On Mon, 26 Nov 2012, Kohsuke Kawaguchi wrote:

> 
> Hi,
> 
> Moving this conversations to [email protected].
> 
> Yes, what you are talking about is [1], which is an attempt to
> define blackbox acceptance tests in a way that lets us test our
> final deliverables (rpm, deb, etc.) It contains the page objects,
> among other things.
> 
> I think having a java library that defines page objects for Jenkins
> is useful. Go for it!


When I first started working on the selenium-tests repo, I tried to come up
with some page-object abstractions that made sense but I just couldn't find
anything that I didn't find crappy. Some of the abstractions that I ended up
defining were more focused on logical objects (Job, Build, Slave, etc).


> There's also a conversation of adding another functionality (either
> in this plugin or another one) that exposes Groovy scripting console
> in a way that lets you verify side-effects that happen on the server
> more easily.


I'm not sure if you're alluding to hooking groovy-based validation into the
selenium-tests but I would caution against it. There be dragons! Such
functionality would start effectively breaking down the testing pyramid (unit,
integration, UI tests), and becomes a huge pain in the ass over time.


Integration testing to verify side effects of things internal to Jenkins, I'm
all for though :)


> And then maybe it's possible to run selenium-tests in JRuby so that
> we can run some Java-based selenium tests in the same environment,
> which allows you to leverage pluggable layers in the selenium-tests
> and VirtualBox integrations.

Ick, I'm not going to tell anybody that's volunteering to write tests how to do
it, but mixing different platforms is going to be a lot of work, even with
JRuby, it's no silver bullet.


Just my few cents :)

Cheers
- R. Tyler Croy
--------------------------------------
    Code: https://github.com/rtyler
 Chatter: https://twitter.com/agentdero

Attachment: signature.asc
Description: Digital signature

Reply via email to