Flight company’s booking system, not only riddled with software glitches, but they are so un-customer friendly that in order for ME to do THEM a favour and let them know why their web sales are so poor – alerting them to the software glitch, I either have to pay £s per minute to call their website helpdesk or post a blog about their immense screw-up (thereby shame-facing them at the same time) There is simply no other option available on their website for getting in touch with their webmasters – Come on guys a simple email address wouldn’t hurt, would it?
Sorry, deep breath, this has just got me so riled!
So, knowing that I was planning a short break in Spain later in the year, a friend kindly forwarded me a promotional mail he had received from Aer Lingus, about a sale on flights to Malaga.
Well, I’m in Marketing, so I was never going to be fooled that it would actually only cost £29.99 (not when you take into count that I wanted to fly at the weekend, I actually wanted a seat and to take a bag etc), but anyway, £102.56 each for me and my husband (plus a handling fee of £20) didn’t seem too awful.
Happy with the flight times – check.
Think an extra £370 pp to be able to take checked bags, have seat selection and airport lounge is a bit extreme, but happy to run with the cheap option – check.
Flights cheaper than taxes – typical – £20 handling fee snuck in there – typical – check.
Well, ready to complete the booking then…
Click continue.
Start filling in my details, scroll down – OMG!!! How come the price has suddenly hiked itself right up to almost £600!!!!!
Ouch, that’s a nasty software glitch. Working at a company that actually manufactures software testing solutions, I can tell you straight off, that this here is an application error – and not just a malicious ploy by the budget airline to part you from more of your hard-earned cash! It seems that Aer Lingus have not been testing their application very well. We’ve seen time and time again that choosing not to invest sufficiently in quality assurance can really come back to bite you on the rear end! With ample time devoted to testing, better test coverage, some validation rules and database effects, this would have been picked up way before it went public. Oh well.
Maybe some customers might not have double-checked the price again, and unknowingly proceeded, only to be ripped off by the good chaps at Aer Lingus, but more likely, customers like me, would have just been put off and decided to check out prices at BA instead! I guess I’ll do a good dead and let them know – perhaps suggest they buy our software while they are at it… An hour later and after crawling their website, I discover, to my annoyance, that unless I want to waste even more of my time and money phoning a premium rate number, there really is no way to contact their technical team! GRRRRRR! Oh well, at least it is good fodder for our ‘Software Testing Hall of Shame’!
In trying to re-create the error before I put this post live, I came across another, completely different bug , where the system now wouldn’t process to that last screen, but looped me back to the start again. Do they do ANY testing on their systems? It’s just riddled with bugs.
Software errors force Boris to back-pedal on hire charges
A cycle hire scheme championed by Boris Johnson is having to refund thousands of pounds of taxpayers’ money after it was hit by glitches.
This week my favourite London Mayor, the enigmatic Boris Johnson, makes it into our Software testing Hall of Shame!
Mr Johnson apparently told LBC radio: “I have no doubt lots of things will go wrong.”
He insisted it would prove a “great success” in the long run but added: “The reality is that the software issues and technical issues of getting the whole thing up and running for tomorrow has been extremely difficult.” “It will be more of a gradual launch than a big bang. I have so many concerns it’s hard to pick one out.”
It seems that the blonde haired buffoon had not allowed enough time for adequate testing of the scheme and ironing out the kinks. By this point they were left with little choice but to go ahead with the launch despite concerns, but as the saying goes: Failure to plan is planning to fail.
Unsurprisingly Monday morning came with numerous reports of mayhem in the capital over the weekend.
The London Evening Standard reported that a quarter of both the promised bikes and their docking stations weren’t ready in time. The Boris bike section of the TfL website froze on Sunday and users yesterday claimed their electronic keys were not working despite activating them online. Problems have emerged with people who purchased multiple electronic keys from the same bank account – the daily or weekly subscription charge is automatically levied on all keys as soon as one is used, even if the others are not in use.
Oh dear, Boris.
The Telegraph humorously reported that ‘Mr Johnson, who describes himself as cycling like a “very elderly French onion seller”, admitted he is expecting “delirious” criticism of the scheme.’
