- ruby-talk - ml.ruby-lang.org

[ruby-talk:444715] [ANN] minitest-sprint 1.5.0 Released
by Ryan Davis 11 Dec '25

11 Dec '25

minitest-sprint version 1.5.0 has been released! * home: <https://github.com/seattlerb/minitest-sprint> * rdoc: <http://docs.seattlerb.org/minitest-sprint> Runs (Get it? It's fast!) your tests and makes it easier to rerun individual failures. Changes: ### 1.5.0 / 2025-12-11 * 10 minor enhancements: * --rerun now takes a style (lines, names) and defaults to lines. * Add --simplecov (and MT_COV=1) option to load before minitest. * Added all the filtering options to minitest/complete. * Folded BinstubReporter into SprintReporter and added line number reporting. * Folded rake_reporter.rb and sprint_reporter.rb into sprint.rb * Normalized namespacing inside of Minitest::Sprint * Renamed --binstub to --rerun and added -r shortcut. * Switched minitest/complete to prism. Should be much more solid. * Vendored latest version of path_expander. * Updated Minitest::VendoredPathExpander to 2.0.0 and cleaned up overrides. * 4 bug fixes: * -I flag now prepends onto load path. * Fixed minitest cmdline to use require_relative to test against current version. * Made RakeReporter play better with superclass. * Prepend test and lib onto load path before loading minitest.

1 0

[ruby-talk:444714] [ANN] path_expander 2.0.0 Released
by Ryan Davis 11 Dec '25

11 Dec '25

path_expander version 2.0.0 has been released! * home: <https://github.com/seattlerb/path_expander> * rdoc: <http://docs.seattlerb.org/path_expander> PathExpander helps pre-process command-line arguments expanding directories into their constituent files. It further helps by providing additional mechanisms to make specifying subsets easier with path subtraction and allowing for command-line arguments to be saved in a file. NOTE: this is NOT an options processor. It is a path processor (basically everything else besides options). It does provide a mechanism for pre-filtering cmdline options, but not with the intent of actually processing them in PathExpander. Use OptionParser to deal with options either before or after passing ARGV through PathExpander. Changes: ### 2.0.0 / 2025-12-11 * 3 major enhancements: * Added #pre_process and #post_process and call them from #process. * Modified #process to return self if block given, otherwise Enumerator. * Modified #process to take a block for processing files directly.

1 0

[ruby-talk:444713] [ANN] path_expander 2.0.0 Released
by Ryan Davis 11 Dec '25

11 Dec '25

path_expander version 2.0.0 has been released! * home: <https://github.com/seattlerb/path_expander> * rdoc: <http://docs.seattlerb.org/path_expander> PathExpander helps pre-process command-line arguments expanding directories into their constituent files. It further helps by providing additional mechanisms to make specifying subsets easier with path subtraction and allowing for command-line arguments to be saved in a file. NOTE: this is NOT an options processor. It is a path processor (basically everything else besides options). It does provide a mechanism for pre-filtering cmdline options, but not with the intent of actually processing them in PathExpander. Use OptionParser to deal with options either before or after passing ARGV through PathExpander. Changes: ### 2.0.0 / 2025-12-11 * 3 major enhancements: * Added #pre_process and #post_process and call them from #process. * Modified #process to return self if block given, otherwise Enumerator. * Modified #process to take a block for processing files directly.

1 0

[ruby-talk:444712] [ANN] rodauth-oauth 1.6.4 released
by Tiago Cardoso 11 Dec '25

