June 9, 2011 11:22 pm

Rollback Strategy…

Filed under: Software Quality,Software Security,Software Testing — ramsblog @ 11:22 pm

Rollback testing is one of the hot topics in the software application deployment where you push the updates into production frequently or even if once in a while bug talking about number of applications and number of patches into production.

What is your Rollback Strategy?

How would you plan for one?

what is your understanding of application Rollback at deployment?

June 8, 2011 1:52 am

Nice story about Software quality – a Fable

Filed under: Software Quality,Software Testing — ramsblog @ 1:52 am


Fable about knowingly taking shortcuts and not adhering to the quality of the outcome…

read the rest and enjoy the rhubarb cake…

February 11, 2011 12:55 am

NaN error on weather forecast in place of date?

Filed under: Software Quality,Tech — ramsblog @ 12:55 am


Looking at the weather forecast for next 10 days, indicates the NaN (not a Number) error on page


Interesting … is there some testing going on the production or is just that the data not available and the computing or rendering made it as NaN?

December 16, 2010 12:13 am

Purchase price different from Advertised price?

Interesting to see the purchase price different from the advertised price on the same site between two pages…

This company has put up a front page banners saying 80% off on 100pk DVD spindle that turns out to $9.99/- . Proceed further to place an order for this item, the item results page shows $14.48.

Now proceed further to check out page or the shopping cart, it shows the $14.48 and other summary of delivery charge etc…

Amount inconsistency? Where might this have gone wrong?
data Configuration?
latency in price points?  so the images for advertisement did not have this price?




Looking at the coupon page, i realized if the offer was In-Store only, and went back to the home page to check if it was available online too. Well, it looks like it does the advt shows “in store and online” .

#WhatIsWronghere #PricingBug #AmImissingSomething

ps: i am not posting this to pin point an issue on the site here. The intent is to probe other software professional visitors here to think what might have happened and how such scenarios can be prevented in their work situations.

Tags: Software quality, User Experience

December 11, 2010 7:31 pm

Great learning on TFS2010 and warehouse ….

I have been playing around the TFS2010 (Team foundation Server) and the way it manages the data between the operational database and its counterpart Data Warehouse. It has been a long time since I had ever tried the warehouse concepts and learning some of those concepts…

At the same time Microsoft Team Foundation Server also offers the Client side of the tools like Test and Lab Manager and they are part of the Microsoft Visual Studio Test Professional SKU. Being a v1.0 of this client utility, obviously we can say there are some limitations, but on the positive side there are great features to leverage with Source control, work item associations, work item management overall and several others…

December 9, 2010 2:39 am

Computer account trust relationship – Excel to Analysis Server


This issue popped up on my computer several weeks ago on my Excel instance when  connecting to the SQL Analysis services. Enquired several people and searched several blogs.

Here is how I had installed the applications on my machine .

Client config
                Windows 7   [7600]
                Office 2010 
                VS 2010 ultimate

Sequence of installation, if that matters:
     Windows 7 –> Office 2010 –>  VS2010 –>  uninstall/reinstall Office 2010.

This is pretty straight forward, right? I don’t see any conflicts here of course, and the add-ins that were required were still in place. And I thought i didn’t have to reinstall VS2010 at this time.

Here is how i would connect the Excel instance to an Analysis services

1. Launch Excel
2. Data –> from other sources –> Analysis services
3. Provide the server name on data connection dialog, and connect if you are connecting with windows authentication
4. takes a while and boom – gives this following error message:

“The following system error occurred: The Security database on the server does not have the computer account for this workstation trust relationship”


5. well, apparently i tried logging in with the same user account but via providing UserName and the Password.
6. it would go through the first level ok and gives the list of cubes and perspectives available
7. select one of those and proceed to create a Pivot –> now again, gives the same error message as above….

It is at times frustrating when the message doesn’t say exactly what is wrong and how it can be fixed. It gets more frustrating if the issue wasn’t resolved for longer time.

Well anyways, finally found the solution – who would have thought the issue was something completely *repeat* completely unrelated in nature with an Excel application connecting to a SQL Analysis service, and where did Live ID assistant come into picture.

Yes, thanks to the following MSDN forum post – http://social.msdn.microsoft.com/Forums/en-US/sqlanalysisservices/thread/b058635f-6d3e-4240-a555-1dbd945ae25a – that said three options
1. to update the kernel on the Server
2. check the windows Live ID assistant on client
3. connect as NTLM from the client machine

…and recommended #1. Well, why should I be looking at the server level configuration when i know another person could connect to  the same server and i was unable to. Well, I was also able to connect to the same server from different client machines and not from this current machine…  So, i chose option #2..

yes, uninstalled the Windows Live Id Sign in Assistant:
1. Launch Control Panel –> Programs and Features
2. look for Windows LiveID sign in Assistant
3. right click and uninstall
4. if cautions close all the applications suggested (Outlook, Visual Studio, etc)
5. Restart the machine
6. go over the Excel –> Analysis service connection process
7. wow! connects through and able to retrieve the data..

ufff! this hard just because the message was not that friendly and the event log didn’t have the right message?

