posted 01-11-2002 09:59 AM         

I am due to be sitting in the COO office Monday a.m. The topic of discussion is the approval of a tools qoute for my teams use.

I have been trying to get tools for over 2 years now and this is the closest that I have ever been, the inside line states that it is all but a done deal. During this two year period I have done tons of reading/research into tools, their use, their pluses and minuses what are some of the perceived best practices.

What has this gotten me? An increased fear that I am destined to fail. Why do I feel as though I predestined to fail in this endeavor? Well mainly due to the fact that I, nor anyone else within my testing team, is a knowledgable programmer by any stretch of the imagination. And due to the fact that everything I read seems to have this "use at your own risk, unless you are a code writing guru" undertone to it.

(OK so maybe it was a really big nut shell)

My question is this. If any of you have entered into the dark world of automation without any automation experience on your staff, including yourself, did you survive the ordeal? And should I not expect to get anything out of automation if I don't hire someone with experience with the tools we purchase?

Any opinons on this would be greatly appreciated.

posted 01-11-2002 10:34 AM         

I think you need to make sure that everybody knows that automation is not the magic bullet. You need to make sure that everybody realizes that it will take training and time to implement. The return on automation is not an immediate thing. For somebody to buy you a tool today and expect 90% of all tests to be automated tomorrow is unreal.

As long as the cards are on the table and you and your company are dedicated to learning and implementing it, you should be fine.

posted 01-11-2002 10:44 AM         

If they're expecting it to be an overnight solution, it probably won't be.

There are some very good articles that should help. If you can get your hands on the Nov/Dec 2001 edition of STQE Magazine there is a very good feature about whether your shop is truly ready for automation.

Also, check the article "Cost-Benefit Analysis of Test Automation" on www.stickyminds.com

And, if you get your tools, check out Automated Software Testing by Elfriede Dustin, and/or two great articles of hers on www.stickyminds.com: Lessons in Test Automation and the Automated Testing Life Cycle Methodology.

Have you already evaluated and chosen a specific tool, or are you attempting now to get the permission to evaluate and choose one?

posted 01-11-2002 11:31 AM         

quote:

---

Originally posted by QAGirl:
I agree with skootz on this one..There IS a learning curve, but if your company is aware of that, then it's a part of the expected schedule...
If they're expecting it to be an overnight solution, it probably won't be...

---

quote:

---

...And, if you get your tools, check out Automated Software Testing by Elfriede Dustin, and/or two great articles of hers on www.stickyminds.com: Lessons in Test Automation and the Automated Testing Life Cycle Methodology...

---

quote:

---

Have you already evaluated and chosen a specific tool, or are you attempting now to get the permission to evaluate and choose one?

---

posted 01-11-2002 12:05 PM         

quote:

---

Originally posted by e-tester:
Yes we have chosen one, Rational Suite. I have evaluated so many tools for so long most come to me, or are sent to me, for all of their tools related questions, even if they are from a different group. As I stated earlier I feel as though I should not be answering the questions because I have no experience with the tools but then again I was showing the Rational Tech guy how to do somethings during the POC we did.

---

Ah, then try not to worry too much! If this is the case, you've at least shown the ability to pick up the things that the tool can do, and you're reading up on approaching things correctly. Nine times out of ten, if automation fails, it's because of false expectations and bad planning - it doesn't sound like either is the case for you.

Do keep us updated and let us know if the purchase is approved!!!

posted 01-11-2002 01:19 PM         

For additional info that you can present to upper management, the "White Papers" from the different vendors should be a good reference too! I know it helped me a lot (especially the topic that explains the Benefits from the tools are not seen during the 1st or 2nd pass...) while going thru what you're going thru now. Good luck!!!

posted 01-11-2002 03:54 PM         

quote:

---

Originally posted by e-tester:

...And should I not expect to get anything out of automation if I don't hire someone with experience with the tools we purchase?

...

---

Hi ER,

The good news is that you finally have this meeting Monday, the bad news is that expectations are often way too high for test automation - using capture/playback tools.
Make sure to manage the expectations. Since you have my book "Automated Software Testing," make sure to read section 2.1 “Overcoming False Expectations for Automated Testing.”
Also make sure to read section 2.3 “Acquiring Management Support,” which includes a discussion on tool training costs.
Rational provides for some good tool training courses - ask your manager to allow for training on the tool. I would even suggest taking an "Intro to Programming" course.

...and then continue reading the suggestions in the book "Automated Software Testing."

By the way, what is your application written in? What type of 3rd party control is your application-under-test using?

Let me know about your software engineering environment. I'd be curious.

posted 01-14-2002 10:25 AM         

Training is part of the quote and high on my list of must have in order to purchase the tools.

Sorry for this being so short but I have typed this three times now and none of the responses made it. Twice a browser crash and the last time my comments just didn't show (only the quoted message)

[This message has been edited by e-tester (edited 01-14-2002).]

posted 01-14-2002 11:45 AM         

quote:

---

Originally posted by e-tester:
BTW This morning I received preliminary approval for the purchase.

---

Congratulations!!

What's your next step going to be?

posted 01-14-2002 12:30 PM         

There are a few cursory tasks the COO wants completed and then it is time to generate the formal request paperwork for his sig. Then once that is approved it is off to training.

------------------
ER,
Sr. QA Lead

posted 01-15-2002 07:21 AM         

You've taken some big steps so far! Still got quite a road to take in this trip, but it looks like you are on your way.

Just remember to maintain your focus on the task at hand. Plan, prepare, disect, then attack. Its a wonderful world of automation that you are entering. But there be dragons, so watch out.

posted 01-15-2002 07:57 AM         

quote:

---

Originally posted by e-tester:
Prayer...

There are a few cursory tasks the COO wants completed and then it is time to generate the formal request paperwork for his sig. Then once that is approved it is off to training.

---

Good luck! We went through what you are about to do now 2 years ago (with Rational) and it was a big disaster.

The rational trainers where a joke, we paid thousands for training only to have some bloke with no idea show up and basically read from the manual word for word, so we ended up trying to go for it ourselves..big mistake!

In our case though, we were a fairly new company trying to set up the whole UML process from start to finish and it just proved too much. If you are only going to focus on getting Robot up & running then you will probably have more success then us because 2 years later, that is the only piece of Rational Software that we still use out of all the suites on offer!

posted 01-18-2002 07:22 AM         

It can be a crutch that people get used to.

You need to consider:
1. Vision. How do you tie in automation to reporting. What about defect tracking? What happens two years from now. (Looking for a new job?)
2. Individual contributor's roles. There are automation specialists, test script specialist, and test case specialists. Some people are good at breaking the stuff, while others are good at making some things continue to work properly. These are two separate, yet equally important roles.
3. Tools you choose. I don't recommend mercury or Silk. I've worked with both, but there are in the business of making money for themselves. Consider Perl, and Java. You can implement full automation strategy from development unit test to user cases with J-Unit, Http-Unit, and Bugzilla. The model is quite feasible.
4. Hire an architect. Forums are great, but hire someone who has wide knowledge of various tools. Don't hire a 'Silk' or 'Mercury' expert who pushes you to one particular language. Every product has it's speciality and you want someone who knows strenghs and weakness of each product.