11 Dec '25

rodauth-oauth 1.6.4 has been released. rodauth-oauth is a rack-compatible toolkit for building OAuth 2.0 authorization servers, as well as OpenID Authentication Providers. rodauth-oauth is certified <https://openid.net/certification/> for the following profiles of the OpenID Connect™ protocol: Basic OP, Implicit OP, Hybrid OP, Config OP, Dynamic OP, Form Post OP, 3rd Party Init OP Session Management OP, RP-Initiated Logout OP, Front-Channel OP, Back-Channel OP # as simple as rodauth do enable :oauth_authorization_code_grant # or enable :oidc end Among its features, it supports: * Authorization Code Grant * Refresh Token Grant * Implicit Grant * Client Credentials Grant * Device Code Grant * Token Revocation * Token Introspection * Auth Server Metadata * PKCE * Resource Indicators * JWT Access Tokens * mTLS Client Authentication * Assertion Framework * SAML 2.0 Bearer Assertion Grant * JWT Bearer Assertion Grant * JWT Secured authorization requests (JAR) * JWT Secured authorization response mode (JARM) * Pushed Authorization requests (PAR) * Demonstrating Proof-of-Possession at the Application Layer (DPoP) * Dynamic Client Registration * OpenID * OpenID Discovery * OpenID Multiple Response types * OpenID Self Issued Tokens * OpenID Connect Dynamic Client Registration * OpenID Session Management * OpenID RP Initiated Logout * OpenID Frontchannel Logout * OpenID Backchannel Logout It can also be used with Rails (via the "rodauth-rails" gem). Website: rodauth-oauth · honeyryder <https://honeyryderchuck.gitlab.io/rodauth-oauth/> Documentation: Rodauth OAuth: OAuth 2.0 and OpenID for rodauth <https://honeyryderchuck.gitlab.io/rodauth-oauth/rdoc/> Wiki: Home · Wiki · OS / rodauth-oauth · GitLab <https://gitlab.com/honeyryderchuck/rodauth-oauth/wikis/home> CI: https://gitlab.com/honeyryderchuck/rodauth-oauth/pipeline These are the release notes since the last update: ### 1.6.4 #### Bugfixes * on `/userinfo` request, enforce existence of oauth grant before proceeding with the userinfo requests. #### Chore * linting + tweaks to example scripts. ### 1.6.3 #### Improvements * all routes can now be used via rodauth `internal_request` feature. #### Bugfixes * `oauth_application_management` feature: fixed "new oauth application" link to account for for prefix usage. ### 1.6.2 #### Bugfixes * fix oauth_base `supported_response_type?` when `grant_types` column is set in the database, but `response_types` is not. ### 1.6.1 #### Bugfixes Fixed regressions in rails generated templates: * some `:oauth_dpop` form fields not being rendered; * `check_box_tag` rails helper usage was wrongly setting all checkboxes to "checked" #### Chore * Updated latest features (`:oauth_dpop`) options documentation. * rails tests will now exercise the rails generated templates again.

1 0

[ruby-talk:444711] [ANN] cel 0.5.0 released
by Tiago Cardoso 11 Dec '25

11 Dec '25

cel 0.5.0 has been released. cel is a pure Ruby implementation of Google Common Expression Language, https://opensource.google/projects/cel. The Common Expression Language (CEL) implements common semantics for expression evaluation, enabling different applications to more easily interoperate. ```ruby require "cel" # set the environment env = Cel::Environment.new(name: :string, group: :string) # parse ast = env.compile('name.startsWith("/groups/" + group)') # check prg = env.program(ast) # evaluate prg.evaluate(name: Cel::String.new("/groups/acme.co/documents/secret-stuff "), group: Cel::String.new("acme.co")) #=> true # or do it all in one go env.evaluate('name.startsWith("/groups/" + group)', name: Cel::String.new("/groups/acme.co/documents/secret-stuff"), group: Cel::String.new("acme.co") ) ``` Here are the updates since the last release: ## [0.5.0] - 2025-12-11 ### Features #### Custom extensions A new `:extensions` kwarg is added to `Cel::Environment.new` which allows adding custom extensions, in a similar manner as what the standard extensions (like `math` or `string`) are done: ```ruby module Ext # defines a random function which takes no arguments and returns 42 end Cel::Environment.new.evaluate("ext.random()") #=> raises error Cel::Environment.new(extensions: { ext: Ext }).evaluate("ext.random()") #=> 42 ``` ### Backwards Compatibility The ractor safety introduced in 0.4.1 has been relaxed in order to allow extensions of core classes by custom extensions, And you'll need to explicitly call `Cel.freeze` before using `cel` inside ractors. This is a direct consequence of how extensions patch `cel` core classes. ATTENTION: Changes may be introduced in the way core classes are patched by extensions, towards making `cel` ractor-safe by default. If you rely on custom extensions, do follow the migration instructions in subsequent releases. ### Bugfixes Fixed checker type inference when using nexted expressions (like when using the `bind` extensions to evaluate cel sub-expressions). ## [0.4.1] - 2025-11-25 ### Improvements * Literal class can now mark which methods are CEL directives, the remainder being lib private helpers. * `cel` is now ractor compatible. * Documentation on how to support abstract types has been added. ### Security A remote execution attack vector has been fixed, which allowed executing arbitrary Ruby code within a CEL expression when calling functions on a variable declared as a CEL map. Example: ```ruby env = Cel::Environment.new(declarations: { webhook: :map }) env.evaluate("webhook.payload.send('eval', 'File.write(\"test.txt\", \"Hello, world!\")')", webhook: { payload: {} }) ```

