Today at the CukeUp conference in London, Aslak Hellesøy spoke about the current development and plans for Cucumber.
Hellesøy said that Cucumber had over 120000 downloads and he estimates that there are 20000 people using it now. Since it was first released, more than 300 people contributed to this open source project. Apart from the main project, there are more than 300 affiliated projects related to Cucumber on GitHub. Based on this, Hellesøy claims that Cucumber is today the most popular tool for automated acceptance testing.
In terms of future developments, he identified four important areas that he will be focusing on:
- Improving documentation: Hellesøy and Matt Wynne are working on a Cucumber book.
- A native Java implementation: Cuke4Duke allows teams to run Cucumber on JVM through JRuby but this is slow and tricky to set up. A native Java implementation will be better integrated into the Java development ecosystem, for example it will run from JUnit, and it will be faster according to Hellesøy.
- Improving internal design: while working on the native Java version, Hellesøy identified several areas that could be improved internally even in the Ruby version. He mentioned that the formatter API will change soon and that configuration will be much more flexible in the future – allowing teams to configure Cucumber from Ruby, not just text files. Another improvement will be the possibility to embed Cucumber and run it as a library, not just as a command line tool
- A WYSIWYG editor – under the working name ACE – should allow business users to access and edit Cucumber feature files easier. Hellesøy announced that ACE will be web based and provide syntax validation, highlighting, auto-completion and integration with version control systems. An early preview version of the editor is now live on http://cukes.info/trycuke/ – with syntax highlighting functionality.
|
|
I tried http://cukes.info/trycuke/ but I can’t see any syntax highlighting functionality.
Google Chrome 11
Firefox 4
This question might sound a bit nasty, but isn’t a pure Java version of Cucumber already available, named JBehave
?
But I definitly agree that the current Java support in Cucumber is more than tricky and as we have to use it from/with Java I stepped back from Cucumber and had a closer look at JBehave instead.
@gojko: The Gherkin/Cucumber editor will be based on http://ace.ajax.org – I’m just adding Gherkin smarts to it.
@beo: Currently the Gherkin/Cucumber editor will only detect syntax errors (by parsing the source with the javascript gherkin parser) and highlighting the line where a syntax error occurs (if any). Syntax highlighting, code completion and source control integration will come later.
@Thomas – JBehave, Cucumber’s grandfather is of course there to use today, but I think Cucumber.java will have more to offer. Stay tuned!
560.000 downloafs actually: http://rubygems.org/gems/cucumber
This is awesome news! I am especially psyched for a native Java implementation. I love cucumber but getting cuke4duke configured for various JVM projects (grails, groovy, jruby, etc) could be taxing at times.