Can we use reverse image search to bypass image CAPTCHAs? has this been done before?

So has there been any group that managed to bypass the Image CAPTCHAs using reverse image search?

whats stopping people from writing a script that bypasses these CAPTCHAs by reverse image searching via many services like google and make a choice?

and can anyone provide me useful web services that do reverse image search other than google?

the only one i know is

and also can this be done? i can imagine writing a python script that cuts one of the 9 images in the captcha and then reverse searches it using a website, but the real challenge is how to make sense of the results and come to conclusion that whether this is for example a store front or not? any ideas?

A definition of “Done” in case of many Development Teams working on a single product

The open scrum test contains the following question:

When many Development Teams are working on a single product, what best describes the definition of “Done?”

A proper answer according to the test is:

All Development Teams must have a definition of “Done” that makes their combined work potentially releasable.

However, there is another option there, which is:

Each Development Team uses its own but must make their definition clear to all other teams so the differences are known.

And my question is that I don’t really understand what’s the significant difference between these two options? I deem them as very similar. They both complete each other. And I don’t really think that there is anything wrong in the second statement.

Could someone give me a glue how should be more confident about why the proposed correct answer is only one correct in this particular case?

Python: skipping values close to what we’ve already done

The main piece of code is this.

    thislist = []     for pt in zip(*loc[::-1]):         skip = False          for previous in thislist:             if abs(previous[0] - pt[0]) < 10 and abs(previous[1] - pt[1]) < 10:                 skip = True                 break         thislist.append(pt)         if not skip:             #do some stuff here! 

Basically I have a bunch of tuples in loc that I am looping over. Each tuple has 2 integers in it. I want to do some stuff with each (x,y) but only if it differs significantly from any previous such tuple.

So the way I do that is I create a list to which I add every tuple. Every time I get to a new tuple, I loop over this list and check if any of them are significantly close to the new tuple (I measure closeness by testing if it differs by less than 10 on each coordinate). If no tuple in the list are close enough, we can do the stuff we need to do.

Can this be sped up, possibly?