Matthew Heusser explores ideas in software development and knowledge work using critical thinking - mostly with an agile and testing slant
Updated: 10 hours ago
Sheep of a different fold
For a few years now I've been listening and reading to the work Mary Poppendeick has produced with increasing appreciation. Then last year I had the opportunity to interview Mary and Tom for InformIT, as part of InformIT's coverage of the Agile 2009, where Mary was giving a keynote speech.
(Mary and Tom) and I have very different backgrounds. We worked in different kinds of organizations and our careers and interests took us in very different directions.
Yet here was this other person that had both studied the history of the philosophy of management -- and studied the actual effects of those ideas in practice -- and come to the same conclusions as I had.
For that matter, the way that the Poppendeick's approach the subject is different than my stuff, and I think worth studying. So when I found out that they had a google tech talk, I just had to link to it here:
I've spent tens, if not hundreds of thousands of words trying to explain the ideology of "process improvement" and some of my concerns about it. For a quick summary introduction, I gotta say, this video by Mary is shockingly good. Throw in a copy of The Management Myth by Matthew Stewart and you end up with a very good survey of the literature in two source documents.
sigh.
If you need me, I'll be in the corner, licking my wounded pride, trying hard not to cry.
:-)
Seriously, this is a good stuff, and I am pleased to recommend it.
UPDATE: A cursory glance at Poppendeick LLC website finds several 'sound byte' level things that you might take issue with in regards to /testing/. My advice: Ignore the sound bytes that are so easy to misconstrue; watch the video instead. Check out what she actually /says/ about software development, management, and leadership. I expect it will resonate with you. It did with me.
(Mary and Tom) and I have very different backgrounds. We worked in different kinds of organizations and our careers and interests took us in very different directions.
Yet here was this other person that had both studied the history of the philosophy of management -- and studied the actual effects of those ideas in practice -- and come to the same conclusions as I had.
For that matter, the way that the Poppendeick's approach the subject is different than my stuff, and I think worth studying. So when I found out that they had a google tech talk, I just had to link to it here:
I've spent tens, if not hundreds of thousands of words trying to explain the ideology of "process improvement" and some of my concerns about it. For a quick summary introduction, I gotta say, this video by Mary is shockingly good. Throw in a copy of The Management Myth by Matthew Stewart and you end up with a very good survey of the literature in two source documents.
sigh.
If you need me, I'll be in the corner, licking my wounded pride, trying hard not to cry.
:-)
Seriously, this is a good stuff, and I am pleased to recommend it.
UPDATE: A cursory glance at Poppendeick LLC website finds several 'sound byte' level things that you might take issue with in regards to /testing/. My advice: Ignore the sound bytes that are so easy to misconstrue; watch the video instead. Check out what she actually /says/ about software development, management, and leadership. I expect it will resonate with you. It did with me.
Categories: Blogs
What Matt has been up to
Woa. Some cobwebs on Creative Chaos, eh? Seems like Matt Hasn't been very busy, doesn't it?
Gosh, I sure hope you don't think that. Let me bring you up to speed:
a) I've split up my blogging into two places - here (always free, no-registration required) and for the Software Test Performance Collaborative - still free, registration required.
b) That blogging now includes recruiting, running and managing a weekly podcast "This Week in Software Testing" - also free for the new shows. To get the old content, you'll need to be a paid member of SoftwareTestProfessionals.Com. Don't want to pay? You can download every episode as they come out, or you can participate in various contests on the blogs. Or write an article for the magazine.
c) I'm still micro-blogging on twitter under the username mheusser.
d) I'm still producing a column for the magazine, now known as "Software Test & Quality Assurance" magazine, or STQA. After two years of writing a encyclopedia-style column were we defined key terms used in the "theme" of the issue, we decided to shake things up a bit and write an interview column. Each issue we'll gather questions from the community for a 'test expert' and have them answer. For the August issue we interviewed Michael Bolton; the extended interview is now available on-line. I'm at the point now where I need recruits to interview, and shortly after that, I'll need questions ...
e) Beyond working with the fine folks at STP, occasionally I get a bit of time to work with other publishers, including the folks at SearchSoftwareQuality. That includes a few podcast-style interviews, a tutorial on the Selenium IDE, a good place for classic testers to learn about Selenium, as well as a two-part tutorial on Selenium RC, which is a start for programmer-types. (Link to Part I and Part II here). I also just wrote a small piece on Effective Bug Reporting Techniques for SSQ.
f) We finally put the Conference for the Association for Software Testing (CAST 2010) to bed last night with a conference retrospective. We held it in Grand Rapids, Michigan in early August. Instead of presenting, I helped out with the local logistics, recruiting some sponsors, and organizing and funding the evening receptions. CAST 2011 will be in Seattle, Washington. With Jon Bach as the conference chair and James as the program chair, I suspect it will be amazing.
g) With one conference to bed, it's time for me to worry about the next one! STPCon is going to be October 19-21 in Last Vegas, Nevada. It starts the 17th if you grab a two-day pre-conference tutorial. I'm a "track chair" for the hands-on track sessions, I'm running one of the track sessions, running a panel on how to decrease costs in testing, and organizing a lightning-talk like session. Oh, and there will likely be a Monday night reception.
h) Day job! Full time as a member of the technical staff specializing in test for Socialtext. At least I managed to skip the commute, otherwise this stuff would be impossible.
i) It's about time for me to re-start teaching religious education for fourth and fifth grade at my Church during the school year, plus coaching soccer for this fall. (See, I have a life outside of work. Really. Occasionally. Sorta.)
j) I just wrapped up a two-year night teaching position at Calvin College. It was really great, but due to a-i, plus not commuting into Grand Rapids anymore, something had to give. I have small children at home; it would be nice to occasionally see them.
... and then I went crazy.
No, at least semi-seriousy. Based on a discussion on the LinkedIn Discussion list, I just signed up to be the lead editor on a collection of essays on how to reduce he cost of software testing to be published by CRC Press in early 2011.
We've got a good team. We had some solid progress before we signed the contract.
But I did just sign and email the contract last week, and our completed, publisher-ready draft is due Nov 1st.
More to come; at the very least, I'll try to blog pointers to interesting work elsewhere.
But forgive me if I haven't been blogging here much. As I hope you can see, I've been ... kinda busy.
And that was before I went crazy. :-)
Gosh, I sure hope you don't think that. Let me bring you up to speed:
a) I've split up my blogging into two places - here (always free, no-registration required) and for the Software Test Performance Collaborative - still free, registration required.
b) That blogging now includes recruiting, running and managing a weekly podcast "This Week in Software Testing" - also free for the new shows. To get the old content, you'll need to be a paid member of SoftwareTestProfessionals.Com. Don't want to pay? You can download every episode as they come out, or you can participate in various contests on the blogs. Or write an article for the magazine.
c) I'm still micro-blogging on twitter under the username mheusser.
d) I'm still producing a column for the magazine, now known as "Software Test & Quality Assurance" magazine, or STQA. After two years of writing a encyclopedia-style column were we defined key terms used in the "theme" of the issue, we decided to shake things up a bit and write an interview column. Each issue we'll gather questions from the community for a 'test expert' and have them answer. For the August issue we interviewed Michael Bolton; the extended interview is now available on-line. I'm at the point now where I need recruits to interview, and shortly after that, I'll need questions ...
e) Beyond working with the fine folks at STP, occasionally I get a bit of time to work with other publishers, including the folks at SearchSoftwareQuality. That includes a few podcast-style interviews, a tutorial on the Selenium IDE, a good place for classic testers to learn about Selenium, as well as a two-part tutorial on Selenium RC, which is a start for programmer-types. (Link to Part I and Part II here). I also just wrote a small piece on Effective Bug Reporting Techniques for SSQ.
f) We finally put the Conference for the Association for Software Testing (CAST 2010) to bed last night with a conference retrospective. We held it in Grand Rapids, Michigan in early August. Instead of presenting, I helped out with the local logistics, recruiting some sponsors, and organizing and funding the evening receptions. CAST 2011 will be in Seattle, Washington. With Jon Bach as the conference chair and James as the program chair, I suspect it will be amazing.
g) With one conference to bed, it's time for me to worry about the next one! STPCon is going to be October 19-21 in Last Vegas, Nevada. It starts the 17th if you grab a two-day pre-conference tutorial. I'm a "track chair" for the hands-on track sessions, I'm running one of the track sessions, running a panel on how to decrease costs in testing, and organizing a lightning-talk like session. Oh, and there will likely be a Monday night reception.
h) Day job! Full time as a member of the technical staff specializing in test for Socialtext. At least I managed to skip the commute, otherwise this stuff would be impossible.
i) It's about time for me to re-start teaching religious education for fourth and fifth grade at my Church during the school year, plus coaching soccer for this fall. (See, I have a life outside of work. Really. Occasionally. Sorta.)
j) I just wrapped up a two-year night teaching position at Calvin College. It was really great, but due to a-i, plus not commuting into Grand Rapids anymore, something had to give. I have small children at home; it would be nice to occasionally see them.
