It is sooooo slow at the office right now. Almost all the developers in my team are on a well deserved vacation, but that leaves us testers a bit bored. So when we finally got a fix to verify we went bananas with it. We did the usual routine; creating a playbook on a whiteboard and sorting out how things worked before and how it should work now. We went through the risk areas and settled on a couple of charters to run sessions on.









Then the fun started. Two big TV screens have been installed in one of the meeting rooms so we high jacked that room. We thought it would be a good idea to test run the TV also. So we got two Macs hooked up to a screen each and then we had some sessions the three of us. It was quite neat, everybody had their own computer but we had a good overview of what was happening in the logs and the environment since we could see the same things and discuss them. So my plan is to make the room into our new test lab. :)

Our new test lab:

More sunshine!

Ah, I can't believe we are having such lovely weather Stockholm! It just keeps being warm day after day and I'm loving it! It is even okay to be at work! Well, kind of. We compensate by eating a lot of ice cream. And since I don't have any vacation until September I just have to be positive about it. And still, I rather work and have the sun shining outside than rain... Any day of the week...

But, there is not so much to say about testing at the moment. I'm working on a presentation for SAST 2010 (Swedish Association for Software Testing) in October. Hopefully we will get started in looking at our processes and improve them during the summer, but as more and more people are going on vacations even that will be hard. So I guess I'll have to live with a slow pace the coming couple of weeks. That is quite alright with me :)

Enjoy the sun!

Summer time!

Finally, we are getting a bit of sun in the great capitol! It's about time I say. My objective during sunny days is to convince my colleagues that every meeting should be held outside down by the water and with ice cream. So far I managed to hijack one meeting this week. I'll hopefully get to more next week. Otherwise things are going into a nice slow summer rhythm. With so many going on vacation it's not so much to do at the office which is a nice change after a quite stressful spring with the opening of the France market. We went live with real money yesterday so now we have three sites. Or as my boss pointed out today, we go into a new market every summer. Two years ago it was Italy, last summer we went live with the new poker engine P5 and this summer France. I suppose next year will be Spain. And Denmark...

I think most of Europe will open up their markets in five years since they can not resist all the money they will make in taxes. And that is good, I am not a believer in monopolies. And the US market will follow as well. Even if it is a hassle with different governments having different requirements, based on the best promoting lobbyist it is better that we have regulated markets. Hopefully the swedes will get it some day as well, and knowing how much we looove our taxes I can't believe it will take them long to realize how much money they are loosing on keeping their monopoly.

So, go outside, dream about a world without monopolies and enjoy the sun and have an ice-cream!

India baby!

This sprint has finally come to an end. That's good, I'm ready for some swedish midsummer and nubbe. Most things have worked out quite well in the sprint, but my concern that we were too many sharing the testlog came true. It has been a bit messy, and nobody remembers what happened 2 weeks ago. AP (action point) for next sprint is that everybody has their own test log that can be merged into one test report at the end of the sprint.

Otherwise there is a new buzz in the office; bwin is discussing outsourcing to India! That is so cool! I attended a lunch meeting today where we discussed some of the changes that might take place and how we plan to handle them. I think this outsourcing will be great. It will be very interesting to work with these new circumstances, even if I won't be allowed to move to India. It will be challenging, but I'm game!

Have a great midsummer everybody!

It's getting messy

As I mentioned last week, in this sprint we have been working with teams within the team. It has been ok but it has also proven to create a new kind of problem. Because we are working on different features we test on different releases (not because we use the great system of branching, but because we ignore the releases that weren't meant for our specific feature.). The problem is that with every release we should update the changes in the testlog/playbook. After a while it is hard to know which risks are connected to which release. And right now I have no idea any more.

It would create a bit of overhead having different testlogs (I'm naming the document that we use to visualize risks/changes and ends up being the test report. Until it is a test report, it is a testlog.) And, if we have two different testlogs, do we merge them into one testreport at the end on the sprint? Or wait until we have done our sprint end testing and then merge everything together? Suggestion anybody?

Enough about that, today it is bwin Games summer party, so I'm off for some food and drinks and good times, see you next week!

Getting everybody on board

This sprint we are trying a new thing in order to strengthen the relationship between developers and QA in the team. We are doing teams in the team, i.e. we have teamed up in two smaller teams working with their own features. So far it is going quite well, it is much more easy to focus on a specific feature and not having to pick up every release and test it. It lessens the context switching which is something that we have suffered quite severely from.

Now we are sitting together, behind us is the whiteboard with the playbook from which we can discuss changes that might have affected other areas and if the risks have changed. Also we try to update our written playbook with every release. It's quite neat actually, we increase the area that needs more testing with 1, and in the end we can see if we are testing where the risks are. It is powerful, it is visual and there is no escape. The playbook tells us if we are testing the right things based on the risks. But see here comes the problem, which is the same as before; it doesn't feel like we have enough time to test all the changes. The developers are pretty fast in writing and changing the code, and we would have to work overtime to match all the changes. But, the upside is that we don't have to. Our responsibility is to be able to present what we have tested, and it is up to the company if they want to release it. They are presented the risk, in a far more obvious way then before.

I am so glad I don't have to make those decisions, if it was up to me nothing would be released ;)

But that is enough test talk for this week, I'm off to Riga to have a lovely weekend, I hope you do too!

The Joy of Visual Feedback

And we are done. Well, not done done as we say here at bwin Games, but more or less done with the testing of our new component TRS. Today I started working on collecting all the sessions I have written and checking them into Subversion. From there my colleague Per's fantastic php-thingy does it's magic and I can see how many minutes we spent on setup time and on testing time. All the data can then be moved to Excel where we have a nice cross-point diagram where you can see how much we have tested in different areas. Had we done a proper risk analysis during the sprint planning we could also have shown how much we have tested vs. how big the risk is. Unfortunately we didn't do this, but we have had discussions with dev about where to focus our attention, so we might be able to work with that somehow. I guess we learn as we go along. One important lesson I learned is that we need to keep the Playbook alive during the sprint and update it with changes so that we can migrate the risks correctly.