July 25, 2010 12:31 am

WTANZ 07 – Web page automation with Ruby / Watir

Filed under: Software Quality,Software Testing,Weekend Testing — ramsblog @ 12:31 am

As usual, I was not sure if I was going to login tonight. However, just made it somehow. The first question I was asked was if i have done the homework. For not looking at the mission completely and not being sure of joining, obviously I had not done my homework.

Oliver got me started with links and pointers right away. I had to do the following:

1. Download Ruby and Install http://watir.com/installation/#win
2. go over the wiki for some details to update and install further components
3. install Watir.

Mission for this session can be found here: http://weekendtesting.com/discussions?vasthtmlaction=viewtopic&t=126.0 

it took a while with several glitches to get through, but got back to track with the session. Oliver walked us through basic commands to navigate the page, locate the page controls, etc. and put the exercise for the team to put the learning into practice. hmmm.. great approach, the lessons have to be applied immediately to get better at it and theoretical knowledge is not good enough.

as we went along, we tried several commands and controls to get hand of it.

What I learned:
> certainly a great start learning another tool/language
> teamwork in helping each other getting upto speed
> got some hang of ruby/watir which i had never used before
> while following the chat, i tried copying the commands so i could get upto speed and worked well at the same time.
> I still couldn’t get the regular expression search working, i need to try it over again.
> attend  another similar session :)

my sample code:

>> b = Watir::Browser.new
>> b.goto("http://weekendtesting.com")
>> b.text.include?"Forum"
>> b.link(:text,"Forum").click
>> b.text.include?"Next weekend Testing Session"
>> b.text.include?"Forum"
>> b.text.include?"Next Weekend"
>> b.text_field(:id,"log").set "<username>"
>> b.text_field(:id,"pwd").set "<password>"
>> b.checkbox(:id,"rememberme").click
>> b.button(:value,"Login").click
>> b.text_field(:id,"s4").set "comments via Watir on july 24 WTANZ 07"
>> b.button(:value, "Submit Comment").click
>> b.link(:text, "Next Weekend Testing Session").click
>> b.link(:text, "WTANZ session #07 – Test Automation with Watir").click
>> b.text.include?"Reply"
>> b.link(:text, "Reply").click
>> b.text_field(:name,"message").set "Reply WTANZ 07 message via Watir on july>
>> b.button(:value, "Submit").click

What would have helped me in this session:
> a little homework would have helped, but with the team’s sharing all through the session, helped get back on track
> when working on Web page automation, getting to know the control properties
> using home computer, i didn’t have the IE web developer add-in or fiddler tool  installed, and having those would help get the control properties sooner.
> the posts, and search worked well on this session, I need to venture into validating if the actions really took place in addition to visual verification.

There was a great debriefing / retrospection session after the learning and applying tasks. There were conversations around Watir, Cucumber, Selenium, and other tools. No, i haven’t used any of these. It would be interesting to know how these several tools scale up to the robust web automation.

There is always a challenge of changing the Control IDs and Names in regular coding cycle. How would a tool cope up with such changes. We had once taken an approach of keeping these in configuration / resource file, so we wouldn’t have to change the code with any of those controls changes.

Overall, it was a great session. Thank you Marlena (@marlenac), Oliver (@Oliver_NZ) and the team.

May 11, 2010 12:42 am

How is file transfer rate computed?

Filed under: Software Quality,Software Testing — ramsblog @ 12:42 am

Is there a problem here? or am I reading wrong?


i am uploading some files over to the server and the file transfer progress bar indicates the speed as 24.2KB/second and about 29kb of file size remaining to be uploaded. With a simple observation, it should take a little over a second to upload the remaining data. Then, is there a reason for it to compute as over 4minutes as estimated time of file transfer?

May 1, 2010 9:43 am

EWT#16 – Find Barcode value of an individual

Filed under: Software Quality,Software Testing,Weekend Testing — ramsblog @ 9:43 am

I thought i was late for this mission today when i woke up 15 minutes into the session. logged into Skype and didn’t see anyone around. waited for sometime and then realized my clock wasn’t adjusted well to the UTC time.

EWT16 sent me a message and we started off.