... and then I went crazy.
No, at least semi-seriousy. Based on a discussion on the LinkedIn Discussion list, I just signed up to be the lead editor on a collection of essays on how to reduce he cost of software testing to be published by CRC Press in early 2011.
We've got a good team. We had some solid progress before we signed the contract.
But I did just sign and email the contract last week, and our completed, publisher-ready draft is due Nov 1st.
More to come; at the very least, I'll try to blog pointers to interesting work elsewhere.
But forgive me if I haven't been blogging here much. As I hope you can see, I've been ... kinda busy.
And that was before I went crazy. :-)
Categories: Blogs
Kaner On Testing
Those of you who read this blog know I've probably spent tens, if not hundreds of thousands of words discussing the applicability of hard metrics to the management of software development.
You likely know that I'm not keen on it.
Yet I struggle to make the point sharply and quickly. Cem Kaner wrote something on metrics today that summed it all up in a hundred words or so:
Capers Jones sometimes talks disparagingly about the (claimed) fact that 95% of American software companies have no metrics program. On the surface, this sounds terrible. But what I saw as a consultant was that most software companies have tried a measurement program, or have executives with lots of experience with metrics programs in other companies. The problem is that their experiences were bad. The measurement programs failed. Robert Austin wrote a terrific book, Measuring & Managing Performance in Organizations. When you start measuring something that people do, people will change their behavior to make their scores better. People will change what they do to get better scores on the measurements—that’s what they’re supposed to do. But they don’t necessarily change in ways that improve what you want to improve. Often, the changes make things worse instead of better (a problem commonly called “measurement dysfunction.”) This problem happens more often, and worse, if you use weak, unvalidated metrics. I keep meeting software consultants, especially software process consultants, who say that it’s better to use bad measurements than no measurement at all. I think that’s’ a prescription for disaster, and that it’s no wonder that so many software executives refuse to harm their businesses in this way.
I thought it was brilliant.
If you want more, you can read the source that quote comes from - part I of a series of interviews with Cem on uTest.com.
Or come to the Conference for the Association for Software Testing - CAST 2010 - next week in Grand Rapids, Michigan, where Cem is giving a keynote-level speech.
UPDATE: I've been thinking about it, and it certainly depends how you define your metrics. For example, one kind of measurement that I am in favor of is the "slip chart." In other words, you look at every deadline the team has committed to and how late they are, and you figure the rough percentage of how late they /always/ are. With that, you can predict when you will really be done. The folks in the extreme programming community codify this into story points and burnup charts, and that's fine. I don't have a problem with these used as approximations; as first-order measurements. The problem comes when they are reduced to 100-word silver-bullets without the context of how they can be used well.
So I wouldn't say I'm totally opposed to metrics are part of a balanced breakfast. I'm just leery of the common ideology of measurement by numbers alone, without context.
You likely know that I'm not keen on it.
Yet I struggle to make the point sharply and quickly. Cem Kaner wrote something on metrics today that summed it all up in a hundred words or so:
Capers Jones sometimes talks disparagingly about the (claimed) fact that 95% of American software companies have no metrics program. On the surface, this sounds terrible. But what I saw as a consultant was that most software companies have tried a measurement program, or have executives with lots of experience with metrics programs in other companies. The problem is that their experiences were bad. The measurement programs failed. Robert Austin wrote a terrific book, Measuring & Managing Performance in Organizations. When you start measuring something that people do, people will change their behavior to make their scores better. People will change what they do to get better scores on the measurements—that’s what they’re supposed to do. But they don’t necessarily change in ways that improve what you want to improve. Often, the changes make things worse instead of better (a problem commonly called “measurement dysfunction.”) This problem happens more often, and worse, if you use weak, unvalidated metrics. I keep meeting software consultants, especially software process consultants, who say that it’s better to use bad measurements than no measurement at all. I think that’s’ a prescription for disaster, and that it’s no wonder that so many software executives refuse to harm their businesses in this way.
I thought it was brilliant.
If you want more, you can read the source that quote comes from - part I of a series of interviews with Cem on uTest.com.
Or come to the Conference for the Association for Software Testing - CAST 2010 - next week in Grand Rapids, Michigan, where Cem is giving a keynote-level speech.
UPDATE: I've been thinking about it, and it certainly depends how you define your metrics. For example, one kind of measurement that I am in favor of is the "slip chart." In other words, you look at every deadline the team has committed to and how late they are, and you figure the rough percentage of how late they /always/ are. With that, you can predict when you will really be done. The folks in the extreme programming community codify this into story points and burnup charts, and that's fine. I don't have a problem with these used as approximations; as first-order measurements. The problem comes when they are reduced to 100-word silver-bullets without the context of how they can be used well.
So I wouldn't say I'm totally opposed to metrics are part of a balanced breakfast. I'm just leery of the common ideology of measurement by numbers alone, without context.
Categories: Blogs
Is Your Software Development Organization Agile?
Elisabeth Hendrickson gave a wonderful keynote on Agile-Testing at STAREast this year, and my Friend Dan Mondello took her definition of Agile and codified it with an article. It's a nice read.
We were sitting together for the keynote, and Dan threw in a picture of us at the bottom of the article. You can see Selena Delesie and Lanette Creamer at the left of the photo, Dan at right, and me in the middle.
If you're not interested in the article, check out the photo. That someone as dopey as me has managed to have some modest amount of success in this field should be encouragement to dopey people everywhere! :-)
We were sitting together for the keynote, and Dan threw in a picture of us at the bottom of the article. You can see Selena Delesie and Lanette Creamer at the left of the photo, Dan at right, and me in the middle.
If you're not interested in the article, check out the photo. That someone as dopey as me has managed to have some modest amount of success in this field should be encouragement to dopey people everywhere! :-)
Categories: Blogs
Selenium IDE
The folks at TechTarget just asked me to write an article on Selenium IDE, the integrated, simple, easy-to-use, free browser-driving automation tool for FireFox.
And they just published it!
You can get the article right now at http://searchsoftwarequality.techtarget.com/tip/0,289483,sid92_gci1516589,00.html. (Free Registration Required)
And they just published it!
You can get the article right now at http://searchsoftwarequality.techtarget.com/tip/0,289483,sid92_gci1516589,00.html. (Free Registration Required)
Categories: Blogs
Matt's Big CAST Announcement - Part II
Did you know that CAST will be serving breakfast, lunch, and snacks? They are come free with your conference registration.
More than that, we are currently working on sponsors to provide pop and snacks at dinner every night -- Monday, Tuesday, and Wednesday. And it's possible those turn into dinners.
If you are attending CAST on your own dime, you'll be able to load up on free food and only have to pay for a minimal number of dinners. Likewise, if the company is sending you, between hotel discounts and all the food, the cost will be significantly less than your typical "Big Box" conference with the sixty dollar dinner buffet. You can practically afford to send two for the price of one. I will be personally sponsoring snacks in the Rebel Alliance hospitality suite on Tuesday night. Don't say I never gave ya nothin'. :-)
Of course, since CAST doesn't work with rebels the way, say, a 'STAR' conference might, we may go for a different theme. "The CAST Aways" probably works. Just don't call me "little buddy" and hit me with your hat!
Other stuff - there's a tool called "Is the website down or is it just me?" that seems handy-dandy, and I've got a lot more blog material up at the Software Test Professionals site. If you're hungry for posts from me, check out that site. I'll have a blog post every week (or more), plus, new, a "This Week in Software Testing" podcast up once a week or more.
More than that, we are currently working on sponsors to provide pop and snacks at dinner every night -- Monday, Tuesday, and Wednesday. And it's possible those turn into dinners.
If you are attending CAST on your own dime, you'll be able to load up on free food and only have to pay for a minimal number of dinners. Likewise, if the company is sending you, between hotel discounts and all the food, the cost will be significantly less than your typical "Big Box" conference with the sixty dollar dinner buffet. You can practically afford to send two for the price of one. I will be personally sponsoring snacks in the Rebel Alliance hospitality suite on Tuesday night. Don't say I never gave ya nothin'. :-)
Of course, since CAST doesn't work with rebels the way, say, a 'STAR' conference might, we may go for a different theme. "The CAST Aways" probably works. Just don't call me "little buddy" and hit me with your hat!
Other stuff - there's a tool called "Is the website down or is it just me?" that seems handy-dandy, and I've got a lot more blog material up at the Software Test Professionals site. If you're hungry for posts from me, check out that site. I'll have a blog post every week (or more), plus, new, a "This Week in Software Testing" podcast up once a week or more.
Categories: Blogs
Matt's Big CAST Announcement - Part I
So CAST - the Conference for the Association for Software Testing - 2010 is coming up, and it's going to be about 40 miles from my place, at the Prince Conference Center at Calvin College.
Last time I checked, the Prince Center was very close to booked solid; people keep asking me where to stay.
