I would like to have the old tests run with Watir 3.0 before we release. Do
we need a volunteer to take this on?

If they are finding valid issues, then we should consider migrating those
tests to WatirSpec.

Bret

On Tue, Feb 7, 2012 at 12:56 AM, Jarmo <[email protected]> wrote:

> I  don't think that we should increase the changes even further. The less
> big bang it is the better.
>
> All failures in WatirSpec are failures which are failing with latest final
> version of Watir too. Most of the failures are from nested :xpath usage
> (element.nested_element(:xpath => "..") won't work). #send_keys works like
> it did before and Tables API has some differences and then there's some new
> API-s like Cookie API, which isn't there in Watir yet anyway.
>
> In that light i'm not seeing that it has any critical failures since as
> said above everything which is failing in WatirSpec is the same in latest
> final Watir too and using xpath is highly discouraged anyway.
>
> I'm not sure about existing tests though. If you noticed there any
> failures which doesn't make sense then we should look at those, but if
> there are failures just caused by backwards incompatibility changes then
> they're okay.
>
> My point is that if there aren't any failures, which didn't exist before
> or if failures are caused by incompatible changes, which are unavoidable
> due to API differences then we shouldn't wait "until everything is ready"
> since it would just take too much time to come out with a release. I'm
> always endorsing "fail fast, fail often, release often" cycles :) This
> approach wouldn't make developers to feel too comfy :)
>
> On Tue, Feb 7, 2012 at 12:26 AM, Hugh McGowan <[email protected]>wrote:
>
>> I've been thinking more about this release and I think there are some
>> additional things we need to do before rolling this out. By moving to
>> WatirSpec we're basically saying that we want Watir and Watir-WebDriver to
>> have the same API. I think this is fantastic but I don't feel like we're
>> there yet. We still need to:
>>
>> * Make all WatirSpec tests pass for Watir
>> * Look at the old Watir tests not covered by WatirSpec and migrate those
>> tests into WatirSpec
>> * Remove old Watir tests covered by WatirSpec
>> * Look at our API vs Watir-Webdriver. We should understand where we
>> differ and have a good reason why we might deviate.
>>
>> My concern is that we're going to break existing functionality. If I run
>> the old Watir tests I'm seeing 56 failures/errors. If I run the new
>> WatirSpec tests I see 32 failures.
>>
>> I guess I feel a little naked without passing tests :)
>>
>> Hugh
>>
>>
>> _______________________________________________
>> Wtr-development mailing list
>> [email protected]
>> http://rubyforge.org/mailman/listinfo/wtr-development
>>
>
>
> _______________________________________________
> Wtr-development mailing list
> [email protected]
> http://rubyforge.org/mailman/listinfo/wtr-development
>



-- 
Bret Pettichord
Director, Watir Project, www.watir.com

Blog, www.testingwithvision.com
Twitter, www.twitter.com/bpettichord
_______________________________________________
Wtr-development mailing list
[email protected]
http://rubyforge.org/mailman/listinfo/wtr-development

Reply via email to