Sunday, December 30, 2012

Where do you stand in this debate?

Inspired by Elisabeth Hendrikson's blog post 

[Updated 25th Jan 2013]
I am disappointed to see no responses to this post. While I expected some responses in agreeing or disagreeing. Whenever I see such condition where my post does not get any comments - I think of following possibilities (thanks to Michael Bolton)

1. The post is not very engaging - there is way too much information there. Everyone and every thing is seeking attention. This post simply failed to get any
2. Its dumb idea - completely useless
3. Post is simply a question which either is too simple to answer (so no one would like to feel insulted by answering) or something deep and intriguing (why bother answering)
4. Why Author is not saying anything? Trick to get some free survey done for some homework?
5. No comments

I will attempt to expand on this topic sometime in the future. This situation made me to learn something - no comments - will make you think.

Dear readers - thanks for not commenting and teaching me something.


Sunday, November 04, 2012

A bizarre idea called "Software testing factory"

"Persistence in the face of a skeptical authority figure is priceless" - Seth Godin

Paul Holland (twitter handle @PaulHolland_TWN) shared this amazing video of Seth Godin on education systems. As I listened to Mr Godin talking about how present system of schools evolved from schools churning about labors for factories. Alas - even in our software testing "industry", we still need laborers as testers and companies take pride in setting up software testing factories. This post is about how bad and dangerous is the idea of "software testing factory"

According to Godin, about 100-150 years ago - schools used to be for a different purpose. He says - large-scale education was not developed to motivate kids or to create scholars. It was invented to churn out adults who worked well within the system. Scale was more important than quality, just as it was for most industrialists. A day in school started with "good morning" represented the notion of respect and obedience that was injected into students as a virtue. School was about teaching compliance, fitting in for the students into larger social context when pass out. Schools, according to Godin were established as public education to produce people who could work in factories - create set of people who can comply, fit-in and follow the orders of the supervisor.  

Emerging industrialization brought the focus on profitable factories - Godin points out. Factory owners thought "there aren't enough people, if we get more, we can pay them less - if we can pay less we can make more profits. When we put kids into factory that is called as school - we indoctrinate them into compliance.  Godin points out another key feature of factories - idea of interchangeable parts - when translated to schools - it meant producing people who are replaceable just as "standard part" of a machine.  When it comes to work - if you do more - there is always "ask" for little more. This is because - we are products of industrial age. The term productivity was brought the center of the things.

Key idea that I was attracted in this talk was about "Factory and how factory worked". I strongly believe that software and software testing work is "knowledge work" in contrast to "factory work". Here, thinking humans, in collaboration with humans assisted by computers create stuff that we call software that has changed and continues to change our life.  Wholesale lifting of idea of factory - thanks to strong association of "quality" to likes of toyota and promotion of idea of  "sick-sigma" (Cem Kaner used this phrase first, I think) - we have indoctrinated software people as factory workers.
I am troubled by this. When I ask people  - "does it what we deal in factory - machines/concrete things vs abstract ideas and machine instructions - matter? Should or is software produced like a machine in assembly line",  I get no clear response. Many simply think since our industry (Software) is immature and nascent - we must learn from engineering disciplines like manufacturing.

I am fine with learning from other disciplines as  I believe software testing is multi disciplinary - we constantly import ideas from multiple fields such as natural sciences, maths and statistics, behavioral economics, neuro sciences, cognitive psychology, philosophy, epistemology and list continues. I am against  wholesale and mind less import of ideas from the areas where we deal with a totally different type of things and we must exercise caution.

Coming back to factory - many IT services companies take pride in saying "we have successfully implemented software testing factory for a client" or "software testing is now commoditized" - what a shame !!! What happens in a software testing factory? There are dozens of "brain dead" people called software test engineers whose job is to produce test cases, bugs, test results, automation code (sorry popular word is "script"), metrics and tones of reports.  The intellectual pursuit of software testing that seeks to discover, investigate and report interesting and strange problems in software that requires - thinking, skeptic and open mind - has been reduced to "mindless" factory work. As a passionate tester, I would never want to associated with this deadly idea.