I've got two concrete suggestions for you:
#1 Country Inn and Suites - Here's the details from an email forward for 35% off!
See all participating Country Inns & Suites By Carlson hotels or check out participating hotels in your favorite regions:
- U.S. - Midwest
- U.S. - Northeast
- U.S. - Southeast
- U.S. - Central Midwest
- U.S. - Southwest
- U.S. - West
- U.S. - Mountain
- Costa Rica
You'll want the Country Inn and Suites in Grand Rapids, Michigan, on East Beltline Road. It's about 3 miles north of the Prince Conference Center.
#2 Any Marriott You'd like
Specifically, the FairField Inn Grand Rapids, which is maybe 1.5 miles from Prince, probably more like 1. If you really feel like pushing it, you could sign up for a Marriott Rewards credit card, get the points and certificate, and use those. It should be about enough combined for two free nights stay -- but read the fine print. The points often aren't credited to you until the first statement is cut. So you might want to stay at Mariott next time, and use the 35% off the Country Inn and Suites this time.
I hope that helps.
Categories: Blogs
Technical Debt: Refired
Phil Kirkham has the cover story of T.E.S.T. magazine this month, talking about technical debt. It's a good article and I recommend it. One thing I like about it is that Phil tries hard to provide a framework for thinking about tech debt, that he borrows from Martin Fowler.
All of it reminds me of the technical debt workshop we ran at Calvin College in 2008. Not a whole lot poured out of that workshop -- it was only two days long, and we spent most of the time trying to come to get to the 'gelled' state so we could make progress. Ron and Chet weren't keen on the term, Chris McMahon came out against analogy shortly thereafter. There wasn't a lot published afterward; I saw a few proposals go to magazine editors but they weren't wild about the formats we proposed. There were a few good presentations we recorded, but I'm afraid there's lots of editing required, and the videos from the event remain locked on a hard drive on my desk.
It wasn't until a week after the conference, on the e-mail discussion list, that we started to see some of the collaborative, building comments I had hope to have during the conference.
Perhaps, if it had been three days, we'd have gotten there. Perhaps, if it had been three days, I'd be saying "if it had only been four days." I don't know.
Using metaphors to describe our work does have certain risks, but I still think that in many cases the tech debt metaphor can have more value than the risk it creates.
It may be time for me to start writing about this again -- or considering a 2011 or 2012 workshop. I'm not sure.
In the mean time, I've got three other new projects. First, a series of interviews with testers called "Testers at work", currently on the back-burner. Second, a book project on changing the cost/value ration of software testing, that i've announced here, and third, a (near) weekly series of podcast interviews with testers I'm nick-naming "This Week in Software Testing", or TWiST, the first of which is up here.
In other news, you can now catch my blogs in two places -- both here and at www.softwaretestpro.com/blog. The majority of my blogging will likely be at STP, but if the topic is the kind of thing that needs a disclaimer, you'll likely find it here.
More to come.
"Welcome all my friends to the show that never ends; I'm so glad you can attend. Come inside, come inside ..."
All of it reminds me of the technical debt workshop we ran at Calvin College in 2008. Not a whole lot poured out of that workshop -- it was only two days long, and we spent most of the time trying to come to get to the 'gelled' state so we could make progress. Ron and Chet weren't keen on the term, Chris McMahon came out against analogy shortly thereafter. There wasn't a lot published afterward; I saw a few proposals go to magazine editors but they weren't wild about the formats we proposed. There were a few good presentations we recorded, but I'm afraid there's lots of editing required, and the videos from the event remain locked on a hard drive on my desk.
It wasn't until a week after the conference, on the e-mail discussion list, that we started to see some of the collaborative, building comments I had hope to have during the conference.
Perhaps, if it had been three days, we'd have gotten there. Perhaps, if it had been three days, I'd be saying "if it had only been four days." I don't know.
Using metaphors to describe our work does have certain risks, but I still think that in many cases the tech debt metaphor can have more value than the risk it creates.
It may be time for me to start writing about this again -- or considering a 2011 or 2012 workshop. I'm not sure.
In the mean time, I've got three other new projects. First, a series of interviews with testers called "Testers at work", currently on the back-burner. Second, a book project on changing the cost/value ration of software testing, that i've announced here, and third, a (near) weekly series of podcast interviews with testers I'm nick-naming "This Week in Software Testing", or TWiST, the first of which is up here.
In other news, you can now catch my blogs in two places -- both here and at www.softwaretestpro.com/blog. The majority of my blogging will likely be at STP, but if the topic is the kind of thing that needs a disclaimer, you'll likely find it here.
More to come.
"Welcome all my friends to the show that never ends; I'm so glad you can attend. Come inside, come inside ..."
Categories: Blogs
A couple of articles, coming at you
I've always felt that the test automation literature is a little ... odd.
On the one hand, you have the floofly high-level tutorial, written so that it applies to every test tool. By trying to apply to every tool, the tutorial isn't specific enough to provide step-by-step instructions.
Then you buy something and try to use it. UGH. That's no fun. So maybe you hire some consultants to build the framework and basic test suite. That's great, but it costs you $8,000 a week - and - $320,000 and six months later, and the consultants leave, do you have the expertise to maintain the suite?
All too often, two years later, you've got some really expensive shelfware.
So I'm interested in tutorials about test tools. The thing is, once you get specific, the market is much smaller -- so it's rare to find a mass-market book about testing with a specific test tool.
But that's the other thing. With some test tools today, you don't need one. All the readers need is an article bit enough to get them started -- or a short series to keep them learning.
That said, the folks at SearchSoftwareQuality invited me to do a small series on Selenium, and have published the first two parts - a two-part mini-series on Selenium RC. Read all about it - part one and part two, online, for free.
The next article in the series will address Selenium IDE, the development environment for Selenium. I hope you find them enjoyable and helpful.
On the one hand, you have the floofly high-level tutorial, written so that it applies to every test tool. By trying to apply to every tool, the tutorial isn't specific enough to provide step-by-step instructions.
Then you buy something and try to use it. UGH. That's no fun. So maybe you hire some consultants to build the framework and basic test suite. That's great, but it costs you $8,000 a week - and - $320,000 and six months later, and the consultants leave, do you have the expertise to maintain the suite?
All too often, two years later, you've got some really expensive shelfware.
So I'm interested in tutorials about test tools. The thing is, once you get specific, the market is much smaller -- so it's rare to find a mass-market book about testing with a specific test tool.
But that's the other thing. With some test tools today, you don't need one. All the readers need is an article bit enough to get them started -- or a short series to keep them learning.
That said, the folks at SearchSoftwareQuality invited me to do a small series on Selenium, and have published the first two parts - a two-part mini-series on Selenium RC. Read all about it - part one and part two, online, for free.
The next article in the series will address Selenium IDE, the development environment for Selenium. I hope you find them enjoyable and helpful.
Categories: Blogs
Awards and Decorations
A little over ten years ago I interviewed for a developer job with Steve Hoek, at a tiny division of McGraw-Hill. Steve asked me if I saw myself collecting an award in five years - what would it be?
I replied that I'd like it to be some sort of industry award, but those sort of things don't really exist for do-ers. I hope I never forget how Steve replied -- he agreed with me. Those kind of things do not exist for do-ers.
Maybe they should.
Oh, yes, there are lots of problems. There are conflicts of interest. The inner ring-ers tend to get them disproportionately, without making significant or meaningful contributions to testing. Plenty of people do good work, dedicated work, for decades without doing any PR, and are virtually unknown. But I do think that a good faith effort, on balance, would be better than none at all. The Agile movement has the Gordon Pask Award; can Software Testing do something similar?
I think we can.
So I was very please earlier in the month when the folks at Software Test & Quality Assurance Magazine announced the Luminary Award. Please allow me to read the qualifications from the nomination page:
The software test and quality community will decide who receives the award, and the expectation is the community will choose a nominee with some or all of these qualities:
•A person that has inspired individuals to be better at their profession
•A person that has dedicated their career for the betterment of software testing and quality
•A person that has shown exceptional leadership in promoting the advancement of our industry
•A person that has shown exceptional ability in promoting and educating concepts for the advancement of the industry
•A person that has published and promoted the industry that has resulted in greater recognition and respect for the industry
•A person deserving of an award based on merit not particular to personality (although nominees may be very popular we want this to be about career achievements)
When I read those requirements, a few words jump out at me. Some appear on the page, others are just how I summarize the tone:.
* Testing AND Quality
* Published
* Community
* Lifetime
When I think of those words, one name comes to mind: Cem Kaner.
Why Cem Kaner? Well, he's been around a long time. In the early 1980's he got a PhD in Cognitive Psychology, and he has been advancing the cause of software as a human-driven activity ever since. After doing some programming and testing in that PhD Program. Dr. Kaner moved to Silicon Valley where he worked as a tester, programmer, test manager, documentation manager -- and get this -- he got a job at night in a retail software store while he held those jobs so he could get closer to the customer.
