Comments on: Hiring Testers Who Can Code: An Exercise http://patrickwilsonwelsh.com/?p=44 Agile coding, agile testing, agile coaching, the agile enterprise, and Network Weaving. Sun, 29 Aug 2010 13:06:14 +0000 hourly 1 http://wordpress.org/?v=3.0.1 By: Dawn Cannan http://patrickwilsonwelsh.com/?p=44&cpage=1#comment-2133 Dawn Cannan Mon, 19 Oct 2009 20:54:39 +0000 http://patrickwilsonwelsh.com/?p=44#comment-2133 Ok, Patrick, I just sent out this project to my first 3 candidates. I'll let you know as I have feedback for you :) Ok, Patrick, I just sent out this project to my first 3 candidates. I’ll let you know as I have feedback for you :)

]]> By: Fredrik Wendt http://patrickwilsonwelsh.com/?p=44&cpage=1#comment-2042 Fredrik Wendt Mon, 21 Sep 2009 20:48:07 +0000 http://patrickwilsonwelsh.com/?p=44#comment-2042 This may be a bit off topic, but ... I wrote something similar to Selenium grid, but around Selenium Core (browsers were started and killed on computers, pointing at an HTML file that started a test run). That way we could use regular HTML files from the Selenium IDE as source for our (GUI) tests. As long as coders provide sensible id attribute values on the HTML tags, the recorded tests are even easy to follow for anyone new to Selenium or perhaps the web GUI and the product itself. I'm sad to see that Selenium still lacks an HTML driver. Most testers can work with HTML and CVS, but asking them to submit uncompiled, untested Java code wasn't really accepted. Going with HTML source files on the other hand was. This may be a bit off topic, but … I wrote something similar to Selenium grid, but around Selenium Core (browsers were started and killed on computers, pointing at an HTML file that started a test run). That way we could use regular HTML files from the Selenium IDE as source for our (GUI) tests.
As long as coders provide sensible id attribute values on the HTML tags, the recorded tests are even easy to follow for anyone new to Selenium or perhaps the web GUI and the product itself.
I’m sad to see that Selenium still lacks an HTML driver. Most testers can work with HTML and CVS, but asking them to submit uncompiled, untested Java code wasn’t really accepted. Going with HTML source files on the other hand was.

]]> By: Dawn http://patrickwilsonwelsh.com/?p=44&cpage=1#comment-1872 Dawn Sun, 23 Aug 2009 22:27:12 +0000 http://patrickwilsonwelsh.com/?p=44#comment-1872 I agree with Abby completely, and as I am currently interviewing for 2 testers in a company about to go through an agile transition, I have had a hard time finding people with the right balance. I have said to those who have asked me "do they have to be computer science students?" (I am hiring entry level), "no, they don't have to be, but I want them comfortable with reading code and able to learn to program pretty quickly." I was actually *just* a few weeks ago looking for *exactly* such a sample to give my selected interviewees. Glad to have met you, Patrick, and looking forward to meeting Abby as well. I will be going through this example more in the coming weeks and will be sure to give feedback. I agree with Abby completely, and as I am currently interviewing for 2 testers in a company about to go through an agile transition, I have had a hard time finding people with the right balance. I have said to those who have asked me “do they have to be computer science students?” (I am hiring entry level), “no, they don’t have to be, but I want them comfortable with reading code and able to learn to program pretty quickly.”

I was actually *just* a few weeks ago looking for *exactly* such a sample to give my selected interviewees.

Glad to have met you, Patrick, and looking forward to meeting Abby as well. I will be going through this example more in the coming weeks and will be sure to give feedback.

]]> By: abby, the hacker chick blog http://patrickwilsonwelsh.com/?p=44&cpage=1#comment-1498 abby, the hacker chick blog Thu, 23 Jul 2009 23:39:21 +0000 http://patrickwilsonwelsh.com/?p=44#comment-1498 I'm really curious to see the responses to this. In planning for our workshop on having developers & testers pairing, Nate and I keep coming back to the question "do the testers HAVE to be able to code?" Our current answer is "no, but... it'd be AWFUL nice." At the least they have to be able to understand code well enough to pair with a programmer to help guide the programmer towards writing more effective tests. But, how much more they could do (& I have to believe, how much more empowered they'd feel) if they could code as well. We don't want their main STRENGTH to be coding, of course - because the whole goal of pairing programmers with testers is that when we bring different perspectives to the table, we do better. So, if you have testers that are kick ass coders it might really not be any better than just pairing programmers together. But, enough coding to be able to automate the tests (NOT through point & click) would definitely be a Good Thing. I’m really curious to see the responses to this. In planning for our workshop on having developers & testers pairing, Nate and I keep coming back to the question “do the testers HAVE to be able to code?”

Our current answer is “no, but… it’d be AWFUL nice.” At the least they have to be able to understand code well enough to pair with a programmer to help guide the programmer towards writing more effective tests. But, how much more they could do (& I have to believe, how much more empowered they’d feel) if they could code as well.

We don’t want their main STRENGTH to be coding, of course – because the whole goal of pairing programmers with testers is that when we bring different perspectives to the table, we do better. So, if you have testers that are kick ass coders it might really not be any better than just pairing programmers together.

But, enough coding to be able to automate the tests (NOT through point & click) would definitely be a Good Thing.

]]>