SPONSORS:






User Tag List

Thanks Thanks:  0
Likes Likes:  0
Dislikes Dislikes:  0
Page 1 of 2 12 LastLast
Results 1 to 10 of 11
  1. #1
    Junior Member
    Join Date
    Jun 2002
    Location
    Irvine, CA, US
    Posts
    25
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Can Developer be a QA

    Can developer be a QA, here is my delima, I have been doing Software Quality Assurance for 5 years, and I just took this job and In my QA group I have several developers trying to do QA, my project is a Data Warehouseing, all of the people are old timers developer, and they don't even know the definations of typical Quality Assurance and and i have been asked to run this group, and it is very diffcult to covince them on any QA process, for example they want to develop a automation tool to test the back-end ETL process, which I think is a manual process and since they are developers they disagree and having havey risistance to domanual testing, so I am looking for some documents that describe the QA standards and processes. I don't want to be disrespectfull to any one but it just a thought proces is totally diferent of a developer from a QA process, I have no problem with automation, I know Rational, Win Runner, Silk, but they don't apply to my back-end testing of the application.
    Thanks
    Neil

    ------------------
    Neil Chaudhry
    Cornered Tigers

  2. #2
    Moderator Joe Strazzere's Avatar
    Join Date
    May 2000
    Location
    USA
    Posts
    13,170
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    1 Thread(s)
    Total Downloaded
    0

    Re: Can Developer be a QA

    I'm not sure what your question might be, but to answer the title of your post - Yes. The right developer can learn to be an effective QAer.

    ------------------
    - Joe (strazzerjATaolDOTcom)
    Joe Strazzere
    Visit my website: AllThingsQuality.com to learn more about quality, testing, and QA!

  3. #3
    Senior Member
    Join Date
    Oct 2002
    Posts
    259
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Can Developer be a QA

    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>Originally posted by mugarmuch:
    Can developer be a QA, ...<HR></BLOCKQUOTE>

    Yes. I am a developer in various languages and I still consider myself to be relatively good at various aspects of quality assurance and/or testing.

    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>...all of the people are old timers developer, and they don't even know the definations of typical Quality Assurance and and i have been asked to run this group, and it is very diffcult to covince them on any QA process,...<HR></BLOCKQUOTE>

    You will probably have to start with a very graduated type of "QA process". Focus on the tactical effort initially. We would need to know more about why it is so difficult to convince them of anything. Is this, in your mind, because they are "old timers"? Or is this because they see no value in the processes that have been laid out? Or is it because they already follow certain processes and they are reluctant to give them up?

    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>...for example they want to develop a automation tool to test the back-end ETL process, which I think is a manual process and since they are developers they disagree and having havey risistance to domanual testing, ...<HR></BLOCKQUOTE>

    You say you "think" it is a manual process. Perhaps the developers might know differently. Or they might have a way to automate what is a tedious manual process. You will want to talk with them about this rather than just disagreeing outright. That said, manual testing always plays a part and should be encouraged. But focused tactical automation efforts can also save you a lot of time. Find out about this process that the developers want to automate and see how workable that strategy might be.

    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>I don't want to be disrespectfull to any one but it just a thought proces is totally diferent of a developer from a QA process, ...<HR></BLOCKQUOTE>

    In some ways there are differences and in some ways there are a lot of similarities, if we, for a moment, speak strictly of "testing" when we say "QA process". (Of course, it largely depends on the person as well and the attitudes they bring to their respective disciplines.) There is a lot more to a general "QA process" than just reactive types of manual or automated testing. It is not a truism that the "thought process" of a developer is not amenable to a QA process. However, sometimes you have to look at how both groups deal with things tactically in their day-to-day tasks. There are different frames of reference (viewpoints, if you prefer) that both types of people (testers and developers) bring to the overall process of software engineering. Yet, even with differing viewpoints, in some cases, that does not mean you cannot establish an effective process that harnesses the abilities of both types of people.

    As it is, however, we know way too little about your situation to do more than comment in a very general way on what you might be facing.