Am I biased as tester about my profession as some highly complex rocket science? Is my rational mind blocked or misdirected by confirmation bias? I think that is possible. If I am thinking about software testing as a business - like any other business say hotel, garments, manufacturing or engineering hardware - I would love the idea of factories. I would want to maximize my profits per dollar of investment. I would want to train cheap labour - teach them how to write test cases, report bugs and automate test scripts. I would then deploy them in "mass" to a client and charge handsome money in the name of testing. This business apparently works and it is perfectly legal, by and large ethical.

If I imagine myself as a tester in such factories (flip my context from factor owner to a factor worker or a supervisor) - I see a dark future for myself.  Just as factory works are expected to "comply" and follow a set pattern of work - when factor owner does not need me - I don't have any skills that I can trade outside factory. Over a period of brain dead work - I have lost my thinking and questioning mind. Unless I gain skills in becoming factory owner myself (that is a business development and management skill) - I must leave the factory quickly and move to an environment where I can grow my skills as tester  as a thinking individual.

In short - if you are managing software testing as a business - software testing factory is good for you. If you are a software tester working in a software factory - get out of the place fast or change the career to become factory owner or supervisor.

As a  tester in me roares - I wish for "End of compliance as an outcome - it is too boring for a curious, skeptic mind to simply fall in line".

Additional Notes: 
Following are few statements - that I liked that strikes chord with my belief in "software (testing)" as a knowledge work as opposed to factory work
  • Why we would not want to have our kids to figure it out and go do something interesting
  • Are we asking our kids to "connect dots" or "collect  dots"
  • We are good at measuring how many dots we collect - how many boxes are collected, how many facts memorized, 
  • we don't not teach kids how to connect the dots. You cannot teach connecting dots in dummies guide, text books. By putting kids in to situations where they can fail, experiment
  • Grades are an illusion - passion and insight are realities
  • Your work is more imp than your answer in congruence to answer key
  • "Fitting in" is a short term strategy to go no where.

Do not forget to read this pdf "stop stealing dreams" by Seth Godin.

Sunday, October 21, 2012

Divisions in Testing, Slotting People - How bad is idea of schools?

This post is an offshoot of discussion with friends Rahul Verma and Vipul Kochar on twitter. It started off from a blog post from Rahul on "exploratory testing" - one approach to testing that many in context driven testing community are working hard to be be good at.  When Vipul joined the debate - to me, two key things stood out as following "long" tweet of viper suggests -

" ...classification, definitions are good. When one starts to use them to divide and slot people, it becomes counter-productive."

Vipul followed up with a detailed post here

Divisions amongst people

Take for example - the idea of schools in software testing by Bret Pettichord.

Rahul wrote a good summary and analysis of schools of testing way back in 2007.  Rahul's main complaint was schools concept divides people.  My view is different. To me, idea of schools has been very helpful to identify myself and my approach to testing distinct from others I see around. It helped me to develop my skills in the framework of context driven testing school. I think, testing as a multidisciplinary field was (will always be) divided. It is just that few refused to recognize the differences. Still worse, some insisted that their's is some sort of universally agreed way of doing testing.

What Bret did is phenomenal but at the core he simply named the groups/schools that he saw. In other words - schools of testing idea did not divide people - it gave "names" to different sets of practices "using" the name of testing. Having names to things arounds us helps to talk about the things, debate about them, understand them and improve them. That is exactly what Bret's idea of schools - did to some of us.

