Re: [15.3 Element] takesElementScreenshot question

Ask and you shall receive. I've just added ITakesScreenshot to
RemoteWebDriver in the open-source project's .NET bindings. I've used the
current spec's information on the command (GET
"session/{sessionId}/screenshot/{elementId}"). Using it with current driver
implementations will yield a NotImplementedException.

On Thu, Jun 25, 2015 at 1:43 PM, David Burns <dburns@mozilla.com> wrote:

> I think 1. should be done. The Java bindings, I am sure there is a good
> reason, is weird in that it forces people to do a cast or it uses an
> Augmentor to get what it needs. The c# bindings have similar things to keep
> parity.
>
> The client bindings should be able to try do anything that it wants and
> the server should just error if it can't do it. This is what the Python and
> Ruby clients do currently.
>
> David
>
> On 25 June 2015 at 18:32, John Jansen <John.Jansen@microsoft.com> wrote:
>
>> In section 15.3 [1] it talks about taking an element Screenshot.
>>
>> Currently (in the C# bindings ) the RemoteWebElement class does not
>> implement the ITakesScreenshot interface and hence we cannot take an
>> element screenshot. This means we have a choice to make:
>>
>> 1. Should we add the interface to the RemoteWebElement and then all other
>> C# browser bindings will benefit from it (assuming that the respective
>> drivers implement it)?
>> Or
>> 2. Should we implement this method in the our own class (which derives
>> directly from RemoteWebElement)? This would not affect the bindings for the
>> other browsers but  has some the disadvantage of forcing people who want to
>> use this functionality to write specific code for Edge. Generally tests
>> that are executed remote, use the RemoteWebElement class.
>>
>> What do you think?
>>
>> -John
>> [1] https://w3c.github.io/webdriver/webdriver-spec.html#element
>>
>>
>