBSD >= 7.3 so that code sections remain readable, as needed for closure marshaling. (Xavier Leroy and Anil Madhavapeddy, review by Anil Madhavapeddy and Sébastien Hinderer)

Tool Fix (ocamlmktop)

  • #11745, #12358: Debugger and toplevels: embed printer types rather than reading their representations from topdirs.cmi at runtime. This change also removes the ocamlmktop initialisation module introduced in #11382 which was no longer useful. This change breaks toplevel scripts relying on the visibility of Topdirs in the initial toplevel environment without loading topfind. Since the opam default .ocamlinit file loads topfind, it is expected that only scripts run with ocaml -noinit are affected. For those scripts, accessing Topdirs now requires the compiler-libs directory to be added to the toplevel search path with
      #directory "+compiler-libs";;
    

as was already the case for the other modules in the toplevel interface library. (Sébastien Hinderer, review by Florian Angeletti, Nicolás Ojeda Bär, and Gabriel Scherer)

Documentation Changes

  • #12201: in the tutorial on modules, replace priority queue example by a simpler example based on FIFO queues. (Xavier Leroy, review by Anil Madhavapeddy and Nicolás Ojeda Bär).

  • #12352: Fix a typo in the documentation of Arg.write_arg (Christophe Raffalli, review by Florian Angeletti)

Dune 3.9.3

The fix to sendfile in 3.9.2 was not quite enough so here is the last part of the fix. It brings compatibility with filesystems where sendfile is not available, in particular when ecryptfs is used.

See full changelog

Fixes

  • Fix flushing when using sendfile fallback (#8288, fixes #8284, @alan-j-hu)

Dune 3.10.0

We're happy to announce the release of Dune 3.10.0. It comes with some internal fixes as well as some interesting features:

  • some changes in dune describe that will allow a better implementation of opam-dune-lint to check the consistency between library and package dependencies
  • more commands made available under dune show for a more consistent command line interface
See full changelog

Added

  • Add dune show rules as alias of the dune rules command. (#8000, @Alizter)

  • Add dune show installed-libraries as an alias of the dune installed-libraries command. (#8135, @Alizter)

  • Add dune build --dump-gc-stats FILE argument to dump garbage collection stats to a named file. (#8072, @Alizter)

  • Add dune describe package-entries to print all package entries (#7480, @moyodiallo)

Fixed

  • Fix %{deps} to expand properly in (cat ...) when containing 2 or more items. (#8196, @Alizter)

  • Fix the severity of error messages sent over RPC which was missing. (#8193, @Alizter)

  • Fix bug with ppx and Reason syntax due to missing dependency in sandboxed action (#7932, fixes #7930, @Alizter)

Changed

  • Improve dune describe external-lib-deps by adding the internal dependencies for more information. (#7478, @moyodiallo)

  • Re-enable background file digests on Windows. The files are now open in a way that prevents race condition around deletion. (#8262, fixes #8268, @emillon)

If you want to contribute to a new release announcement, check out the Contributing Guide on GitHub.