    ------------------

  4. #4
    Junior Member
    Join Date
    Jun 2002
    Location
    Irvine, CA, US
    Posts
    25
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Can Developer be a QA

    Thanks for all your thoughts guys, like i said I don't disrepect any developer. I have tried giving them a test plan and test strategy, which is a true QA process and what and how i think we should be doing testing, I believe in "you have to crawel before you walk and you have to walk before you run" but I get a lot of resistance about testing the ETL process manually, and to answer your question my title = QA Manager In this project we and migrating a lot of different source feed data throught a ETL mapping into new Oracle database, which means i have to test all the business rules for ETL like every field and every table how it is mapped, and i understand we can automate some back-end process like row count etc... but this QA group have set thier mind on developing a tool ( which they have no idea that will be written in which language) that will minuplate the data coming in, to test the business rules, and inorder to minuplate the data for example source feed = Username, my incomming data is very large so this tool DEV/QA suggests will go and change all the records of user name, my question why would you do that when you can change it only one field and get the results, and ontop of it the suggest we should keep a copy of all the test data every time they make these changes for regression purposes, which means, i have to create a huge data pool for all the fields that we need to change the data for, and i have suggested that there is no need to go and change all the 1 million records for user name with the exact same value, because every time i will run this ETL if change all the records it will take a lot of time to run and complete just one test also My ETL error handeling will stop the process on the first error so why should i change all the records, because it will generate a huge Error log showing all the Data change vales and insted i can do it manually and be much more faster. So my problem is they have already set their mind on this and not considering that it will be time consuming Now you all now more delima, so what i am looking for is some document that I can share with them about QA process. I understand that a developer can be a good QA but I also believe that in order to become a good QA a person should consider learning a new process, I don't say that they are not good developers or good people I am sure they are they have been doing development for 10-15 years.

    ------------------
    Neil Chaudhry
    Cornered Tigers

  5. #5
    Senior Member
    Join Date
    Oct 2002
    Posts
    259
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Can Developer be a QA

    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>Originally posted by mugarmuch:
    I have tried giving them a test plan and test strategy, which is a true QA process and what and how i think we should be doing testing, ...<HR></BLOCKQUOTE>

    Well, that all depends on the details of the "test plan" and the "test strategy". I have seen people overdo these things initially rather than, as you say, "crawling before you walk and walking before you run". As you say, since you are the QA Manager and you are there to establish some process, you should certainly go foward. But recognize when you might need to take graduated steps.

    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>...I get a lot of resistance about testing the ETL process manually, ...<HR></BLOCKQUOTE>

    Okay. But we have no details of the "ETL process" and thus we really cannot comment on whether it is best run manually, automated, or a combination of the two. Given what you state regarding mappings and business rules, there is certainly a type of automation that could be done there combined with some manual testing as well. However, then you state:

    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>... but this QA group have set thier mind on developing a tool ( which they have no idea that will be written in which language) that will minuplate the data coming in, to test the business rules, and inorder to minuplate the data ...<HR></BLOCKQUOTE>

    Well, perhaps that is ambitious. Perhaps it is a good way to go. Perhaps a good approach for you is to have the developers write up a sort of business case for developing this kind of tool. I am not saying a completely formal document, necessarily. Just a document that makes them write out the exact details of the tool and a list of the benefits. Make sure you ask them operational questions like: Will this improve testing? If so, how? Will this cut down on time that we take to test? How will we be sure that the tool is not covering up errors? Sometimes by forcing people to answer questions like this, they start to think about their own ideas a little more. It will also help you better understand what you are criticizing.

    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>... my question why would you do that when you can change it only one field and get the results, ...<HR></BLOCKQUOTE>

    Have you asked them this operational question? If so, did they give you a good response? Ask them why they want to do it the way they suggested. Perhaps they have knowledge of some part that you do not. On the other hand, perhaps you, in fact, are looking at it a little better because you are not so close to it and thus you can provide a better overall picture. I know you have provided some details but, to be honest, I am having trouble merging everything you say into a coherent picture in my mind of how your system works, except at a general level.

    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>So my problem is they have already set their mind on this and not considering that it will be time consuming<HR></BLOCKQUOTE>

    Well, I have no idea if it will or it will not. It sounds like you are convinced it will. They are convinced it will not. You have to determine some way to meet each other in the middle on this. Is it possible for them to do a proof-of-concept, assuming that does not adversely effect any time constraints you have? Could they give you specific details on why they feel it will not take a lot of time to execute these scripts?

    The key here is that both of you are convinced of the "rightness" of your propositions. If that is the case, you should be able to sit down together, present your points, and have a feedback conversation to try and see the viewpoint of the other person. If there are decisive arguments for or against this approach, it should not be difficult to get them out in the open for examination.