In 1988 Dr. Kaner wrote the first edition of "Testing Computer Software", a landmark book in the field, which he brought in two co-authors to help revise into the 2nd edition in 1993. After Silly Valley, Dr. Kaner got a law degree, practiced law for a short time, and went on to write Bad Software, one of very few books about the legal implications of defective software. In 2000 he moved to Florida where he started teaching software engineering at Florida's institute of technology, the only school in the nation to my knowledge to offer an earned minor in software testing. Some of Dr. Kaner's former students are heavily involved in the test community today.
Then there is testingeducation.org, the source for free and open testing training materials. There is the Association For Software Testing (AST) that Dr. Kaner helped champion and organize. There is the Black-Box Software Testing (BBST) course that the AST runs for free to it's members, that Dr. Kaner helped create. There's Lessons Learned in Software Testing, the most recent book Dr. Kaner has served as lead author on. Not to mention his publication list, which would likely use up more than all of the ink currently in my printer. And I haven't even mentioned the LAWST-Style peer conferences that Dr. Kaner lead, inspired, and served as an initial host for.
That's the short list; I've tried to hit the highlights. But notice something about those publications: He was almost always the lead author. They weren't done alone; they were done as a part of a community. And he doesn't do the classic professor trick of sticking his name on top and letting the "little people" do the work: Dr. Kaner is a do-er through and through.
Now look at the dates - this is a gentleman who views career as a marathon. He started his first testing job on his resume in June of 1983, and hasn't stopped since. I already mentioned his publication list.
I don't think it takes a genius to connect the dots back those initial requirements, nor to see how Dr. Kaner is uniquely qualified for the luminary award.
Yes, I said uniquely.
When I think of a 'luminary award', sure, there are other names that come to mind, but when I think of those four bullet points, only one name passes the test. But, just as an exercise, let's talk about some of the folks who might deserve an award, just not this one:
Dr. Jerry Weinberg: Author of over forty books on computing, consulting and quality, Jerry has also ran an enviable marathon career, going back to being the test lead on the Apollo spacecraft in the 1960's! Jerry built his own special community and inspired a generation of consultants. James Bach went so far as to refer to him once as the "Prince of Testers." But that other requirement is to focus your career on quality and testing, and I'm afraid I can't say that was Jerry's focus. Consulting, certainly, and Delivery, yes, and maybe Quality, but testing is not the main thing of the main thing of the main thing for Jerry the way it is for Cem.
Dr. Paul Jorgensen: My old professor at Grand Valley, Paul Jorgensen spent twenty years in industry testing telephone switches before getting his PhD and teaching software engineering for twenty more. Paul is also the author of Software Testing: A Craftsman's Approach, now in it's third edition. Massively published, long career, testing and quality, yes, and I have great admiration for Paul. What I can't say is that he's built a community around him the way Cem has. Every one of Cem's project is collaborative, every one seems to get new people involved. Paul deserves and award for individual effort, certainly, and has done a great deal to foster community in his decades of service. I just don't think he's a perfect fit for this particular community award.
Buccaneer Scholar James Bach: Is probably the closest. James also contributed to create AST, he's keynoted at every single international test conference, he also co-authored LLST, and he's invested a fair amount of time in helping new testers get into the craft. Like Cem, his publication list is a mile long, and he's done a significant amount of training of testers, as well as doing. Yet he started later, and he was a contributor on one book, not a lead author of three. I've no doubt James is worthy of a luminary award, I'm just not sure it should be the very first one. Also, I expect in the years to come, he will become more and more worthy.
There are a few other names you could mention. Michael Bolton, for example, probably falls into the same category of James, and so do my friends Mike Kelly, Ben Simo, and a handful of others. There are also several people like Glenford Meyers, Richard Bender, or Boris Beizer who did very promising work but didn't stick around to wrestle with how their ideas were implemented in practice, to see if they worked. One thing Cem and James have done is come back to revise those ideas in light of feedback from the field - both in terms of what works, and how to explain the concept so there is a high transference rate.
Likewise, there are some agile-testing names that may be eligible for the award in a few years, but right now, I'm looking at the intersection of lifetime, published, community, testing and quality. And I'm looking at Cem Kaner.
Now I am open to other people interpreting those requirements differently, and I am open to debate. You can suggest anyone you please, and nominate and later vote for anyone you please. But I hope that, after some reflection, you will join me in nominating Dr. Cem Kaner for the STP Luminary Award within the next week. After nominations, STP will open the community to vote on the top three or four selections.
The nomination form is up on the web right now; please join me in nominating someone you respect, value, and appreciate.
Thank you!
I replied that I'd like it to be some sort of industry award, but those sort of things don't really exist for do-ers. I hope I never forget how Steve replied -- he agreed with me. Those kind of things do not exist for do-ers.
Maybe they should.
Oh, yes, there are lots of problems. There are conflicts of interest. The inner ring-ers tend to get them disproportionately, without making significant or meaningful contributions to testing. Plenty of people do good work, dedicated work, for decades without doing any PR, and are virtually unknown. But I do think that a good faith effort, on balance, would be better than none at all. The Agile movement has the Gordon Pask Award; can Software Testing do something similar?
I think we can.
So I was very please earlier in the month when the folks at Software Test & Quality Assurance Magazine announced the Luminary Award. Please allow me to read the qualifications from the nomination page:
The software test and quality community will decide who receives the award, and the expectation is the community will choose a nominee with some or all of these qualities:
•A person that has inspired individuals to be better at their profession
•A person that has dedicated their career for the betterment of software testing and quality
•A person that has shown exceptional leadership in promoting the advancement of our industry
•A person that has shown exceptional ability in promoting and educating concepts for the advancement of the industry
•A person that has published and promoted the industry that has resulted in greater recognition and respect for the industry
•A person deserving of an award based on merit not particular to personality (although nominees may be very popular we want this to be about career achievements)
When I read those requirements, a few words jump out at me. Some appear on the page, others are just how I summarize the tone:.
* Testing AND Quality
* Published
* Community
* Lifetime
When I think of those words, one name comes to mind: Cem Kaner.
Why Cem Kaner? Well, he's been around a long time. In the early 1980's he got a PhD in Cognitive Psychology, and he has been advancing the cause of software as a human-driven activity ever since. After doing some programming and testing in that PhD Program. Dr. Kaner moved to Silicon Valley where he worked as a tester, programmer, test manager, documentation manager -- and get this -- he got a job at night in a retail software store while he held those jobs so he could get closer to the customer.
In 1988 Dr. Kaner wrote the first edition of "Testing Computer Software", a landmark book in the field, which he brought in two co-authors to help revise into the 2nd edition in 1993. After Silly Valley, Dr. Kaner got a law degree, practiced law for a short time, and went on to write Bad Software, one of very few books about the legal implications of defective software. In 2000 he moved to Florida where he started teaching software engineering at Florida's institute of technology, the only school in the nation to my knowledge to offer an earned minor in software testing. Some of Dr. Kaner's former students are heavily involved in the test community today.
Then there is testingeducation.org, the source for free and open testing training materials. There is the Association For Software Testing (AST) that Dr. Kaner helped champion and organize. There is the Black-Box Software Testing (BBST) course that the AST runs for free to it's members, that Dr. Kaner helped create. There's Lessons Learned in Software Testing, the most recent book Dr. Kaner has served as lead author on. Not to mention his publication list, which would likely use up more than all of the ink currently in my printer. And I haven't even mentioned the LAWST-Style peer conferences that Dr. Kaner lead, inspired, and served as an initial host for.
That's the short list; I've tried to hit the highlights. But notice something about those publications: He was almost always the lead author. They weren't done alone; they were done as a part of a community. And he doesn't do the classic professor trick of sticking his name on top and letting the "little people" do the work: Dr. Kaner is a do-er through and through.
Now look at the dates - this is a gentleman who views career as a marathon. He started his first testing job on his resume in June of 1983, and hasn't stopped since. I already mentioned his publication list.
I don't think it takes a genius to connect the dots back those initial requirements, nor to see how Dr. Kaner is uniquely qualified for the luminary award.
Yes, I said uniquely.
When I think of a 'luminary award', sure, there are other names that come to mind, but when I think of those four bullet points, only one name passes the test. But, just as an exercise, let's talk about some of the folks who might deserve an award, just not this one:
Dr. Jerry Weinberg: Author of over forty books on computing, consulting and quality, Jerry has also ran an enviable marathon career, going back to being the test lead on the Apollo spacecraft in the 1960's! Jerry built his own special community and inspired a generation of consultants. James Bach went so far as to refer to him once as the "Prince of Testers." But that other requirement is to focus your career on quality and testing, and I'm afraid I can't say that was Jerry's focus. Consulting, certainly, and Delivery, yes, and maybe Quality, but testing is not the main thing of the main thing of the main thing for Jerry the way it is for Cem.
