Feature #17303: Remove webrick from stdlib

• Status: Closed
• Assignee: hsbt (Hiroshi SHIBATA)
• Target version: 3.0

[hsbt (Hiroshi SHIBATA)]

I propose to move webrick to bundled gems or remove it from stdlib of ruby.

We have several vulnerability issues in webrick gem.

https://www.ruby-lang.org/en/news/2020/09/29/http-request-smuggling-cve-2020-25613/

The ruby core team don't have enough time to handle them. We should remove webrick from default gems at least.

Patch for this feature: https://github.com/ruby/ruby/pull/3729

[jeremyevans0 (Jeremy Evans)]

I am in favor of this change. I prefer removing webrick from stdlib, as otherwise we are still likely to be shipping vulnerable code if there is a security issue in webrick. Moving webrick from default gems to bundled gems doesn't change much security wise, other than making it slightly more difficult to use an separately installed webrick gem.

[Eregon (Benoit Daloze)]

Doesn't RubyGems depend on WEBrick (notably for gem server)?
It seems also RDoc depends on it.
And I know ruby -run -e httpd . -p8080 depends on it as well.

I think having a basic HTTP server in stdlib is important (bundled gem is fine for that).
Notably for properly testing Socket and new IO APIs.

Also removing it entirely without any kind of deprecation first seems like it might break lots of things.

[Eregon (Benoit Daloze)]

It seems like the more fundamental thing we need is more maintainers for WEBrick.
Even if it's removed from stdlib, people will still install it for some existing use cases.

[sawa (Tsuyoshi Sawada)]

• Subject changed from Make webrick to bundled gems or remove from stdlib to Move webrick to bundled gems or remove it from stdlib
• Description updated (diff)

[mame (Yusuke Endoh)]

@normalperson (Eric Wong) What do you think?

[normalperson (Eric Wong)]

[e-mail] wrote:

@normalperson (Eric Wong) What do you think?

I don't really care about Ruby, at the moment...

But if I were paid to care in the future, there is absolutely no
way I could use GitHub (due to Terms-of-Service, JS CAPCHA, etc)
and there needs to be a mailing list (e.g. ruby-core) and
(optional) Redmine tracker for me to work on it.

[mame (Yusuke Endoh)]

Thank you for your reply.
I'm sorry but I cannot afford to hire you and have no bugdet. If I were a billionaire..
Anyway, you are not against the removal of WEBrick from Ruby package.
If you are, let me know.

This is my opinion.
If you are willing to continue to maintain WEBrick, GitHub is not mandatory.
We will keep WEBrick on GitHub, but I think we can transfer the source code to you everytime.
You can maintain it in your own Git server and mailing list, like the unicorn project, if you want.
And I hope so if possible.

@matz (Yukihiro Matsumoto), do you decide the removal of WEBrick?

[ioquatix (Samuel Williams)]

Personally, I'm in favour of a smaller stdlib. Therefore I support this change.

I also believe that we see more innovation when things are allowed to grow under their own terms rather than part of stdlib.

[hsbt (Hiroshi SHIBATA)]

• Subject changed from Move webrick to bundled gems or remove it from stdlib to Remove webrick from stdlib
• Status changed from Open to Assigned

[matz (Yukihiro Matsumoto)]

Go ahead.

Matz.

[naruse (Yui NARUSE)]

Recently vulnerability handling for webrick is heavy load for CRuby development.
Reports related to webrick is low S/N rate though the importance of bundling webrick with ruby tarball is decreasing.

We remove webrick in ruby repo and separate it as dedicated project.

[mame (Yusuke Endoh)]

• Target version set to 3.0

[hsbt (Hiroshi SHIBATA)]

• Status changed from Assigned to Closed

Removed at https://github.com/ruby/ruby/pull/3729

I'm going to resolve the issues related this until the final release of Ruby 3.0.

[hsbt (Hiroshi SHIBATA)]

• Related to Feature #15657: Make webrick to bundled gems added