1 0

[ruby-talk:444710] [ANN] llm.rb v2.1.0
by Robert 06 Dec '25

06 Dec '25

Hi everyone, I'd like to share and announce the release of llm.rb v2.1.0. The source is available at https://github.com/llmrb/llm.rb > What is llm.rb? llm.rb is a zero-dependency Ruby toolkit for Large Language Models that includes OpenAI, Gemini, Anthropic, xAI (Grok), zAI, DeepSeek, Ollama, and LlamaCpp. The toolkit includes full support for chat, streaming, tool calling, audio, images, files, and structured outputs. > Example A simple chatbot that maintains a conversation and streams responses in real-time: #!/usr/bin/env ruby require "llm" llm = LLM.openai(key: ENV["KEY"]) bot = LLM::Bot.new(llm, stream: $stdout) loop do print "> " bot.chat(gets) print "\n" end All the best, Robert

1 0

[ruby-talk:444675] [ANN] Jarbler 0.4.3 released
by Peter Ramm 03 Dec '25

03 Dec '25

Pack a Ruby application into an executable jar file. Jarbler creates a self executing Java jar file containing a Ruby application and all its Gem dependencies. The included Ruby program is directly executed inside the JVM using the JRuby runtime jars. https://github.com/rammpeter/jarbler Changes: https://github.com/rammpeter/jarbler/blob/main/CHANGELOG.md

4 4

[ruby-talk:444692] Ruby, Rubinius, and AI
by Brian Shirai 01 Dec '25

01 Dec '25

Hello Rubyists! If you're interested in Ruby+AI, you may be curious about some recent developments with Rubinius. I've written them up here: https://vivarium-ai.com/newsletters/0-hello-world-who-is-out-there/ Cheers, Brian

1 1

[ruby-talk:444707] [ANN] DB GUI 0.3.0 & Glimmer DSL for LibUI 0.13.1 Released
by Andy Maleh 01 Dec '25

01 Dec '25

DB GUI 0.3.0 & Glimmer DSL for LibUI 0.13.1 have been released. DB GUI now supports remembering multiple database profiles, including which one was last selected. Glimmer DSL for LibUI now supports combobox items data-binding. https://raw.githubusercontent.com/AndyObtiva/db-gui/refs/heads/master/scree… Blog post announcement : https://andymaleh.blogspot.com/2025/11/db-gui-030-glimmer-dsl-for-libui-013… GitHub Projects : DB GUI : https://github.com/AndyObtiva/db-gui Glimmer DSL for LibUI : https://github.com/AndyObtiva/glimmer-dsl-libui Ruby Gems: db-gui : https://rubygems.org/gems/db-gui glimmer-dsl-libui : https://rubygems.org/gems/glimmer-dsl-libui

1 0

[ruby-talk:444700] The Embarrassing Ruby/Rails Subreddit Chronicles 2025-11-24
by Andy Maleh 25 Nov '25

