Take the 2-minute tour ×
Software Quality Assurance & Testing Stack Exchange is a question and answer site for software quality control experts, automation engineers, and software testers. It's 100% free, no registration required.

I somehow stumbled across some references to the new ISO/IEC 29119 Software Testing standard that is currently draft stage (apparently).

I have not seen the details of this new standard, however I would expect that a truly modern testing standard should not just acknowledge, it should embrace agile development practices, so that it can be applied in such a way that you can be in compliance with the standard whilst using Agile methodologies like Scrum.

Has anyone "tested" the new standard and actually applied it to an Agile project? Will the new ISO/IEC 29119 Software Testing standard work with agile methodologies like Scrum?

Edit: Whilst James' point of view is interesting, it doesn't actually answer the question. I would like to hear from people who have actually used it.

share|improve this question
4  
Good question - it looks like the drafts aren't publically available, but you can request a copy by emailing your national standards body: softwaretestingstandard.org/gettinginvolved.php I'd be very interested to hear what the Agile community in general think of the standard - if it can be compliant, or made to be compliant, or if it can't - if it isn't compatible, it would be very useful to be able to say to senior management that "The general community consensus is that this standard isn't compatible with/relevant to our chosen development process." –  testerab May 22 '11 at 12:04
    
I've requested a copy of the draft in order to review it and I've been informed that my request has been passed to the committee responsible for the standard. They are unable to give a timescale as to when they will be able to get back to me unfortunately. However, the ppt available here softwaretestingstandard.org/downloads.php (you have to register) hasn't allayed very many of my own worries about the standard. –  sean_robbins May 26 '11 at 18:51
2  
The thing is in draft state, not easily available. and a lot of respected people in the test community are saying it's a waste of time. When you add that all up, I think the chances of finding anyone that has actually 'used' then standard are low. –  Chuck van der Linden May 30 '11 at 1:53
    
@sean_robbins You've got further than I have then. I requested a draft copy at the time of the original post, and have yet to even get a response. Agreed on the ppt not looking promising. –  testerab May 30 '11 at 2:02

5 Answers 5

up vote 17 down vote accepted

This "standard" is being created in the face of direct and repeated pleas by citizens of the Context-Driven testing community to cease and desist.

Whatever this thing says or is based on, it will not be based on a truly representative community of testers, nor will it be based on research that passes muster among competent social scientists.

It's yet another tool that will be used by European testing consulting companies to blind their clients and promote outdated, ignorant, and abusive testing practices.

We don't need a standard of software testing. We need communities of testers who study the craft and push it forward. Please join me in boycotting stupid standards.

share|improve this answer
1  
Agreed, as long as the project is successful the standards doesn't matter –  Aruna May 22 '11 at 23:19
2  
I guess that is a no then ... :-) –  Bruce McLeod May 23 '11 at 4:00
1  
What James Said. He is after all the father of all that is the life of a Pirate tester. Arrr! –  Hannibal May 23 '11 at 7:39
3  
The question was about how ISO/IEC 29119 relates to agile methodologies like Scrum. Mr. Bach didn't actually answer the question, so I'm downvoting. –  user246 May 23 '11 at 20:46
3  
Oh I'm sorry User246. Let me answer the question directly: no. ISO/IEC29119 relates to bullshit. Does that make it clearer for you? –  James Bach Jul 24 '11 at 10:22

As a former tester and now a QA Manager I have worked for over 15 years testing and managing the testing of very complex Geology and Geophysics interpretation software. My test team are all geoscientists first and testers second since you have to have an advanced degree to understand what the software is doing. I worked on a team using RUP, and we were ISO 2000 certified. Talk about the tail wagging the dog. As a team lead, at a major oil services company who produced GeoFrame and now Petrel, I can say the standard, the ISO2000 standards that is, drove everything. The organization was so focused and preoccupied with complying with the standard that more time was spend in worthless meetings, creating artifacts and jumping through the hoops dictated by the standard than doing constructive testing.

Now using Scrum Agile in a different company, the testing is much more streamlined and effective. We are putting together a testing standard for Agile, as part of our Definition of Done for user stories and iterations. We found that the QA team was geing forced to carry forward a lot of technical debt in the form of uncompleted workflow test cases, that could not be completed until all the functionality was in place from the last iteration. It is kind of Catch-22. Can't complete the test cases until all the coding is completed and can't close all the user stories until all the test cases are completed.