Dr. Paul Jorgensen: My old professor at Grand Valley, Paul Jorgensen spent twenty years in industry testing telephone switches before getting his PhD and teaching software engineering for twenty more. Paul is also the author of Software Testing: A Craftsman's Approach, now in it's third edition. Massively published, long career, testing and quality, yes, and I have great admiration for Paul. What I can't say is that he's built a community around him the way Cem has. Every one of Cem's project is collaborative, every one seems to get new people involved. Paul deserves and award for individual effort, certainly, and has done a great deal to foster community in his decades of service. I just don't think he's a perfect fit for this particular community award.
Buccaneer Scholar James Bach: Is probably the closest. James also contributed to create AST, he's keynoted at every single international test conference, he also co-authored LLST, and he's invested a fair amount of time in helping new testers get into the craft. Like Cem, his publication list is a mile long, and he's done a significant amount of training of testers, as well as doing. Yet he started later, and he was a contributor on one book, not a lead author of three. I've no doubt James is worthy of a luminary award, I'm just not sure it should be the very first one. Also, I expect in the years to come, he will become more and more worthy.
There are a few other names you could mention. Michael Bolton, for example, probably falls into the same category of James, and so do my friends Mike Kelly, Ben Simo, and a handful of others. There are also several people like Glenford Meyers, Richard Bender, or Boris Beizer who did very promising work but didn't stick around to wrestle with how their ideas were implemented in practice, to see if they worked. One thing Cem and James have done is come back to revise those ideas in light of feedback from the field - both in terms of what works, and how to explain the concept so there is a high transference rate.
Likewise, there are some agile-testing names that may be eligible for the award in a few years, but right now, I'm looking at the intersection of lifetime, published, community, testing and quality. And I'm looking at Cem Kaner.
Now I am open to other people interpreting those requirements differently, and I am open to debate. You can suggest anyone you please, and nominate and later vote for anyone you please. But I hope that, after some reflection, you will join me in nominating Dr. Cem Kaner for the STP Luminary Award within the next week. After nominations, STP will open the community to vote on the top three or four selections.
The nomination form is up on the web right now; please join me in nominating someone you respect, value, and appreciate.
Thank you!
Categories: Blogs
Claimin' my blog
Part of the process of dusting off Creative Chaos and reusing it again is moving my RSS Feed. The new RSS feed is http://feeds.feedburner.com/CreativeChaos. In order to index the RSS feed, I need to tell Technorati, a blog index-er, that I own the blog and the new feed is correct. To do that, they give me a random 'code' and I put it in a blog post, indicating that I truely have control of the site.
Here's the code: 8Q94285VRTT4
Thank you for your coorperation as we remodel my blog; I apologize for the mess! :-)
Here's the code: 8Q94285VRTT4
Thank you for your coorperation as we remodel my blog; I apologize for the mess! :-)
Categories: Blogs
I'm just not excited about efficiency
... So we're driving out to my daughter's softball game in Hopkins, Michigan, and I choose to go straight instead of turning left at the shortcut. We've got plenty of time, and if I go straight I can turn past Sacred Heart Mission, which is a nicer drive.
My wife points out that this is not the fastest route. This is not the first time I choose the slower route; I commonly turn out the driveway to go downtown instead of taking M-89. Why is that?
I Pause.
Have you ever had one of those moments where you realize there is no right answer? Where the other person asks "why" or "couldn't you just...?" but they actually mean "There is no good reason" or "You should"?
Again, I Pause.
Carefully choosing each word, I say something like "well, I'm reluctant to answer. A proper answer would take a long time. But the short version is this: I suspect that I do not value efficiency as much as you do."
That did not go over well.
So let me share with you the longer version, that went over better.
Efficient ... or Effective?
We tend to use these terms interchangeably, but when you look them up in the dictionary, they mean subtly different things. Efficiency is a measure of utilization; we might say a gas furnace is highly efficient if most of heat goes out the heating ducts and into the house -- as opposed to out the exhaust, or heating the basement. A 100% efficient furnace would let no heat escape. When it comes to technology workers, we typically mean a team is 'efficient' if everyone is working, all the time. To do this with clerical jobs, we can create an always-filling inbox, but with knowledge workers we typically need to have them work on multiple projects at the same time.
Effectiveness, however, is different. Effectiveness is how good we are at accomplishing the task at hand. Consider the typical fire department. Couched in these terms, it is not efficient - 95% of the time a paid staff is sitting around, training or cleaning the equipment. Yet when we do have those 5% of emergencies, we need someone to respond quickly.
In that case, the fire department has a conflict between efficiency and effectiveness. And, for the sake of public safety, they'll choose effectiveness every time.
Likewise, all over the world, there are fighter pilots and entire airborne brigades standing at the ready, collecting salary. 99% of he time, they'll be sitting around playing cards. But you want them standing at the ready, don't you? I sure do.
In some jobs designed a certain way -- perhaps on an assembly line -- efficiency and effectiveness can be interchangeable. Not so in software.
This isn't my idea; Tom DeMarco explains how administrative assistants need a quick response time, thus they need to have built in slack in his book by the same name. Eli Goldratt, creator of the theory of constraints, talks about having factory employees waiting for the instant that the machine is done in order to load quickly in his opus The Goal.
But what's that got to do with driving to a softball game?
Well, it has to do with optimization. Optimization is another word we like to use a lot, and it ties back to efficiency or effectiveness. Optimization doesn't just mean doing a better job at the work. It means pushing the work toward the best possible way of doing it -- the optimal way.
The things, I find that optimal is generally best for a given way of thinking.
When we do the building project, and we get to the point that we are trying to save a penny on each brick, someone will say "hey, with ten thousand bricks, you are starting to talk about real money."
I yawn. Or, more accurately, when we get to an optimization problem, I start looking for a different way of thinking. For example, if we are near the best we can possibly get for brick, let's look into concrete or some other building supply.
Optimization, in my experience, is often a lot of work to squeeze out a little more reward.
I'm not excited about it.
In a similar way, if I'm water-sealing my deck, and I'm moderately fast at it, I don't look for a faster system of using a paint brush. Instead, I'll look into using a paint roller. I don't want to optimize paint-rolling, either -- I discovered spray-on water seal.
It's the same thing in software. Oh, sure, it's good to write a batch script to save us some typing. It's good to have an automated setup of our test environment instead of manually configuring. Don't get me wrong. But there's something that happens -- some tipping point, where we are spending a lot of mental energy to eke out an improvement from 95% to 97.5%. Me, I'll just leave it at 95% and look for something else to improve.
In some cases I just don't find any more improvement. Until hovercraft, low-flying aircraft, or the teleporter become popular, the fastest way to get to Hopkins from Allegan is going to be taking A-37.
It's going to take at least 55 seconds longer to get there if you take 20th Avenue.
Forc me to choose between efficiency and effectiveness and I'll choose effective every time. Yet even when the two align, sometimes the difference in approaches is round-off error. And, when that happens, there are other things to consider, like Aesthetics.
20th Avenue is a drive I haven't taken a thousand times, and it has it's own special charms, including Fat Blossom Farm and Sacred Heart Hall.
I think I'll take it.
My wife points out that this is not the fastest route. This is not the first time I choose the slower route; I commonly turn out the driveway to go downtown instead of taking M-89. Why is that?
I Pause.
Have you ever had one of those moments where you realize there is no right answer? Where the other person asks "why" or "couldn't you just...?" but they actually mean "There is no good reason" or "You should"?
Again, I Pause.
Carefully choosing each word, I say something like "well, I'm reluctant to answer. A proper answer would take a long time. But the short version is this: I suspect that I do not value efficiency as much as you do."
That did not go over well.
So let me share with you the longer version, that went over better.
Efficient ... or Effective?
We tend to use these terms interchangeably, but when you look them up in the dictionary, they mean subtly different things. Efficiency is a measure of utilization; we might say a gas furnace is highly efficient if most of heat goes out the heating ducts and into the house -- as opposed to out the exhaust, or heating the basement. A 100% efficient furnace would let no heat escape. When it comes to technology workers, we typically mean a team is 'efficient' if everyone is working, all the time. To do this with clerical jobs, we can create an always-filling inbox, but with knowledge workers we typically need to have them work on multiple projects at the same time.
Effectiveness, however, is different. Effectiveness is how good we are at accomplishing the task at hand. Consider the typical fire department. Couched in these terms, it is not efficient - 95% of the time a paid staff is sitting around, training or cleaning the equipment. Yet when we do have those 5% of emergencies, we need someone to respond quickly.
In that case, the fire department has a conflict between efficiency and effectiveness. And, for the sake of public safety, they'll choose effectiveness every time.
Likewise, all over the world, there are fighter pilots and entire airborne brigades standing at the ready, collecting salary. 99% of he time, they'll be sitting around playing cards. But you want them standing at the ready, don't you? I sure do.
In some jobs designed a certain way -- perhaps on an assembly line -- efficiency and effectiveness can be interchangeable. Not so in software.
This isn't my idea; Tom DeMarco explains how administrative assistants need a quick response time, thus they need to have built in slack in his book by the same name. Eli Goldratt, creator of the theory of constraints, talks about having factory employees waiting for the instant that the machine is done in order to load quickly in his opus The Goal.