If you disagree with idea of schools - you might be saying one of these

"There is one universal way of doing testing hence idea of schools is absurd"
"I do not agree with Bret's classification - here is mine"
"I refuse the idea that there are patterns in testing that are distinct"

So - it would be not be correct to put the blame on idea of schools in testing to "division" in our industry - divisions always existed, we now have one model in which these differences can be named. I also argued with Rahul that "divisions" are good for our craft - they work like having multiple political parties in a democratic setup. With divisions we can have multiple, diverse ideas to co-exist. I am in favor of division in testing community as we need diverse mindsets, ideas and philosophies each offering solution to unique situations.

Vipul's post on "religions" and his apparent suggestion on being like "water" - indeed is a support of view of "divisions" are good.  If there are differences and divisions - cherish in diversity instead of trying to bring unification.

Slotting people, calling people by names

As strong supporter of schools concept - what I condemn is slotting people where they don't want to belong or identify with. There are factory or analytical school practices not factory or analytical school testers. Likewise there is Agile Testing (some form of testing that happens in Agile projects) but there are no Agile testers. There is exploratory testing and testers can chose to be good at it - but when they master it - they don't become exploratory testers - but testers with mastery over the approach of exploratory testing.

When people get slotted in groups/labels (for example if we call someone as factory tester) - for few it sounds "offensive". Personally, I am proud to be context driven tester. I don't have problems of me getting slotted in a category that Bret proposed. But that is me only speaking. By speaking of me as a context driven tester - I will let others know my testing philosophy and to some extent help others what to expect from me. This label for me is helpful to identify my approach and grow it in a framework driven by the principles of my school. 

Vipul approaches this from a different direction - he talks about dangers and obsession of belonging to a school (akin to type of fundamentalism that we see in religion). He says "Test matters and the test result matters not the division" - Well - I say - how does one test? what principles and values one approach the act of testing? The values, beliefs and approaches that one uses in testing define what Bret called as school. These elements of school are not independent and separate parts of a testers life and work. When we become conscious of them - we can work to improve them,  add few, modify few and delete few. How can one chase objectives and goals of testing without having a value system of individual about testing? If you think young testers struggle to define terms like of GUI testing or agile testing etc or struggle to belong or not belong to any school - it is sign of they trying to find their value system.

While a person can be FREE thinking person to choose and adapt - I can always see in the person - a subtle value and belief system about world, work (testing) - a view. Even choices of Free thinker are subtly guided by these values and beliefs. Instead of trying to deny the existence of these values and beliefs (in involuntary pretext of freedom to chose and adapt) - I urge likes of Vipul and Rahul to explore to find these subtle values that drive them. Bret's idea of schools and influences of James Bach, Cem Kaner and Michael Bolton - personally helped me to find my values or to be precise - they shaped up my fluidic and rather vaguely defined testing philosophies, values and beliefs.

I am proud to stand up as a context driven tester - I can talk about my values and beliefs about testing. While I do this - one thing that these great teachers (James, Cem, Michael) taught me is - not to get biased by one unilateral thinking. I constantly question my beliefs and values - I try to hang around with people who think and work differently than me. I train to be critical and rational thinker - constantly look to beat "confirmation bias".

I am reminded of this famous quote of Bertand Russel "Do not absolutely be certain of anything" - So…as a tester - I keep doubting my own ideas and that of others - that keeps me learning.


Friday, August 24, 2012

How different Software Industry segments see Testing ...

Consider these views expressed by few real people about testing - cutting across the software industry segments. You (a tester) might be surprised by few of these comments - but take it from me - these reflect true state of how stakeholders see testing as.

A manager from a Software Product Company : "We follow Agile model - every team member in the team is responsible for quality and will do a bit about testing. We believe in Agile practices like test driven development, continuous integration, automated unit testing - our code is naturally comes out with good quality. We do not employ any "plain vanilla" black box testers. That is waste of our time. We would get all our testing done by developers mostly or in some cases - testers cover the rest through automated testing. We dont have anything called "testing" phase in our process. We hire testers that are capable of writing production level code - as most of their time will be spent in writing unit tests and automation to help developers.

