docs: improve wording in the "CI on MRI" section

Signed-off-by: Edwin Garcia <egarciavalle2014@gmail.com>
ko1 · Jan 25, 2024 · eb627d5 · eb627d5
1 parent ecf1118
commit eb627d5
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/EN/1_culture.md b/EN/1_culture.md
@@ -93,7 +93,7 @@ The list of Ruby's supported platforms can be found at: <https://bugs.ruby-lang.
 Because MRI is used in many enviroments, it is preferable to run tests on as wide a variety of environments as possible.
 It's common practice to use Continuous Integration (CI) to run automated tests. Ruby is no exception.
 
-In addition to using the popular Travis-CI service, we also run the <http://rubyci.org> site to collect test results run on a wider variety of environments. Typically, a CI system uses its own computing resources. However, our resources are limited. So, instead of preparing and managing the computers for the multitude of environments we need, we gather the results from tests run by volunteers in the community who run tests on their own computing resources. The tool [chkbuild](https://github.com/ruby/chkbuild) builds Ruby, runs tests, generates results, and performs a diff on the output so that we can determine which versions of Ruby have particular bugs.
+In addition to using the popular Travis-CI service, we also run the <http://rubyci.org> site to collect the results of tests performed on a wider variety of environments. Typically, a CI system uses its own computing resources. However, our resources are limited. So, instead of preparing and managing the computers for the multitude of environments we need, we gather the results from tests run by volunteers in the community who run tests on their own computing resources. The tool [chkbuild](https://github.com/ruby/chkbuild) builds Ruby, runs tests, generates results, and performs a diff on the output so that we can determine which versions of Ruby have particular bugs.
 
 `chkbuild` is good test/CI framework but, for various reasons (for example, chkbuild downloads source code each time) it can be quite slow (typically tens of minutes). To overcome this limitation, we use another CI system <http://ci.rvm.jp/> that can reuse previous builds, and can build/test in parallel, reducing the time required for testing to the order of 2-3 minutes. This allows us to run our tests hundreds of times every day, which can be helpful for revealing hard-to-reproduce bugs (e.g. timing bugs).