This is a bit of personal musing.  Feel free to skip.

Originally I was going to tweet this but then I started coming up with quite a few tweets and decided a short blog would be better.  There is such a thing among testers on twitter as tweeting your own horn a bit too much.

I have a test talk to give tomorrow.  It’s an internal Microsoft talk to a bunch of senior Test Manager and Architects.  It’s for this group called QTEC that we started about two or so years ago with the help of Grant George, Alan Page, JW, and a few other folks.  It’s a cool group with some really great talks.  It will be my first time presenting to the group but as I know most of the members quite well I feel like it should be familiar turf. 

The talk is to be me presenting how to test services or is it S+S or Client and Cloud.  My thinking on the subject has evolved quite a bit since I wrote on the topic in, "How We Test Software at Microsoft."  but that thinkin has evloved because of my interactions with other great, smart testers.  @Setheliot is top notch on experimentation.  @venkatna has taught me quite a bit in his talks about how they tested Bing.  I'd mention more but they don't tweet or blog that I'm aware of.  My conern is that I have just picked others brains and so now I get to stand in front of a crowd and pontificate (I love that word almost as much a penultimate).  There must be more and I think that’s why I’m feeling feisty.

 

I wouldn’t be a good tester if I didn’t have opinions and if I didn’t then test those opinions out.  I mean I don’t want to get as feisty as @Jamesmarcusbach and try to take on anyone that has a differing opinion on exploratory testing (apparently now referred to as ET but not the type that goes home).  I don’t want to get into the kind of back and forth as chronicled by Seth Eliot in his great blog post, “Your Software Has Bugs,” in which he chronicles the debate on Code Coverage by a few more public testing luminaries. 

I’m feeling feisty though.

I do belive that I will spice things up a bit.  I really do strongly believe certain things about services testing and Microsoft’s approach to it.  I believe we have a great chance to beat our major competitors and that test can be a key asset in helping us move our company forward.  I believe quite a few things about services testing.

1.       The traditional on premises test lab is dead and we should move all equipment to data centers and VMs

2.       All load testing should be done in production and if you’re a v1 service just order production hardware or VMs 6 months early and conduct your load tests on that hardware.  Buying test equipment for load testing is a waste.

3.       Ship into production more often and finish your testing there.  Test lab results are good for unit and basic functional testing but integration testing needs to happen in production.

4.       The OneBox is the key to agility for services testing

5.       Emulation is the second key to agility for services testing

6.       Test automation needs to run against production

7.       Deployment break = build break

If you want a regular dose of feisty check out I. M. Writes blog

 

Yep, I’m feeling feisty.

 I think that’s got to be the why me part.  I’ve got red hair and at times a hot temper.  I have opinions and I don’t mind ruffling a few feathers when the mood strikes.  I don’t quite know what I’m going to say tomorrow.  I’ve got some time to work on my slides.  If Alan Page reads this post before tomorrow he’s going to be nervous.

Related Blog posts

·         TIP-ING SERVICES TESTING BLOG #1: THE EXECUTIVE SUMMARY

·         1+1 Redundancy Just Isn't Good Enough

·         The Demise of the Test Lab