Drilling deeper into the World Quality Report
Some good headlines about emerging industry trends came out of the recent World Quality Report produced by HP, Sogeti and Capgemini, but as I read deeper, some things started niggling at me, so much so that I decided they were worthy of a blog post.
For instance:
‘Only 4% of IT professionals agree that their ALM investments are fully paying off, and only slightly more than a third said that half or more of their solutions have been fully implemented and are helping to improve application quality.’ (Pg.8)
HP and Sogeti state that ‘QA organizations need to focus on increasing the adoption rates of their purchased technologies’. This is of course great advice, but both companies have a vested interest in these users throwing more and more money at consultancy, training, and in HP’s case, overpriced and difficult to implement products, (that consequently need a lot of ‘services’ cash thrown at them).
In my opinion it would be more prudent to look in more detail at why these companies are not getting a return on their investments or managing to fully implement their products.
The answers seem to fall into three categories. Company issues, Resource issues and Technology issues, with technology resoundingly winning the prize for biggest stumbling block.
20% failed due to lack of an internal process or support from management. The question does beg to be asked – how on earth did they ever manage to purchase their ALM technologies without some form of internal support? To make ALM successful, it must touch more than just the QA team: support needs to be gained, processes mapped out and business goals and requirements defined way before you make decisions about what technology to buy. It’s really not surprising that these projects failed.
26% stated that not enough resources were invested into the adoption of the technologies. Now I could have classed this under company issues, but I think you really need to look deeper than just writing this off as a staffing issue. It could equally be a technology issue. Was it that not enough staff were trained? Was the project badly planned? Was the technology too complicated for users outside the dev/test team to adopt?
In terms of technology, a whopping 41% struggled because their ALM investment was the wrong choice of technology; it failed to integrate with other technologies or was too complicated and required specialist skills that were thin on the ground. I’ve lost count of the amount of times we hear stories like that when meeting companies across the globe. To truly embed a solution in your company, you need to empower all stakeholders. Unless everyone involved in the delivery of IT projects can collaborate using the solution, it’s just not going to work. When choosing a solution, you need to think about how easily management, business analysts, business users, developers, project managers and testers are able to get what they need out of the solution.
‘Companies prefer testers who have both strong technical skills and relevant domain and business knowledge’ (Pg.11)
Well the stats don’t really allude to that. The question that was asked was – When hiring testers which of the following skills are most important to you? Well obviously QA skills came out tops at 31%. Having a good grounding and understanding of the principles of Quality Assurance is key for testers, I’m actually more surprised that this figure wasn’t higher, but interestingly, the second largest desired skill is business knowledge (22%). This is something we come across time and time again with companies we talk to; so many of them utilise business users for the testing phase. Take SAP testing for example, business process is key. You really need to leverage the knowledge that the business users have about how the system is supposed to perform and exactly how they all use it. So many of the accounts we’ve been into have been literally banging their heads up against a brick wall trying to work out how to capture this knowledge or utilise these testers, knowing that it is impossible with their current toolset – HP is just too cumbersome to get non technical business users to adopt. Development skills 9% and scripting skills 10% are actually rated incredibly low when you consider that the market dominating tools actually force these prerequisites upon QA and make these skills imperative at sites where these traditional tools are embedded.
‘Nearly three quarters of respondents say that they do not follow [common test management methodologies]. Instead, their organisations develop and document their own best practices that are followed in the majority of development and testing projects. (Pg.10)
Different groups in the organisation may adopt their own ‘versions’ of the standard practices, and as a result, the company as a whole is not fully realizing the benefits of standardization, economies of scale, common metrics, unified reporting and asset reusability’
Not all companies are equal and each has different ways of doing business. One size DOES NOT fit all, so surely it is good for the industry that companies develop their own best practices? These companies are just using their brains and working out what best suits their own unique needs and circumstances.
Software vendors should be supporting this very obvious progression of development maturity. Why shouldn’t they be able to all work slightly differently, yet still enjoy the benefits of unified reporting, asset reusability, common metrics etc.? Perhaps the answer lies in the fact that HP hasn’t built its software to be this flexible? Maybe it’s time for the dominant market player in test automation and management to listen to what businesses need rather than telling them how they should be working!