A manager from IT/Captive Unit : "We believe in providing agility and value to our customers. Testing is one small bit in that whole process. We don't actually worry about how testing is done as long as it aligns to our business purpose. Bulk of testing that happens is done by our partners. We constantly seek to commoditize testing and aggressively deskill so that - we can gain the cost efficiencies in testing. More than testing skills - we value business domain skills. Testers eventually either become managers (and manage customers, IT services deliver/management and other stakeholders) or become business analysts.

A manager/consultant from IT services Industry: Testing is all about assuring quality and process improvement. We constantly develop tools and frameworks to help our customers to do testing efficiently and cheaply. We provide value driven testing services based our process maturity and experience in setting up large scale test factories. Our number 1 aim  is to reduce cost of quality - we do it by focussing in tools, processes and domain skills.

A consultant from Software Tools Company: Testing is an essential part of SDLC that can gain significantly from Tools - Automation tools. Usage of Automation aggressively can help reduce cost of testing. Software Testing tools help in implementing Software Test factory so that non technical and business users can use them and achieve faster cycle time and enhanced quality. Not to forget our strength in terms of Six Sigma, CMMi and other Software Quality models. We endorse software quality management through rigorous metrics and quantitative measures.

Now - dear tester - identify yourself where are you working and how are you improving skills in testing to suite the industry segment you work now or hope to work in the future. Does this sound similar to the view of testing that you read in text books or conferences ? Did you know software industry sees testing in such variety of perspectives?


Sunday, May 06, 2012

A brief introduction of Test Automation...

I was asked by a blog reader to give a quick introduction of how automation helps in testing. Here is how I replied. I thought this might kick off some interesting off shoots...

"Certain portions of testing such data validation etc can be efficiently verified by automation programs than humans in repeated way (humans make mistakes and often are terrible at repeated executions). By carefully identifying portions of application under test that could be "safely" checked (validated) by automation - you can speed up testing (you can run many test cases in parallel, in the night etc) through automation. 

But beware - automation is a dumb and (humble?) servant - will do exactly what you ask it to do million times without cribbing - it does not have intelligence. A good tester can recognize something that is not in test script and looks like a problem. Automation cannot do this."

Do you like it?

One offshoot I am reminded of when wrote this piece - Automation is like people trying to losing weight. It requires patience, discipline and dedication. There are many quacks that operate in both automation and "weight loss" industry that promise "over-night" benefits.

If you are aware of how weight loss works or does not work - you can safely extend the analogy to benefits of automation.

Do not expect your testing or your application to become slim and trim with automation - overnight and most importantly - do not expect it remain so with no investment on ongoing basis. The later part - neither automation consultants (especially those who sell tools) nor those folks that run weight-loss industry - will tell you.


Sunday, April 01, 2012

Testing is Dead - in which world?

Few weeks back I participated in VodQA event by Thoughworks. It was a day filled with lots of power packed sessions and discussions around the topic testing - sorry QA (that is how TW calls testing as).

My talk was around the topic about alleged death of testing and implications of the same on people who live and make their living by testing.

The slides of the talk are here and video of the talk is on youtube (thanks TW)

Folks organizing the event did a wonderful job of arranging a platform and have people exchange their views on testing. There was this "open house" where an adhoc group of people assembled at a place to discuss about a topic that one of them wanted talk about. There was passion and energy all around. I said to myself - in such an assembly of about 50-70 people - who could believe "testing is dead". It was  a real thing for the people in the event.

One thing that I wanted the listeners of the talk - was this idea of "two worlds of interpretation" - software makers world and software users world.  More about that later in a separate post.

Saturday, March 24, 2012

Learning from Tenali Raman's crows ...

As kids, like many in southern part of India - I grew up listening to stories of Tenali Raman - a 16th century wise court-poet of King Krishnadevaray of vijaynagar empire. Tenali Raman is also known as Vikat Kavi - meaning intelligent poet. Birbal from King Akbar's court - enjoys similar cult among kid's stories in India. This story of counting crows that I narrated to my 8 year old daughter - made me realize how real are Tenali raman's crows in our day-today life in software.