But what's that got to do with driving to a softball game?
Well, it has to do with optimization. Optimization is another word we like to use a lot, and it ties back to efficiency or effectiveness. Optimization doesn't just mean doing a better job at the work. It means pushing the work toward the best possible way of doing it -- the optimal way.
The things, I find that optimal is generally best for a given way of thinking.
When we do the building project, and we get to the point that we are trying to save a penny on each brick, someone will say "hey, with ten thousand bricks, you are starting to talk about real money."
I yawn. Or, more accurately, when we get to an optimization problem, I start looking for a different way of thinking. For example, if we are near the best we can possibly get for brick, let's look into concrete or some other building supply.
Optimization, in my experience, is often a lot of work to squeeze out a little more reward.
I'm not excited about it.
In a similar way, if I'm water-sealing my deck, and I'm moderately fast at it, I don't look for a faster system of using a paint brush. Instead, I'll look into using a paint roller. I don't want to optimize paint-rolling, either -- I discovered spray-on water seal.
It's the same thing in software. Oh, sure, it's good to write a batch script to save us some typing. It's good to have an automated setup of our test environment instead of manually configuring. Don't get me wrong. But there's something that happens -- some tipping point, where we are spending a lot of mental energy to eke out an improvement from 95% to 97.5%. Me, I'll just leave it at 95% and look for something else to improve.
In some cases I just don't find any more improvement. Until hovercraft, low-flying aircraft, or the teleporter become popular, the fastest way to get to Hopkins from Allegan is going to be taking A-37.
It's going to take at least 55 seconds longer to get there if you take 20th Avenue.
Forc me to choose between efficiency and effectiveness and I'll choose effective every time. Yet even when the two align, sometimes the difference in approaches is round-off error. And, when that happens, there are other things to consider, like Aesthetics.
20th Avenue is a drive I haven't taken a thousand times, and it has it's own special charms, including Fat Blossom Farm and Sacred Heart Hall.
I think I'll take it.
Categories: Blogs
37Signals -- and Inc. Magazine
Do you remember when Monster.com was actually a good source of job leads? Why, when it started, only tech-savvy people, techy-savvy hiring managers, and a handful of recruiters knew about it.
Then something happened. Happy, bright people told their friends, who might not have been quite as bright. They got big. They ran a TV ad During The Superbowl. The recruiters got wind that Monster.com was a place to find talent, and started putting out 'skills' based job ads, so they could build a rolodex, just in case a job popped up later.
Popularity brought with it to the attention of other companies that are interested in meeting people looking to improve their career prospects - colleges, online diploma mills, resume-writing services. These other companies started to offer Monster "Business Development" deals (money) for access to the monster candidate list.
Being good guys, the Monster Management refused, instead allowing you, the customer, to opt-in. Over time, they started to force you to at least re-consider every so often. Then they made the 'yes I'm interested' buttons bigger and bigger and the 'no thanks but let me see the job' buttons smaller and smaller.
Eventually, all the cool kids found craigslist. The problem was, happy, bright people who were pleased with Craiglist told their friends ...
About that same time Joel Spolsky's website, JoelOnSoftware, was hitting the top of it's popularity. Joel had created a regular "gravity well" for developers, and created jobs.joelonsoftware.com. It's a super-easy, super-simple site with only a few hundred listings. When it launched, and to some extent, even now, you don't really need to search. Just scroll down looking for interesting gigs.
Sadly, jobs.joelonsoftware.com doesn't really have a test/QA focus, so the search goes on. Two more sites I find interesting are jobs.freelanceswitch.com, another site that did the gravity-well thing, this time for freelancers, and jobs.37signals.com.
37Signals is that company that built basecamp, the office productivity tool, and has expanded to create a whole series of web-based collaboration tools. It's not really testing focused; if anything, they focus on graphic design -- but one thing they do have is really good writing.
Speaking of writing, the CEO of 37Signals has a monthly column in Inc. Magazine; I just got the June issue Saturday. This month's column is called "Never read another resume." It's for hiring manager, and suggests that resumes are inflated, misleading, or just plain tough to sort out. Jason suggests hiring for people the write a custom cover letter or have a portfolio of work over poring through resumes. (Or, maybe, people who find you at the cool website before it jumps the shark.)
When I started writing this, I intended to link to the article, but I'm afraid it's not available online yet. They may delay it for a few weeks (to give a perk to us paid subscribers), or they might never offer it online.
I subscribe to a half-dozen magazines, but there are only two I devour every time; Inc is one of them. You can pick up a subscription to Inc for ten bucks. For that matter, before you buy that next airline ticket, sign up for a rewards program with that airline. You'll likely earn enough points on a single flight to get a free subscription to Inc.
I have no financial relationship to Inc.; I am simply a fan.
Speaking of being a fan, I'll be at CAST 2010 in August. It'll be at the Prince Conference Center at Calvin College, but last week the hotel sold out of rooms. I recommend the Residence Inn (by Marriott) or Country Inn & Suites (my Mariott), both on East Beltline. I have to drive up east Beltline to get to the Conference Center, so if you're a longtime reader (ideally one I've met in person) and book a room (preferably at the Residence Inn), I might be able to swing by to pick you up.
We may also organize a few 'rebel alliance' after-conference events for CAST.
Stick around, more CAST details to come.
UPDATE: The Prince Conference Center reservation system has issues. (Shocking!) The word on the street is that they still have 5 rooms left for the conference. You can call 1-866-526-7200 to talk to a person.
Then something happened. Happy, bright people told their friends, who might not have been quite as bright. They got big. They ran a TV ad During The Superbowl. The recruiters got wind that Monster.com was a place to find talent, and started putting out 'skills' based job ads, so they could build a rolodex, just in case a job popped up later.
Popularity brought with it to the attention of other companies that are interested in meeting people looking to improve their career prospects - colleges, online diploma mills, resume-writing services. These other companies started to offer Monster "Business Development" deals (money) for access to the monster candidate list.
Being good guys, the Monster Management refused, instead allowing you, the customer, to opt-in. Over time, they started to force you to at least re-consider every so often. Then they made the 'yes I'm interested' buttons bigger and bigger and the 'no thanks but let me see the job' buttons smaller and smaller.
Eventually, all the cool kids found craigslist. The problem was, happy, bright people who were pleased with Craiglist told their friends ...
About that same time Joel Spolsky's website, JoelOnSoftware, was hitting the top of it's popularity. Joel had created a regular "gravity well" for developers, and created jobs.joelonsoftware.com. It's a super-easy, super-simple site with only a few hundred listings. When it launched, and to some extent, even now, you don't really need to search. Just scroll down looking for interesting gigs.
Sadly, jobs.joelonsoftware.com doesn't really have a test/QA focus, so the search goes on. Two more sites I find interesting are jobs.freelanceswitch.com, another site that did the gravity-well thing, this time for freelancers, and jobs.37signals.com.
37Signals is that company that built basecamp, the office productivity tool, and has expanded to create a whole series of web-based collaboration tools. It's not really testing focused; if anything, they focus on graphic design -- but one thing they do have is really good writing.
Speaking of writing, the CEO of 37Signals has a monthly column in Inc. Magazine; I just got the June issue Saturday. This month's column is called "Never read another resume." It's for hiring manager, and suggests that resumes are inflated, misleading, or just plain tough to sort out. Jason suggests hiring for people the write a custom cover letter or have a portfolio of work over poring through resumes. (Or, maybe, people who find you at the cool website before it jumps the shark.)
When I started writing this, I intended to link to the article, but I'm afraid it's not available online yet. They may delay it for a few weeks (to give a perk to us paid subscribers), or they might never offer it online.
I subscribe to a half-dozen magazines, but there are only two I devour every time; Inc is one of them. You can pick up a subscription to Inc for ten bucks. For that matter, before you buy that next airline ticket, sign up for a rewards program with that airline. You'll likely earn enough points on a single flight to get a free subscription to Inc.
I have no financial relationship to Inc.; I am simply a fan.
Speaking of being a fan, I'll be at CAST 2010 in August. It'll be at the Prince Conference Center at Calvin College, but last week the hotel sold out of rooms. I recommend the Residence Inn (by Marriott) or Country Inn & Suites (my Mariott), both on East Beltline. I have to drive up east Beltline to get to the Conference Center, so if you're a longtime reader (ideally one I've met in person) and book a room (preferably at the Residence Inn), I might be able to swing by to pick you up.
We may also organize a few 'rebel alliance' after-conference events for CAST.
Stick around, more CAST details to come.
UPDATE: The Prince Conference Center reservation system has issues. (Shocking!) The word on the street is that they still have 5 rooms left for the conference. You can call 1-866-526-7200 to talk to a person.
Categories: Blogs
Free Teleconference - Today & Tomorrow
Software Quality Engineering is currently running Agile Development Practices West -- and they are partnering with CMC Media to offer the keynotes online!