Mission: This Flash app generates barcodes – but how? Can you find how it’s calculating them? Find the highest value that you can, and report invalid values. 
Product: http://www.barcodeart.com/artwork/netart/yourself/yourself.swf

Participants when we started off:
     Anna Baik [blog] [twitter: @testerAB]
     Stephen Hill
     Jeroen Rosink [blog] [twitter: @jeroenro]
     Ram [blog][twitter: @esaarem]
     Thomas Ponnet [blog]

idea was to find a biggest value with variable parameters

there were few clarification questions: it is good to see participants asking questions to understand the mission. Although there was  a caveat that moderator would sometimes be vague giving the details. I think it makes sense – and helps make the necessary assumptions and justify those assumptions. That’s what we apparently land up in our daily tasks, make assumptions, and ensure why we make such assumptions and clarify those assumptions.

> find the algorithm or thought behind?
> usability?
> how do we find the value from the bar code?
> should the mission end at the finding the value alone or anything else?
> and so on. ….

Anna, answered most of those questions and participants pitched in with their understanding and shared their thoughts. It is helpful to understand from like minds across the regions.

My approach:
I tried to plan around a bit this time:
> Varied input parameters and understand the pattern
> look for source code
> read a bit more on home page
> boundary value analysis
> and a few more that i jotted down on a piece of paper

At first my understanding from highest value was based on the number below bar code – well thats a dumb assumption i guess. I didn’t think of a dollar value until someone mentioned about SCAN button.

The most i got was $9.62

this how the diagram look like and shows the input variables that affects the value on the barcode


I started off with a table like one below and expanded as proceeded. Boundary Value Analysis concept helped here to understand the trend.
STarted off with a median age/weight and
the indications i had was
> if the leading countries or
> most known country would influence the decision.
> If a taller and average acceptable weight influences the value of the person
> of course zeroes, negatives and max outs…
> and several other assumptions in my approach…

i have also read the introductory text on the website

iteration Gender Age Country Weight Height Barcode value dollar value
1 M 51 US 111 6-3    
2 F 99 zambia . .    
3 M . . . .    
4 F 21 US 98 5-8    
. . . . . .    
. . . . . .    

I made a lot of notes on the trend varying the input factors to work backwards on the algorithm being used. it was interesting to see the factors it considers to put out the results it does. the trend analysis helped me in a way within very few iterations to understand what brings the value up, and so I thought. and realized there were few other factors that i overlooked.

Noticed Thomas and Jeroen talking about the BMI calculations in their approach to find the highest value. hmm… that makes sense. I did not think about BMI in my execution.

there was a mention about Insurance companies and how they may arrive the human values… well thats a good correlation i think. How would insurance companies set the value for premiums for individuals? based on Age, height, weight, overall health and his/her hygiene habits? so perhaps…

there were several bugs logged in repository as well – see here http://www.bugrepository.com under project EWT16

i have some other thoughts about the algorithm that i will have to detail it out here later time… it is a bit of writing..

Michael Bolton (twitter: @michaelbolton) joined in during the debrief.

i have a problem with my system clock that resets to older time by itself. It is tricky when the skype messages are coming in and my clock was reset to previous hour so the messages were mixed up and not realizing there were new messages coming in but getting into another sequence. :(

there came up the discussion about Modeling the approach. How different states were defined individually while doing so. @MichaelBolton mentioned about the focusing/defocusing heuristics – i learned the new terms:
OFAT – varying One factor at a time –> to focus one the effect of a particular factor
MOFAT – varying Many factors at a time –> seeking to confirm or disconfirm ideas about factors
HOFAT – holding one factor at a time

@Thomas clarified on Model approach –if considering if gender state made a difference, or / and if weight or height states influences or impacts the results.

@Joroen mentiond about TMap. although I thought i had heard this somewhere in some context, but i was not sure. He pointed me to this website: http://eng.tmap.net/Home/ i will take a look at this website to know more about TMap – Testing Management Approach.

Overall, it was a great session chatting with all the participants.

It reminds me of the blue angels or airshow where each practice is a new and unique mission for those pilots. and they go over the PLAN – BRIEF – EXECUTE – DEBRIEF – PLAN cycle. the challenge is unique each time and learn from those iinstances each time. but the structure and the plan they would adopt is / might be same or similar for each mission. and what they come out with after debrief is new learning and emphasizing on old learning, i suppose

Thank you EWT16. see you next week possibly…

(p.s: i haven’t reviewed this post before publishing, any miss-outs or overstatements regretted – please let me know i will correct them)

April 18, 2010 12:17 am

WTANZ – Weekend Testing Australia NewZealand

Filed under: Software Quality,Software Testing — ramsblog @ 12:17 am

i have been tied up with sorting out months of papers this evening and took so long and almost missed this meeting tonight. I was just about close my computer and go to sleep and realized it was saturday night and I had told Marlena [blog] [twitter @MarlenaC] that i was going to join the WTANZ this time.