    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>...so what i am looking for is some document that I can share with them about QA process.<HR></BLOCKQUOTE>

    As a QA Manager, your job might be to put that together yourself in such a way that it is relevant for your company. In other words, a general "QA process" from someone else might not have direct relevance to your group and it certainly will not necessarily convince them of anything. First you should probably determine what level of focus your "QA process" will be at. Are you going to start from the requirements stage on down? Or are you more going to deal with reactive testing elements at first and then broaden out from there? Remember to consider your audience. You want to persuade them with this "QA process" document. So ask yourself: what will persuade them? To know that, you have to know their current conceptions (and misconceptions) regarding QA and testing in general. Then you can speak to those specific points of concern in your document.

    You also may want to consider talking with them about "QA process" before just shoving a document in their face and asking them to read it. Get them together as a group and ask them about some areas of concern they have with the process of the company. Then, when you know what they see as problems, you can say something like: "Hmmm. Well, that is interesting. Perhaps we could add something to make that go a little smoother." You might even avoid the use of the term "QA process" initially because a lot of people cringe when they hear that. Try to keep it tactical and practical, at least at first, so that everyone can see the actual benefits of doing what you are suggesting.

    A good starting point to get them to see process details might be this automated solution they want to work on. Ask them to come up with a very brief 'justification statement' for it. Ask them to estimate the time it will take to develop this tool. Ask them how you (and they) can be sure that the tool is covering what needs to be covered and not hiding (or simply not finding) various types of errors. Forcing them to think through questions like these will start getting them to think about process in general. But remember not to lay it on too thick initially. You mention these are "old timers" and, as such, you have to consider their mentalities and their approach to problem-solving. But also let them get a glimpse into your mentality and your approaches to problem-solving. You may find more similarities than differences. That is a crucial thing to do when you are balancing the needs of practice and process.

    ------------------

  6. #6
    Senior Member
    Join Date
    Nov 2002
    Location
    Columbus, OH, USA
    Posts
    245
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Can Developer be a QA

    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>Originally posted by mugarmuch:
    Can developer be a QA, here is my delima, I have been doing Software Quality Assurance for 5 years, and I just took this job and In my QA group I have several developers trying to do QA, my project is a Data Warehouseing, all of the people are old timers developer, and they don't even know the definations of typical Quality Assurance and and i have been asked to run this group, and it is very diffcult to covince them on any QA process, for example they want to develop a automation tool to test the back-end ETL process, which I think is a manual process and since they are developers they disagree and having havey risistance to domanual testing, so I am looking for some documents that describe the QA standards and processes. <HR></BLOCKQUOTE>

    Did anyone else run out of breath just reading this question? No offense.

    I think that you will find many members of this board are ex-developers. I think the Crypto made some good remarks about finding out what it is they want to automate and using tactical automation to same time. Tell me, are the only people on your QA team developers? Or do you have other "QA" people with you?

    One suggestion I have is not so start out the "old timers" on newbie/remedial stuff. If you are trying to get thier buy-in on QA processes, you need to go at them with a high-level of knowledge. For instance, don't start out with 'how to write a defect', start them out with the writing of a test strategy or test plan. This will give them the 1000 ft view of the process that these "old timers" are probably used to rather than the 10 ft view that a kid out of school would get.

    Just a thought, and good luck.

    ------------------
    Steve_Jones@SoftHome.net
    Always remember that you are unique. Just like everybody else.

  7. #7
    Moderator Joe Strazzere's Avatar
    Join Date
    May 2000
    Location
    USA
    Posts
    13,170
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    1 Thread(s)
    Total Downloaded
    0

    Re: Can Developer be a QA

    Neil,

    Cryptonomic said it well. It appears that you have a management problem here.

    You have a team of QAers who desire to test one way, and you as the QA Manager want them to test another way.

    That problem could just as easily have happened if the QAers in question had not been developers.

    Talk with your team. Explain how you want them to do their testing. Then hold them accountable for getting it done in the relevant time frame.

    Perhaps you need to train them, but just handing them a book or article doesn't sound like the best way to accomplish your goals.

    ------------------
    - Joe (strazzerjATaolDOTcom)
    Joe Strazzere
    Visit my website: AllThingsQuality.com to learn more about quality, testing, and QA!