You can register on-line and watch from anywhere with high-speed internet and a speaker. Speakers include Tim Lister, Johanna Rothman, and Bob Martin.
The whole thing is free. I'm looking forward to it.
Update: Check it out, I'm watching Tim Lister right now:
You can register on-line and watch from anywhere with high-speed internet and a speaker. Speakers include Tim Lister, Johanna Rothman, and Bob Martin.
The whole thing is free. I'm looking forward to it.
Update: Check it out, I'm watching Tim Lister right now:
Categories: Blogs
Never Read Another Resume - Now on-line
Hey, check it out, the people at Inc have put put "Never Read Another Resume" on their website.
I really do believe that print journalism needs support right now, and subscribing is win-win. If I was a better behavioral psychologist, and my goal was to get you to subscribe, I think I wouldn't provide the link. At the same time, I believe that information wants to be free, and 'walled gardens' of information are kinda silly.
So there you have it. The article is out there, it's good stuff, I hope you enjoy it.
I really do believe that print journalism needs support right now, and subscribing is win-win. If I was a better behavioral psychologist, and my goal was to get you to subscribe, I think I wouldn't provide the link. At the same time, I believe that information wants to be free, and 'walled gardens' of information are kinda silly.
So there you have it. The article is out there, it's good stuff, I hope you enjoy it.
Categories: Blogs
The more things change ...
"We value responding to change over following a plan"
- Some guys in Utah, 2001
It's been fun posting at "Testing at the Edge of Chaos", over at Software Test Performance Magazine, and also fun contributing to a monthly column with Chris McMahon.
At the same time, the magazine is changing. I'm happy about the changes, and I hope you will be too.
I talk a fair amount about web 2.0 and user created content; heck, in my day job, I test software for an Social Media Company. Yet when you think about a magazine, a website, and a conference, the model is something out of 1953:
A) Assemble 'content' for magazine
B) Deliver 'content' in magazine as one-directional communication
C) Sell advertising to slip in between articles
D) Assemble 'content' for conference ...
... You see where I'm going.
I don't think there's anything wrong with that; magazines deliver value, that's why people read them.
I've just wondered if there were a more democratic way to do that.
Turns out, there is. Or at least, the folks at STP want to try, and I'm excited about that.
First: The website is changing to SoftwareTestPro.com. Yes, we'll require a signup, but you have to have some sort of login to track your contributions, just like Facebook or Myspace or anything else.
Once you are in, you'll be able to read the community blog, a place where many people can contribute to STP, not just me. So yes, "Testing At The Edge of Chaos" is going to be split in half, replaced by the community blog and, yes, a return to "Creative Chaos." Yes, my RSS feed is back to http://feeds.feedburner.com/CreativeChaos.
Second: The whole SoftwareTestPro website has been revamped, including the creation of 'crews.' Crews are a professional membership feature, something like a special interest group. The folks at STP are also throwing some support behind a local chapter program, so you can create local user's groups. A professional membership costs $100 a year, but to get a test, they've opened up my "Ask the Tester" Crew for anyone with the free basic membership -- at least for a limited time.
Third: The monthly column I share with Chris McMahon is changing, from a sort of encyclopedia of testing into "Ask the Tester" -- were we invite experts from the community to be interviewed. Chris wants to move on to writing features and other content, so instead of a on-man interview crew, we'll ask the community to come up with questions -- that way, our experts answer your questions. Our next interview will be with Michael Bolton; but to write the article I'm afraid I need, well ... the questions.
So please, leave a comment with a question for Michael about testing, along with your name, city, and state/province or country.
It's an exciting time for STP, and i'm pleased to be able to dual-blog now.
Thank you for bearing with our construction in progress. I hope you have questions, especially for Michael. Please, ask away! :-0
- Some guys in Utah, 2001
It's been fun posting at "Testing at the Edge of Chaos", over at Software Test Performance Magazine, and also fun contributing to a monthly column with Chris McMahon.
At the same time, the magazine is changing. I'm happy about the changes, and I hope you will be too.
I talk a fair amount about web 2.0 and user created content; heck, in my day job, I test software for an Social Media Company. Yet when you think about a magazine, a website, and a conference, the model is something out of 1953:
A) Assemble 'content' for magazine
B) Deliver 'content' in magazine as one-directional communication
C) Sell advertising to slip in between articles
D) Assemble 'content' for conference ...
... You see where I'm going.
I don't think there's anything wrong with that; magazines deliver value, that's why people read them.
I've just wondered if there were a more democratic way to do that.
Turns out, there is. Or at least, the folks at STP want to try, and I'm excited about that.
First: The website is changing to SoftwareTestPro.com. Yes, we'll require a signup, but you have to have some sort of login to track your contributions, just like Facebook or Myspace or anything else.
Once you are in, you'll be able to read the community blog, a place where many people can contribute to STP, not just me. So yes, "Testing At The Edge of Chaos" is going to be split in half, replaced by the community blog and, yes, a return to "Creative Chaos." Yes, my RSS feed is back to http://feeds.feedburner.com/CreativeChaos.
Second: The whole SoftwareTestPro website has been revamped, including the creation of 'crews.' Crews are a professional membership feature, something like a special interest group. The folks at STP are also throwing some support behind a local chapter program, so you can create local user's groups. A professional membership costs $100 a year, but to get a test, they've opened up my "Ask the Tester" Crew for anyone with the free basic membership -- at least for a limited time.
Third: The monthly column I share with Chris McMahon is changing, from a sort of encyclopedia of testing into "Ask the Tester" -- were we invite experts from the community to be interviewed. Chris wants to move on to writing features and other content, so instead of a on-man interview crew, we'll ask the community to come up with questions -- that way, our experts answer your questions. Our next interview will be with Michael Bolton; but to write the article I'm afraid I need, well ... the questions.
So please, leave a comment with a question for Michael about testing, along with your name, city, and state/province or country.
It's an exciting time for STP, and i'm pleased to be able to dual-blog now.
Thank you for bearing with our construction in progress. I hope you have questions, especially for Michael. Please, ask away! :-0
Categories: Blogs
Are testers “going away”?
(Meta: This post is written in response to a specific hiring situation in Silicon Valley and the Pacific Northwest. If your experience is different, I’d like to hear from you. For example – I don’t think this is an issue in Indianapolis. And that’s a good thing …)
Did you know that at my first big testing conference – in 2004 – I presented on Test-Driven-Development?
During the phone interview, the program chair of the conference said to me something like “Well, yes, we’ve been saying for years that the non-technical tester is going to go away. This might be the final nail in the coffin.” Around the same time, I started to hear this term SDET — the tester who writes test-code. I’ve been grilled me on interviews that the SDET is the future of the craft, and several of my friends tell me they are learning Python or Ruby, just to get and keep their jobs. The argument is that with advent of the SDET the traditional tester will “go away.”
I don’t think so. — At least not completely. And I’ll explain why I feel that way.
First of all, there are technology adoption curve issues. Consider Shockwave, Flash, Silverlight, The Web, and AJAX – each of these new technologies were very powerful, allowing developers to do things in a few lines of code that previously might take days to program, or might not be possible at all. (Look at Google Maps. Could you even /do/ that before AJAX? If you could, why didn’t anyone?)
At the same time, for each of these technologies, the test automation was an afterthought. And in today’s race-to-be-first, it just didn’t make a ton of sense to spend time on automation; it makes sense to get version 1.0 out the door in order to make money right now.
This idea — that product innovations general push process maturity back, and that is okay, is not my idea. It’s not a new, either – you can find it in the industry maturity writings of Michael Porter. Porter is the author of Competitive Strategies and a Harvard Business School Professor. Porter’s claim is that industries move from innovation to growth to standardization and decline. On the “innovation” and “growth” stages of his model, new product development looks like a sort of, well … Creative Chaos.
The neat thing about technology is that we can push back the model. For example, retail delivery of books was well standardized by Walden Books, Barnes and Nobles, and Borders, until Amazon.com came and re-invented the whole business.
You can look at what Amazon.com did and say it was not “mature” — yet they acted like a typical technology start-up. Look at all most successful companies in their startup days – Slashdot, Yahoo, Google, Borland, DEC — and you’ll see a pattern. For that matter, a documented pattern.
As long as we have kids in their dorm rooms inventing new technologies that need testers, we’ll need non-SDET testers. And while Napster is the classic example of this, Facebook is quickly becoming an even better one.
What does Technical Vs. Non-Technical Even Mean?
Oh, yes, I’ve worked with the clueless project manager that can parrot back words like “Application Server” and “https over SSL” but doesn’t really know what they mean. The guy you might refer to as “non-technical” and actually mean it. But that’s an exception. After years of meeting testers and asking how they fell into the role, I’ve noticed some similarities: Curiosity, an interest in how things work, and a fair amount of technical aptitude. As computers and handheld devices go everywhere, you don’t need to teach testers how to use the web or download Firefox; they can figure it out.