First, let me quickly run through the story. One day king throws up a strange puzzle to Tenali - asking him to count and report the number of crows in the city. Tenali thinks for a while and asks for 2 days time to come up with the answer. After two days, he comes back and reports to king that that there are One lach (10 lach = 1 million) seventy thousand and thirty three crows in the city. At first, the king becomes frozen and did not know how to respond - after a while, recovering from the shock of the answer - king checks if  Tenali is sure about the answer. King further says that he would conduct a counting (recounting?) and if number does not agree with Tenali's number - he (Tenali) would punished. Tenali being Tenali - responds to qualify his answer. He says it is possible that the recounted number of crows might be different from his number. If the new number is less than old number - then it is due to the fact that few of city's crows have gone out of station (city) to visit their relatives in nearby cities. If the new number is more than the old number, then additional number is due to crows from nearby cities visiting their relatives in vijaynagar city. Listening to this - king has a heart-full laugh and realizes the flaw in assignment/problem. As it happens in all Tenali stories - Tenali gets king's praise and some prizes for the witty answer.

Now, let us come back and see how this crow metaphor is applicable to what we do as project managers, test managers and testers in our day today work.

There are entities we deal that are similar to crows - in following respects :

1. Counting/quantifying is a prized puzzle
2. Counting is asked by an authority, a boss - you cannot say "No" to ( saying "no" can cost you your job or potential label of "incompetent")
3. Often you can fake a number
4. There is no easy, sure way to verify/validate the count
5. Even if someone does a recount and comes up with new (different) count - you can always "explain" the discrepancy, like Tenali did.

One example that comes to my mind is count of test cases. Typically, during test estimation process, as a test manager you would be asked "how many test cases could be written for a given set of requirements". The boss would then do the required math to confirm the number of testers required, time required to execute the estimated number of test cases (note - time required to "execute" test cases - not to test). So, wear hat of Tenali - throw up a number. If asked - show your working (be sure to have your working).  You would be OK from then on.

There are things we deal in software that can not be counted like we count concrete things.  Software requirements, use cases, test cases, lines of code, bugs, ROI from Automation - are abstracts not concrete objects. Counting them is akin to counting crows as in Tenali's story.

[Puzzle : Prove that ROI from automation is a Tenali Raman Crow count]

Cem Kaner says executives are entitled and empowered to chose their metrics. So, King was perfectly right in asking Tenali to count and report number of crows - though objective of King in the story is not to make any important decision for his kingdom. In any case - crow count metric was sought.

What can a tester/test manager do when asked to count "crows" ? While our community develops models and better alternatives to "imperfect metrics" - we need to tread a careful path. We should provide alternatives and possible answers to crow counts.

I have come to realize that refusal to give the count might be counter productive in many cases - trying to ape Tenali Raman might useful. Need for quantification is here to stay - years of persuasion and reasoning why in some cases counting can be bad - has not managed contain the problem.

What do you think about "Pass/Fail Counts"?


Wednesday, March 21, 2012

My Views on Testing certification : 2012

A reader of my blog "Arpan Sharma" writes "What’s your take on certifications these days? I see your wrote about this is 2008 which is almost 4 years ago. Do you think the landscape of certifications have changed in recent times?".

Arpan - Thanks for writing and reminding that my stand on certification on this blog is about 4 years old now. It is interesting that you are checking with me if I have changed views. Here is how I summarize my current thinking on certification.

1. First of all the person seeking certification should be absolutely clear what they are expecting the certification to give them - Knowledge, Skill, skill enhancement, Marketing value, a job, an interview

2. Certifications that do not observe and qualitatively grade a tester - in action "while doing testing" - can not guarantee a certain level of skill in testing. Employers, Recruiters, hiring managers - please take a note.