But it is so awarding to see all your testing being transformed into useful data! You can actually see that you have been doing something for the last three weeks! When we were running scripted test cases in regressions I had no confidence that I had tested enough. You somehow forgot what you did the first week, and it is only what you do during the last week in regression that shows for anything. Now we have a track record for the whole sprint.

So with this happy feeling I am going out in the sun for some ice cream. See you next week!

Three can too play nice

When I was a little girl there were almost always problems when I and two other friends would play together. Being just me and my bff there were no problems, but as soon as that third person came along something started going wrong. It almost always ended with our parents stepping in trying to help us resolve our world wars. I have realized this week that it is much easier playing three as an adult. And as a tester, it is great!

Me and my two colleagues have finally gotten our messy test environment back on its feet and have started running our tests on a new component. During the sprint planning we selected the charters that we needed to cover and created our test plan on the whiteboard. On Wednesday we finally got started! Aah, it was good being back in bug-hunting-territory! Feeling the rush of finding something peculiar, poking it, analyzing it. It such a thrill to find something wrong, and then nail the scenario.

Anyhow, after we covered the basic tests by running some recon-sessions we got to the more complex parts. We did a couple of sessions the three of us and it was really good. All of the sudden there were three pairs of eyes on the logs. We could discuss problems, strange behaviors and, with help of our different experiences of the system, do a more thorough testing.

So hopefully we can do as much multitesters-testing as possible the rest of the sprint, because it was a really good way testing, learning and having fun!

So until next week, happy multitester-testing! (trying to name it, but I am not really sure yet.)

Back to work

For six weeks now I have been working in our system team, preparing a release to go live. During this time I have spent three weeks integrating all of our components, trying to make them play nice with each other. The past three weeks we have been doing system verification, making sure that the release is fit to go live. It has been really fun to see how the other parts of our system work in a big live-like environment. From this week and forward I'm back in my old team, and I can finally start doing session based testing. My test team has come such a long way in these past six weeks! They are really starting to get the hang of it and it's going to be so much fun to take part in that kind of testing again. I almost feel like a rookie after being out of the loop for such a long time. Next week I will hopefully have a nice update of how it’s going, If we ever get that test environment of ours working again…

Until then, happy testing!

Moving on

For two years I worked with the same routines. I got a requirement and broke it down with our product manager and developers. I wrote some nice test cases, tested the thing and at the end of the sprint ran the test cases again. Voilá, tested and done! I got sooo bored with this. Last autumn I was fed up and felt that this kind of work procedure gave me nothing. Not a great feeling to have, especially since I love testing. Luckily for me, I got the chance to go to the Star West conference in Anaheim. There I got to listen to some of the most prominent testers in the world. And my, did my inspiration return! I came back to Stockholm with the feeling that QA is the best job in the world and now I could do anything!

I went to some seminars in Stockholm and came in close contact with Session Based Testing and I fell in love. Here was the answer to how I could structure my exploratory testing. We started to try it out and got the hang of it. After that we ran into some trouble. I don't know how many hours we spent on discussing what the Word document should look like, and a month ago we skipped it all together. One of my colleagues did a great job with some PHP-coding and now we use text documents that are saved into Excel where we can extract nice data such as time spend in different areas and bugs found in each sprint. All of the sudden we can show were we test, and how much we have tested it.

We are still experience some growing pains, but we are getting there. Even if sessions seem so easy to do, it's hard to do it well. It takes practice to learn how to write your notes so that they make sense in another context than when you are testing. But practice makes perfect, and in half a year I think I will be a session master.

That's all for now, see you next week!

Until then, happy testing!

Name it and own it again

This is one of the first posts on my bwin-blog:


When I first came to bwin Games I had a vague idea of what testing was, something about finding bugs. Now that is indeed true. I found the bugs, I did the testing. But as time went by, I started to feel that I couldn't describe my work properly. If somebody asked me - What did you do yesterday? I would say,
-Uhm, well I tested this and that and I found five bugs.
- So what kind of testing did you do?
- Ehh... exploratory? I clicked on a bunch of stuff and then they broke.

I did testing but I had no idea of what kind of testing. I had no words to describe my work. I am certified by ISTQB's. I had a bunch of testing terms in my head for a couple of days. If somebody would ask me today what equivalence partition coverage is, they would get a blank stare. The term means nothing to me. Maybe it should, I took the course. I know we need a common language, but in my opinion that language is useless, unless everyone can speak it and understand it. Using the terms of ISTQB is like having Latin as the common language of testing. Nobody speaks it, and people only know a few quotes. This hardly helps us communicate with each other, neither developers nor management.

So there is a dilemma. We have a vocabulary that we don’t understand, and still we have to be able to communicate what we are doing, something more specific than “um,.. exploratory testing”. With inspiration from James Bach I have tried to take control over my language and my work. I have started to reflect on what I do and write it down. If I have to make up my own name for it, I do. In that way, the language becomes my own. If somebody asks me what I have done I can tell them with my own words. If someone doesn’t understand, it is easy for me to explain what I did and I don’t have to get the feeling: “Sure, I have done a great job but I don’t have a fancy word to describe it with”.

So, take control of your testing, name it and own it!

See you next week, until then, happy testing.

(Inspired by James Bach)

Oops, I did it again

I'm sorry, I did it again. Kind of stopped blogging. One reason being that I started blogging for bwin as well. So my plan now is to move all the posts from bwin to here. Enjoy!