25 Nov '25

Welcome to another edition of the Embarrassing Ruby/Rails Subreddit Chronicles. Attached are screenshots for another embarrassing incident demonstrating the mean/hateful nature of Ruby/Rails subreddit members and how they downvote without first having a nice and loving patient discourse in case of a misunderstanding or an uninformed disagreement. It hit me the other day that the reason I suffer from mean/hateful behaviour in the Ruby/Rails subreddits despite being a 2-time Fukuoka international competition winner (of awards by Matz, the creator of Ruby), a 4-time RubyConf speaker, and a 2-time RailsConf speaker is because they allow in people who are below the minimum bar of professionalism and respectful conduct that is required for hiring in a Software Engineering job at a respectful ethical company. I mean I wouldn't hire any of them in a million years, not just for lack of technical skills, but more importantly for not having the patient loving nice Rubyist personality. So, why do the Ruby/Rails subreddit mods allow in unnice/hateful people who are below the minimum bar of respectful conduct?! This is a negative reflection on the mods just as much as it is on the trash members of the Ruby/Rails subreddits. I say trash because they're no different from trash criminals and drug addicts I'd encounter on the streets in their behaviour. It's not the behaviour of a member of the Software Engineering community, let alone the Ruby/Rails community. In this screenshot, you can see a Ruby subreddit moderator asking for feedback about new rules, which in my opinion are embarrassingly fascist, even if I didn't say that in my comment about them. https://imgur.com/a/yS0W8z9 Here, you can see my response, and how it was immediately downvoted to 0 within 5 minutes of posting, demonstrating the mean hateful non-open-minded unsympathetic nature of Ruby subreddit members (not sympathizing with someone who says they received hateful behaviour means you are a hateful person). https://imgur.com/a/5aExMUF Of course, fascist rules are a red herring. They're a side effect to Ruby subreddit mods holding on to terrorists that they harbour within their group instead of letting them go as explained in this sentence: "If you worry too much about the rules, it means you have the wrong kind of people in the group, and that’s the real problem not being dealt with (eg allowing in haters of Ruby, people who don’t give Ruby the benefit of the doubt in applications that it’s not common in, people who are below the minimum bar of ethical respectful conduct and professionalism, etc…)." Here is my full comment, repeated here textually: "This group allows a lot of hate on Frontend Ruby. It also literally allows people who don’t truly like Ruby or get it to freely discuss JavaScript as better than Ruby while making true Rubyists who understand and appreciate Ruby’s benefits over JavaScript not feel safe in a Ruby group. This automatically renders the group a bad one. Also, I’ve encountered a lot of hateful behavior in this group against anything that’s novel and outside the box of what’s common in the software development community. 15 years ago, Rubyists were open minded about new novel ideas and patiently listened to them without downvoting. In this group today, such ideas get downvoted to zero in a very hateful unintelligent way without giving the poster the benefit of the doubt or even attempting to understand the benefits of what is shared. That discourages Rubyists from sharing new ideas or exploring them in this group in a respectful intelligent open minded manner. I personally make an effort not to downvote anyone, yet to ask questions and facilitate discussions if I disagree with something. In general, rules that are unethical or do not implicitly respect everyone equally only make the group a bad one. Lastly, politics are off-topic. Honestly, this is not a good group if it allows off-topic things while attempting to shift the blame unto the people who are right about the matter. That’s attempting to get out of responsibility, which is also bad. If you don’t understand this, you’re on the wrong side of this. I’m apolitical and don’t vote by choice by the way, so I don’t care to discuss this further because I don’t like discussing politics ever, especially not in a Ruby group not about politics. If you worry too much about the rules, it means you have the wrong kind of people in the group, and that’s the real problem not being dealt with (eg allowing in haters of Ruby, people who don’t give Ruby the benefit of the doubt in applications that it’s not common in, people who are below the minimum bar of ethical respectful conduct and professionalism, etc…). Update: Exhibit A: I have won 2 awards from Matz in very difficult international engineering competitions, and have spoken at RubyConf 4 times, and yet people don’t respect me, discriminate against me and hate me, unintelligently downvoting my posts out of hate and discrimination not for any intelligent reasons. Of course, this behavior is only an indirect reflection on the mods allowing hateful disrespectful people into a Ruby group without respecting the Ruby luminaries of the Ruby community. I was respected more 10-15 years ago when I wasn’t a very well accomplished Ruby Software Engineer back then because the Ruby group was run better back then." Blog post edition for sharing: https://andymaleh.blogspot.com/2025/11/the-embarrassing-rubyrails-subreddit…

5 5