Logged into Skype and noticed Marlena’s message. I was already 20 minutes late into the session. after exchanging few messages with the group, i got the objective http://groups.google.co.nz/group/software-testers-new-zealand/browse_thread/thread/6cbee3a6fb040310

looking over that and the objective was to : (from Oliver’s link above)

Website Tested:

1) Have a look at the mortgage calculator and roughly script/try some
tests to verify functional correctness. Try and be creative about
breaking the solution. (rough time estimate 40min)

2) Have a look and think at usability and if this calculator matches
your expectations or not. (rough time estimate 10min)

3) Have a look at browser compatibility and make a quick assessment.
(rough time estimate 10min)

Please do not…
* …automate your testing
* …performance test
* …security test
* …test the Login functionality. Just the mortgage calculator.

Aim of this exercise is to show & share different methodologies on how
to test something. There is no right or wrong here. Do what you
usually do and share your process in the discussion bit at the end.
Don’t be shy to share. Ask for help if you don’t understand something
or want more details!

well, this time for joining late and having only few minutes left, i jumped straight into exploring than planning my approach and asking any questions although i found some ambiguities there. here is the screenshot of the calculator screen



Marlena and others mentioned about logging the defects and observation on http://www.bugrepository.com – a defect repository where we can log the issues for tracking. Being a first time user for these tools, it took a little lag in coping up between the task, IM chat, bug repository logging defects, etc.

Alright, enough of my excuses and story but what did I learn from the discussion for the time I engaged with:

> learned about following tools to leverage:
         http://validator.w3.org for HTML checker
         http://www.browsershots.org –> a fantastic tool that provides several browser types and runs the given URL in all those selected browsers and provides the screenshot – absolutely Fantastic tool to see how the screen looks on various browsers. There were browser names I had never heard of.

> various different ideas the group shared in how they approached and things they did to get through the objective.

> great communication – kind of questions participants asked during the session; help sought and offered; tools shared, ideas shared, etc…

>Questioning skills – @oliver asked great questions and so also others. It is critical to ask the right questions in solving any mission, which I am still working on, it was great to observe the pattern and learn from the group

Oliver Erlewein [site] moderated the session through out. I liked the way he asked the questions every time a participant reported an observation. some of those were like
      “do you think that’s a defect?
      “how would you report that bug
      “why do you think that’s a defect?” 
      “If you had a luxury of time how would you have approached?” and so on…

How about Planning for problem solving? Looks like some of us being new to skype spent some time figuring out using the chat sessions. however, @Oliver asked if anyone went through the planning process for problem solving. I am not sure if I’ve missed, but noticed most saying jumped straight into the task. I did so too. I would probably have taken another approach if we all had joined the problem solving at the same time. personally, I have no excuse, could still have done some outlining even in the time I had. I missed. Here is what I would have done perhaps:

a. understand the objective to more extent

b. Ask questions for ambiguities – in this case i noticed multiple mortgage calculators, i wasn’t sure which one the group was looking at.

c. plan on the approach for different types  of testing – including boundary, security (well, security was out of scope from mission), and edge scenarios

d. Scenarios – understand how mortgage concept works and come up with the approach / scenarios around the user context – customers first – Customers use the product in ways we don’t imagine – explore being in customers’ shoes.

e. plan on the tools and resources required to complete the mission  (see above for some tools shared by team)

f. ESTIMATE what it takes for the planned approach – so that helps building the estimation skills which we often need to develop for the task completion; and can be tracked during the task progress and after completion

g. attack the mission – well, there could still be situations where we might have to keep the plan aside and do what’s needful – that’s where the Risk Identification and Mitigation would have helped;

It was great to know James Bach [website] [blog] was on the WTANZ too. It was great to join in my 2nd of the series WT in past few weeks, and the 1st WTANZ (Australia and NZ). I hope i will be able to join in these in upcoming weeks too. Problem solving is great at these sessions but the retrospection or debriefing discussion after the exercise that I liked the best that we interact so much and learn their approach and what would have gone right and the improvements could be made in future for similar instances. Thats invaluable.. talk about the participation? I think there were people from across the globe participating in this session. 

It was great to catch up with Marlena since she moved from the states few weeks ago. Thank you @Marlena for mentioning about this session and timing (zone) seems to worked out for me this time :)

if you are interested to participate in Weekend Test – see http://www.weekendtesting.com

<ps: i will update the links and any further findings a little later as appropriate>
<Apr 18 update: updated few links and questioning sections>

Next Page »

The Rubric Theme Blog at WordPress.com.


Get every new post delivered to your Inbox.

Join 55 other followers