I am taking great pains that we do not create a testing standard that puts us in a straight jacket or becomes the tail that wags the dog.

share|improve this answer

I have excellent news for all of you - I do have access to the Jul 2011 draft and I can safely say that Agile testing is reflected in all of the case studies ("annexes") in 29119-3. Specifically, the contrast between two notional large organizations: "Agile Corp" and "Traditional Ltd" and their testing approaches. Here's something that should bring a smile to every Agile practitioner's face: an IEEE standard that features examples like the following:

Test report for: New subscription system (NSS)

Vers.: Sprint 3

Covers:

NSS final sprint result, including result of previous sprints, in preparation for a major customer delivery (for use).

Risks:

The live data risk was retired by creation of a simulated database using historic live data “cleaned” by the test team and customer.

Test Results:

Customer accepted this release of the product based on: 16 user stories with one added after last status report were successful. Technology facing testing with the one high risk story 100% statement coverage was achieved, and for the others 72% statement coverage was achieved on average. Team accepted the backlog on 4 defects of severity 3. Showcase was accepted by the customer with no added findings. Showcase demo sprint features interfaced with “live” data. Performance of the sprint features were found to be acceptable by team and customer.

New and changed risks:

Security of the system may become an issue on future sprints, assuming a follow work task is received from customer

Notes for future work from retrospective:

Sprint team feels a new member may be needed given possible new risk since no one has knowledge in this area. Severity 3 defects that move on to backlog should be addressed in next sprint to reduce technical debt The modified live data worked well and should be maintained. Test automation and exploratory testing is working, but added techniques should be considered, e.g. security and combinatorial testing.

Enjoy!

share|improve this answer
    
Erm - why would that bring a smile to my face? Also, what's the secret to getting hold of it? I never did get any response to my request via the official channel, so I never did get to see this supposedly open for comment draft. –  testerab Sep 5 '11 at 20:46
    
Also - where are my manners? Welcome to the site Andy :) –  testerab Sep 5 '11 at 21:04
    
Welcome to SQA! Could you please format the excerpt you pasted? Would be easier to read –  dzieciou Jun 26 '12 at 13:33
    
I have a copy too, and the words "agile" and "context" are repeated throughout the standard. –  Rsf Aug 1 '12 at 14:09

Why only then think about Agile being complaint with this piece of standard,why not context driven school of testing,why not TDD,why not scripted testing ?

I do not think a 'blanket' compliance is possible or even a good thing to do .

Reason being that each project,organization,business,technology has a different standard & nature of expectations from a testing exercise.

Having said that,there are also occasions when "standards" do get applied in some industries , but they are usually technology standards driven by business needs rather than process/activity standards.

example - I work in a company which manufactures HW and SW for public safety radios(police,fire safety,power,military) .Now,the technology is an open standard and the standards organization has mandated it that the technology should be an open standard to encourage interoperability between manufactures.Hence an important part of the testing activity is to execute a bunch of heavily controlled tests that pass a set of standard documents.However,this is done because all the manufactures have been consulted by the standards authority and taken on board before the standards are published.And the manufacturers accept it because they see this as a way to move their business forward and reduce monpoly.

However,seems like this particular set of test standard does not exactly represent the broad spectrum of testing experts or people from various 'schools' of testing. (Inferring from James Bach's comment)

share|improve this answer

I'm a newbie in software testing and would like to express my view on this.

Some developers/programmers/business people believe that software testing is a "low-end" IT job because they "think" everybody can do testing! They consider it as a "no-brainer" job.
Of course, we know it's wrong. Good QA/Testing requires lot of advanced skills and professionalism.

With this ISO standard, we can finally show them(people who believe everybody can do testing) that QA/Testing is NOT a low-end no brainer job. WE HAVE A STANDARD. WE ARE PROFESSIONAL!(Your boss will treat you as a professional!)

All I can say is we are living in a "dark" world run by Business people. In order to push the QA industry forward, "unfortunately" we need to have a standard to convince them.

share|improve this answer
    
Downvoting this because it doesn't answer the question. –  user246 May 23 '11 at 15:34
2  
I disagree. The problem isn't getting the respect of the business folk, it's finding good technical companies that recognize that software is a unique business that traditional business-folk just don't understand. The most skilled members of the testing community tend to shun standards like these, instead choosing to work at companies that respect their talents. These standards have been around for a while, and the best companies generally ignore them. –  Ethel Evans May 23 '11 at 16:44
1  
I agree with Mr. Y that we do need some sort of standard, but as @James points out below, the standard being created is being done by people who, themselves, are not testers and, therefore not qualified to come up with software testing standards. So, I'm down voting this answer as it is reactionary instead of an objective answer and it is not well researched –  TristaanOgre May 23 '11 at 18:40
2  
I actually have some sympathy with Mr.Y but hope one day he will realise that HE is responsible for convincing his boss that he is a professional and not the standard. –  sean_robbins May 25 '11 at 20:50
    
standards != respect (in many cases it can even be the reverse) –  Chuck van der Linden May 30 '11 at 1:50

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.