3. If you want to learn how to do good testing, how to gain skills in broad testing landscape - certification is not something you should look for.

4. If there is a certification that let us get a job in a given situation/context or gets you a interview shortlisting - you should consider taking that certification. But - be aware - once you get your job - you are on your own. You would then be required to display (depending on the type of organization and nature of job) skills on job. Certifications' role ceases there.

5. Be critical about certification material and tests tell you - question them. Form your own ideas and logic about how things work. Do not take everything that taught or you read as part of certification as "universal truth. Why this is important? Only being critical on what is certification course - can help you to decide what value intrinsically you gained from it and what already existed in you.

6. Reputation is everything in today's world. You gain professional reputation by demonstrating your work and skills to your employer and to out side world (through networking world). Building reputation takes time and real good work. People with confidence in their skills and reputation - do not require a third party to endorse their level of skill. In today's world - people with skill and reputation - don't need certification. What does that tell you about certification?

7. Take special note of qualifiers like "Advanced" when applied to certifications - check out what is advanced and how? More often that not - it is more "jargon-laden".

#4 and #5 specially apply to freshers looking for /some/ job and those 1-3 years experience folks who either had some software job or a lost a testing job.

In terms of landscape of certifications - I don't think there has been change. Prime motive for certification providers is to make money - fast and cheaper. That has only intensified with many job seekers. That is fine as a business objective - we the target audience of such business ventures need to be clear about what we want from certifications and how capable are these certifications to deliver on the promises.

I repeat what I said earlier - if you want to learn, acquire skills, enhance skills in testing - certifications are the things that you should avoid. There are better, cheaper ways of doing that.

Did I answer your question Arpan?


Thursday, March 01, 2012

Patterns in weakness in approaches about testing

I was reading this testing round table discussion and thought this might a blog post. Here I go...

To me, the biggest weakness is the perception or idea about what testing is  and why it is required.

Here are few examples as how companies treat testing.

1. Something that is avoidable to large extent or even can be eliminated if they could get their programmers and analysts get spec and code exactly right. The lousy work that these folks do during SDLC - creates need for testing.

2. Quality Assurance - Straight out of the box comparison with manufacturing assembly line. For these folks, testing is all about process and nothing else. If you get process right - you are done. It does not make any difference who does and when - all they need is to get process script right.

3. Building quality from the grounds up - A variation of #1 above - a growing group of people think that if you have automated tests (checks - actually) you don't have to really worry about testing. You are building quality from the ground - you cannot test quality but need to build it right? poor testers will not define and manage testing (under the name QA)

4. Testing? What? The whole team approach. This is creation of Agile model. Here, testing is everyone's responsibility. There goes "testing" out of the door as a specialist's job. When testing is something anyone in the team does - it is like any other project task.

5. Dont forget - this popular rhetoric - Testing (phase) is dead - That is biggest weakness in approaches about testing.  What else can be the biggest weakness about something other than saying it is dead?

Whole idea of testing is dead is around these beliefs and notions (test yourself if you agree or not)

1.  Testing (phase or role) makes developers complacent - a safety-net - remove it to make developers responsible.
2.  with so much focus on automated unit testing, test driven development, continuous integration - developers are producing quality software anyway
3. Finding problems is no big deal, we know where problems are (this is what James Whittaker said in his EURO STAR 2011 keynote). So do we need what testers for?
4. With cloud as popular software delivery model - you don't bother about bugs leaking. Time and effort to fix and turn around bug is ridiculously LOW - why bother testing?
5. What for you have crowd sourcing? Beta testing? - throw your stuff to users - let them use and tell us where are the bugs (there should not be many as we are group of smart developers and we know where the bugs are)

Thus,  the weakness in testing arises out of how we think about it  and what we want it to do for us. Thinking idealistically about how software is made and used, applying models from other fields without properly customizing them and removing or de-emphasizing the human element in the system  - are the key patterns of weakness in testing approaches

What do you people think?