- This topic has 68 replies, 18 voices, and was last updated 10 years, 2 months ago by
.
-
Posts
-
Because I changed job recently and the website (like many other websites) is not up to date. we are all humans. Linked in is correct.
Good question. The broader the understanding of test and the test techniques are the better quality of testing we should get. If you know what you do and you train it, the ‘product’ is usually of a better quality than if you don’t. ISO 29119 is one way to get a better understanding of what test might be.
AMThings are usually not know until they are dicovered; that goes for anything in the world.
The principle of gradually breaking down what you know (as the presentation shows one way of doing) will inspire you to what you can test and also give you in insight to what you don’t know yet. That principle – in my vue – applies to all the way of testing and developing you mention, and probably to many more.
AMSince this forum doesn’t seem to offer threading. (If it does I can’t find the button). The answers to questions are completely out of context and I don’t know which question they relate to.
Alan, Good point.
@Anne-Mette : please mention to whom (or which question) you are responding?If you get the impression that the activities are rigid, I’m afraid you have not heard everything I said. I said, several times, that this is not about doing things in a specific order, or making documentation, it is simply a way to get from the test item to ideas about how you can test it. If you have a user story all this might take 5-10 minutes depending on the user story. Try to get the perspective right here, please.
AMHi Alan
We have asked Anne Mette to start copying and pasting each question she is responding to when answering so it will make it clearer who she is responding to.
Thanks,
EmmaThis reminds me of a research that was done in the US recently (I think). The researches got some university students to participate and they placed them behind a starting line with a goal line some distance away. Between the start line and the goal line there was two buckets placed for each participant. The participants were asked to move one of the buckets from its position to behind the goal line with the use of a little effort as possible. Most of the students picked up the bucket closed to them selves, and hence the longest way from the goal line, and moved that – causing more use of effort than if they had taken the buckets furthest away. Asked why, they said that they wanted to get started as soon as possible and not waste time on walking to the furthest bucket.
AM@IlariHenrik Aegerter Just to address that issue about Iain being unable to access the discussion. No one is excluded from the chat here. There may have been an access issue. I have contacted Iain to see what that might have been.
Hello Anne –
I read your comment about not wanting to question 29119 in this forum, ok. But I do want to ask (and I believe this is one point to the discussion), can you share:
1. How does 29119 help you? I did not see or hear a direct correlation to 29119. Perhaps the correlation is implicit to you and that is understandable if you work with it and are involved in its foundation and formation – but I am not – I would like you to draw an explicit correlation to your test process as outlined and how 29119 helps guide the process.
thank you,
Karen N. Johnsonalan
Does ISO 29119 say “never ever use the code as a test basis?”I do use the code to inform my testing. A reading of the code can make me think of risks that I then use to test e.g. this code looks inefficient I can use tests to explore this, and this code looks like it block other code so perhaps multiple users aren’t supported, or this code doesn’t look thread safe so might not handle multiple users etc.
Why exclude some valuable sources of information from your testing?
AM:
I’m afraid I was not quite clear on this. Surely you can and should use the code to inform you if you are doing structure based testing. This is your inspiration for input to the test cases. But you should not use the code to inform you on the expected result. That’s what I ment to say – sorry about the confusion.Tim Western
Isn’t static testing, generally more of a Developer concern? Static Testing is often done in the form of peer or formal code review, unit testing, and static analysis by a host of tools that can be run in your CI environment with each build? Why is this a concern of testers?AM: This is matter of opinion, I guess. Many testers are also involved in static testing, using their skills for finding defects in the process. I you take the definitions of testing, I think you can safely include static testing. But static testing is not covered by ISO 29119, for the reasons you mention.
Tim Western
Isn’t static testing, generally more of a Developer concern? Static Testing is often done in the form of peer or formal code review, unit testing, and static analysis by a host of tools that can be run in your CI environment with each build? Why is this a concern of testers?AM: Paul Gerrads model is not included in 29119. I’m Just including another persons view because I find it valuable; I dont care if it comes from a standard or not.
Peter
a feature set is a logical subset of the test item => What is a logical subset?AM:
It is a feature set. No, seriously, it is a part of what your are going to test, that seems like a reasonable coherent and isolate part. I’m afraid I can’t get any closer. May the example can cast some light on it?Tim Western
How many test cases? Why is that even a valid metric? It’s been debunked so many times I would have thought we as a community had moved on from it.AM:
Only how many test cases do you execute out of possible number of test case (i.e. coverage) is valid.Tim Western
last I checked, a lot if not all of these ‘coverage’ metrics you’ve mentioned can be garnered from CI environments automatically using tools. Why is this a testing concern? Why is this part of the ISO 29119 Standard?AM:
I don’t know what ‘garnered’ and CI environments mean, but surely coverage is a testing concern, since it, to my knowledge, is the only measure we have to the thoroughness of test. But if you know others, I’ll love to hear about them.Laurent Bossavit
Regarding the “test condition” nomenclature, what’s the difference between function, feature, transaction etc?