We typically consider those people “non-technical”, but if they can figure it out, does it really matter? Let me give you an example.
Last month, I talked to a development manager who wanted to hire a lead tester for a cloud-based web-hosting product. The general idea was that people would pay him a flat fee, and they’d provision the servers from their own data center, and, if they run out of space, get them from Amazon, with no difference to the customer. (Provision is a fancy word for “request and assign.” Throw it around to sound “technical.”)
The manager’s fear was that he needed someone ‘technical’ to lead the test effort; I got the impression he wanted a four year degree in computer science.
But if this tool was built with a high level of abstraction, what would the testers need to do? Well, they’d need to learn how to telnet into a machine (that’s a webby term for “login” to a text-based remote machine), then run some command-line/text-based utilities, maybe write some utilities on their own.
That is very similar to what a whole lot of testers did for fun in high school in MS-DOS; running programs from the command line and making batch files.
It’s the kind of thing that curious people can figure out.
My guess is that the build infrastructure was complex and did not have clear seams – so to test it might require a bunch of different commands, some mocking, and maybe some coding. In that case, while the company could hire an SDET to write mocks and stubs and code, it might benefit them more to hire a less-technical tester who can step back and say “this app has some usability issues. It needs better seams, more little utilities, more helper functions.”
Saying things like that would result in a more usable problem, which adds value. And once it had those seams, the less-technical tester could get to work and add a lot of value quickly.
Bottom line: In an era where the typical 13-year-old knows more about Texting than I do, I see the line between technical and non-technical becoming very blurred, moving to “more technical” or “less.” And it turns out, having your testers be a little less technical than the developers can be a strategic advantage for usability.
But it works for Google and Microsoft
Another friend of mine pointed out that lots of companies want to “Test like Google” or “Test Like Microsoft”, when they aren’t in the same place as Google or Microsoft. So, while it might make sense for Microsoft to invest 100 person-years in a specialized test automation framework for it’s testers that improves their efficiency by 10% (spread over 5,000 testers it’ll pay for itself in a few months) — that might not make sense for a company with a technical staff of twenty people, only four of which are testers
In that situation the right approach might be to use a mix of strategies, viewing test infrastructure as a business value for the product owner. Thus test infrastructure is an investment that helps shrink the duration of a test cycle. The product owner can then take building that infrastructure as a ’story’ or requirement, comparing its costs and benefits against any other ’story’ or requirement. The test team can take whatever infrastructure is provided, then balance it’s time among a variety of risks.
We may have to be a little more tech-savvy, we may want to make smart use of tools, and our dev-to-test ratio may be a little higher, but going away? Really?
To borrow from Mark Twain, I’d say the death of the traditional tester has been greatly exaggerated.
If anything, I’d go the other way – the need for testers today is greater than ever before.
Testers Analyze Systematic Risk
Have you ever tried to call a company about an account with a routine question, and spent an hour or two on the phone, only to walk away unsatisfied? I submit that this happens enough that companies should see this as a real risk; I’ve walked away from several companies because of bad call-center experiences.
Somebody, somewhere, should have been looking out for risk — and they weren’t.
This is a systemic problem in our society; it impacts everything from Enron (they built a power-plant in an area of India that could not afford to pay for the electricity), to BP Oil’s mess in the Gulf of Mexico, to the May 6th bug in the US Stock Market. It was a big bug, folks; something went wrong and a few key stocks, such as Proctor and Gamble and Accenture, went to trading for a penny a share. When that happened, other companies had stock market ‘triggers’ that created massive sell orders, and the stock market lost 10% of it’s value in a period of fifteen minutes. Then there is Toyota.
What strikes me about these recent crises is the lack of investigation, research, root-cause-analysis, and prevention. Reading the Wall Street Journal, the general response seems to be a shrug of shoulders, a blaming of a “glitch” and some “strange timing of orders.”
No, there was a bug. The bug needs to be fixed. Next time, it needs to be prevented.
Which tells me that we not only need more testers, but we need tester skills in decision-making roles. It tells me that the next logical step for career progression for any individual tester might not be development or project management, but instead general management or the executive suite.
And that, my friends, is something I can live with.
Epilogue
My friend, Jon Hagar, read an early draft of this article and replied “The future is so bright, I gotta wear shades.”
I couldn’t agree more.
Categories: Blogs
Performance Testing Metrics – and Bolton too!
SearchSoftwareQuality.com just published an article I wrote with Oliver Erlewein on Performance Test Metrics; the article is available on-line right now – a free registration is required.
In other news, for the August Edition of “Ask the Tester”, I plan in interviewing Michael Bolton. No, not the singer, nor the guy from office space. Michael Bolton the Consulting Software Tester and co-author of the “Rapid Software Testing” course.
Of course, “Ask the Tester” is a little bit different than a traditional interview, in that we ask /your/ questions. To do that, well, I need some help – I need questions, along with your name, city and state (or province and country).
So please allow me to prime the pump a little. Michael was a tester, then program manager at Quarterdeck, a software firm that made a little piece of software called QeMM-386. You may recall the Bill Gates famously quipped that no one would ever need more than 640K of memory – well, QEMM was the program that allowed people to access over 640K in MS-DOS, and became, in its heyday, a very popular application for MS-DOS.
You might actually say the best-selling piece of software, at least at a few points in time according to the charts at PC Magazine. That’s the best selling software, not among operating-system plugins, not among software utilities, but among everything you could buy at the story -including wordperfect, lotus 1-2-3, wordstar and other office applications. That’s how Michael cut his teeth.
In 1998 Michael went independent, launching his company, DevelopSense, offering testing and test-advisory services. Starting with more of a focus on doing testing, In 2003 Michael started teaching Rapid Software Testing, and has spent the past few years focusing on training and consulting.
Personally, I know Michael best for looking at simplistic solutions to problems of testing, breaking them down, pointing out what they miss, then re-assembling a more complex, but also better, approach to the problem.
He’s a world-class consultant that has offered up his time to answer your questions for free, and get your 15 seconds of fame by having your name in a test magazine.
Please list your questions here, and I’ll pass them on. Or you can email me direct at matt.heusser@gmail.com. We look forward to hearing from you!
Categories: Blogs
Interview with a Creamer – and two podcasts!
So first of all, the folk at uTest have interviewed Lanette Creamer, and yes, The Rebel Alliance came up; you can check out part one and part two of the interview now; the third should appear later in the week. UPDATE: and here it is!)
Meanwhile, Back at the Ranch
Yvette Francino over at SearchSoftwareQuality asked me to participate in a four-part series of podcast interviews; they’ve allready placed the first two, a discussion of The Selenium Test Framework and a second on open source test management up on the web. (Look for “Podcast Link” for a link to the direct MP3’s)
I’m really excited about these podcasts; I think the idea has legs. Personally, I would really like to see the audio interviews continue, see the audio quality get better, and see us start to cover the topics you ask for and invite the guests you suggest.
To do that, we’ve got to get the series renewed.
So I’m going to ask for something. Oh, no, it won’t cost you any money. No, I don’t need you to email anyone. No, there is no grassroots campaign to manage. All I’m asking if you for you to do me the favor of listening to the podcasts. If the ideas are good, blog about them, tweet about them, tell your friends. Then let me know what you’d like to hear covered on a future podcast, and who you’d like to hear from.
I try not to ask for much on this blog; you already honor me with your time by reading it. But if you do invest a little time on the podcasts, well … thank you. I’ll work hard to see that it’s time well spent.
Categories: Blogs
The Long, Dark Tea-Time Of the Soul
So, when should a test be automated?
It turns out this is a non-trivial question, and I spent a fair amount of time working on an answer.
Now I think most readers of this blog respect that it’s not just about one-way ideas; it’s give and take. That in addition to publishing ideas, the blog is also a place to collect and develop ideas.
Yet you also want answers that are reasonably well-thought out and developed.
So I started writing on the subject. I wanted to develop something good; something tight, comprehensive, and helpful. But those three ideals conflict with each other – to be comprehensive, I have to make all of my arguments, so I can’t write tightly.
Likewise, to give specific advice, I need to drill down deep (helpful examples), but those will only apply to a limited set of people, so to be comprehensive, I’d need lots of examples … and all of a sudden, the writing isn’t tight.
Then I decided I might have to settle for (reasonably) tight, (reasonably) comprehensive, and (reasonably) helpful.
Then it became “pick any two.”
Folks, I’m not going to post that. I’m going to keep working on it, in the background.
Longtime “Testing At The Edge” Reader/Contributors will have a login to mattswiki – you can go get the first few hundred words of the essay there, and watch as it develops. Just click on ‘recent changes’ or search for “when testing through the GUI might be appropriate.”
If you are a long-time reader and commenter – someone I’d recognize – and you’d like review but don’t have a login to mattswiki, just email me: matt.heusser@gmail.com and ask for one.
Plenty more to come, but for now, I’m afraid “When should a test be automated?” needs to simmer a bit longer on the boiler. Thank you for your patience.
Categories: Blogs