  8. #8
    Senior Member
    Join Date
    Feb 2002
    Location
    OH, USA
    Posts
    548
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Can Developer be a QA

    I think there is too much weight attached to the word "Automate".
    All you need is a data generator tool, which you can get for less than $ 1000, a few shell scripts (for a bunch of requirements) which have SQL in them.
    I fail to understand why this approach cannot be combined with "Manual" testing. If there is a transformation layer with business rule before the data is transferred to Oracle, quite a bit of testing effort can be concentrated in this layer rather than spreading resources on "Automation".
    I am involved in a project where sources like tables, multifiles and flat files are used and data is transformed across 4 different layers and then staged into an Oracle database. We use a combination of data generator tools, shell scripts (with SQL) and tester's manual testing skills.
    As to your question of manipulating only one record, it may be a risky proposition as we are talking of millions of records. What if the application logic can handle a couple of loops and then fail ? I can think of 5 other reasons where this approach can be risky.
    I may be sounding oversimplistic- IT IS NOT, but what the heck get on with it. ETL testing is lot of fun.
    ------------------


    [This message has been edited by qasleuth (edited 03-26-2003).]

  9. #9
    Senior Member
    Join Date
    Feb 2002
    Location
    OH, USA
    Posts
    548
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Can Developer be a QA

    Apart from 'Datatech' you can also look into DataFactory - Quest and D-Generate from http://www.syntheticminds.com.
    If you do not want to use a data generation tool but use your production files (input files), then you can take a subset of records. Why do you want to include all the million records, take say head -1000 (i am assuming it is unix).
    have your input test data, transformation rule and expected output data in an excel spreadsheet and do your comparison (if manual testing), or use shell scripts if want to 'Automate' the process.

    ------------------

  10. #10
    Junior Member
    Join Date
    Jun 2002
    Location
    Irvine, CA, US
    Posts
    25
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)
    Total Downloaded
    0

    Re: Can Developer be a QA

    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>Originally posted by qasleuth:
    I think there is too much weight attached to the word "Automate".
    All you need is a data generator tool, which you can get for less than $ 1000, a few shell scripts (for a bunch of requirements) which have SQL in them.
    I fail to understand why this approach cannot be combined with "Manual" testing. If there is a transformation layer with business rule before the data is transferred to Oracle, quite a bit of testing effort can be concentrated in this layer rather than spreading resources on "Automation".
    I am involved in a project where sources like tables, multifiles and flat files are used and data is transformed across 4 different layers and then staged into an Oracle database. We use a combination of data generator tools, shell scripts (with SQL) and tester's manual testing skills.
    As to your question of manipulating only one record, it may be a risky proposition as we are talking of millions of records. What if the application logic can handle a couple of loops and then fail ? I can think of 5 other reasons where this approach can be risky.
    I may be sounding oversimplistic- IT IS NOT, but what the heck get on with it. ETL testing is lot of fun.
    <HR></BLOCKQUOTE>

    hey thanks for good advice, can you share with me the name of your data generation tool that you are using and a sample SQL shell script, one thing i would like to mention about my ETL process is the reason i just want to change only one record at a time is because since it's still in deve;lopment phase and my ETL will stop at the first error and will generate a log file of that error so i am sending 1 million record change and it stops at the first one wt's the poit of changing all the records in the database. And your approach is right for testing that kind of application, my application is simillar to what you have, I think what i need is a shell script in SQL which will allow me to go in and run a query on certin table and filed and change that value and excute it, i think that will fix some of my issues and will make my testers and me happy. My current time to run 1 ETL with 1 million record changes is hours not minutes, so you can see why i am not infavor of changing that many record.

    Thanks,
    neil


    ------------------
    Neil Chaudhry
    Cornered Tigers

 

 
Page 1 of 2 12 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
Search Engine Optimisation provided by DragonByte SEO v2.0.36 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 10.00%
vBulletin Optimisation provided by vB Optimise v2.6.4 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
User Alert System provided by Advanced User Tagging v3.2.8 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
vBNominate (Lite) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Feedback Buttons provided by Advanced Post Thanks / Like (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Username Changing provided by Username Change (Free) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
BetaSoft Inc.
Digital Point modules: Sphinx-based search
All times are GMT -8. The time now is 09:43 PM.

Copyright BetaSoft Inc.