AM:
Good question. I’m afraid my English is not good enough to answer that question, but I have a feeling that they are more or less the same and more a question of preference.Laurent Bossavit
More generally, why does the standard introduce a substantial amount of terminology – “test basis”, “test condition”, “test converage item” – in addition to the more familiar “test case” – which, as Anne-Mette noted during the Webinar, is likely to be unfamiliar to most people (and so, arguably, “non standard”)?What is the added value of this complex ontology and terminology?
AM:
Many of the terms you mention are not ‘invented’ by ISO 29119, but used in others contexts as well. The reason for using a comprehensive terminology is to be as precise as possible about what you are talking about.Tim Western
Why is repeatability of a test important? Won’t this leave gaps in testing, because you reduce variation of inputs?AM:
Repeatability is important for retesting and regression. If you can’t repeat the test that reveiled a defect you do’t know if it has been corrected, and if you can’t repeat a test that worked before you don’t know if it still works after defect correct another place in the test item.Peter
Seriously, Do we need a webinar for this? => From the Testers at my clientAM:
There are so many things in this world we don’t need, but that others might find useful.Laurent Bossavit
Early on in the webinar, it was asserted that it was not mandatory to record in writing the various information items produced.This statement seems at odds with the repeated emphasis, including at a very fine level of detail in the dynamic test process, on having the various information items “agreed by the stakeholders”.
Why is the standard so insistent on having “sign off” on things like the “test design specification”, which should be no one’s business but the tester’s? Why is the standard simultaneously so vague on defining who is referred to as “the stakeholders” in these various sections?
AM:
It is not every tiny bit of you work that the standard suggest that you should get stakeholder agree on – as far as I remember, it is the final test specification, i.e. the test procedures. An even those may be agreed on only by discussing them, though that is perhaps not so easy.Wrt. stakeholder stakeholders are defined as anybody who will be affected or affect the system (from my memory). The standard tries to be context driven, by not stating who the stakeholdes are, but making it depend on the context.
Peter
Seriously, Do we need a webinar for this? => From the Testers at my client– See more at: https://huddle.eurostarsoftwaretesting.com/forums/topic/the-core-of-testing-webinar/#sthash.dzMy3VxF.dpuf
AM:
There are so many things in this world we don’t need, but that others might find useful.– See more at: https://huddle.eurostarsoftwaretesting.com/forums/topic/the-core-of-testing-webinar/page/3/#sthash.OGjmohV8.dpuf
This was a serious reaction of four testers who take their craft seriously and saw a webinar that was “explaining” how to test in a certain way. Something we already do for years. We did not see something new, something innovating. What is ISO29119 bringing to us?
Silly question. How can I join the webinar? I have not got any link to it after registering….
O.K. Just found it here: https://huddle.eurostarsoftwaretesting.com/resource/the-core-of-testing-dynamic-testing-process-according-to-iso-29119/
It’s kind of webinar as recording? I see…Will Ms Hass answer any of the ISST board member questions, or continue to systematically ignore them?
Tim Western
last I checked, a lot if not all of these ‘coverage’ metrics you’ve mentioned can be garnered from CI environments automatically using tools. Why is this a testing concern? Why is this part of the ISO 29119 Standard?AM:
I don’t know what ‘garnered’ and CI environments mean, but surely coverage is a testing concern, since it, to my knowledge, is the only measure we have to the thoroughness of test. But if you know others, I’ll love to hear about them.– See more at: https://huddle.eurostarsoftwaretesting.com/forums/topic/the-core-of-testing-webinar/page/3/#sthash.ALFX7jEy.dpuf
I can see i wasn’t very clear. I was referring to CI = Continuous Integration environments, which are the standard in the industry now. Many of them have the ability to provide code coverage guidance, where there is automated checks going on (such as with unit tests, and so forth) I agree that coverage is a ‘testing’ concern, I think I misunderstood what you were saying about the importance of coverage, which IMO is only important in that it can help you focus what you test. If you have non automated testing going on, that testing can be enhanced by knowing the coverage model provided by what the developers have already written.
So, I realize now, that there’s two different ways of describing coverage, that for what’s covered by test ‘code’, and what individuals have to perform. I apologize for the ambiguity.
- You must be logged in to reply to this topic.