This particular bone of contention was revealed earlier this year with a survey of Application Development Managers back in April 2010. The industry is really crying out for flexibility in the way that tools allow them to work, which is one of the reasons that Original Software developed Qualify, a process and methodology agnostic Application Quality Management solution. Qualify allows businesses to map their own processes, use standard methodology templates, tweak them to suit their own needs and even run multiple methodologies across different teams and projects, with – wait for it, all the added benefits of unified reporting, metrics, re-usability and economies of scale.
It's not a bug, it's a feature! Apple: 'Hold your iPhone 4 differently'
Apple and the iPhone 4 feature in today’s topical entry to the Testing Hall of Shame. As reported by computerandvideogames.com, ‘The geniuses at Apple have put their heads together and come out of the lab with a simple fix for the iPhone 4’s bizarre reception issue: hold your phone differently.’
This week saw the release of most sought after smartphone on the planet, the iPhone 4, but a fatal flaw was quickly discovered – that holding the gadget in a certain way kills reception, which means making phone calls or downloading games is rather difficult.
Here’s a video of the issue, courtesy of TechRadar –
Apple’s statement brushes off the defect as a common issue that is found in most devices.
“Gripping any mobile phone will result in some attenuation of its antenna performance, with certain places being worse than others depending on the placement of the antennas,” it says.
“This is a fact of life for every wireless phone. If you ever experience this on your iPhone 4, avoid gripping it in the lower left corner in a way that covers both sides of the black strip in the metal band, or simply use one of many available cases.”
So, if you’re having problems, hold the phone in a different way or spend more money to fix the issue preventing the device from performing its primary function. That’s that then.
TechWorld reports that ‘One user saw a boost in signal strength after wrapping the phone in a Ziploc bag, call it a super-cheap alternative to Apple’s $29 bumper. Another post on Apple’s forums recommends covering the sides of the phone with InvisiShield tape. But as one official Apple forum-goer noted, “Apple needs to fix this problem, we should not have to rig a fix for a brand new phone.”’
Thorough testing should have uncovered this bug, but how are Apple now going to appease the thousands of customers who not only queued up for up to 32 hours to get the buggy phone in the first place? One speculation is that Apple are going to release a propeller for the iPhone 4 so that it can hover next to your ear while you make a call!
George Wilson, Original Software, will be hosting the webinar this Wednesday, May 12th 2010, discussing the reality of test automation in an Agile environment.
Agile is a methodology that requires a change in the way QA and development work together. The use of technology and automation are much more difficult and finding a practical approach to testing is critical for successful Agile projects. This webinar will explore how testing in Agile is different and George will give pragmatic advice to ensure application quality.
Listen in and learn how:
* To overcome the barriers to Agile automation.
* You can progress into automation gradually and easily even in an Agile environment.
* Your application quality can be managed with complete visibility.
Computerworld reported that the Co-operative bank was ‘losing customers’ through system problems.
Co-operative Financial Services has severe system problems that are causing it to lose online customers. Users told the BBC they were unable to access their accounts at times, and that transactions online often do not work.
Marc Palmer, from Gloucestershire, who runs a small business, told the BBC: “A lot of the time you can’t even log in. Other times, you can’t see your bank balances or any of your accounts listed. There comes a time when it’s damaging to your business.”
The bank is now set to upgrade its systems in response to the complaints of customers, who vowed to move to a competitor, BBC Radio 4’s Money Box programme found. John Hughes, director of retail products at the bank, apologised for the “inconvenience, difficulty, frustration and irritation we’ve caused our customers”.
FAA Computer Glitch Delays US Flights
Flights ground to a halt throughout the USA on Thursday morning, reported Stickyminds, after a mysterious computer glitch hampered a key Federal Aviation Administration flight processing system.
The system electronically inputs pilots’ flight plans to computers, telling air traffic controllers the anticipated route and altitude of each flight after taking off. ABC News, who originally broke the story, said Controllers they spoke to had to enter those plans manually, resulting in a slowdown of takeoffs and landings.
Bird Brain Collider
Oh, and because it made me chuckle, I’m including a tweet this week, courtesy of @danfusion – http://twitter.com/danfusion/status/5460071784. Speaking about the latest in a string of failures concerning The Large Hadron Collider, he says ‘Thinking about software testing: How many times has an app broken because of a bird brain with bread?’ If you haven’t read the story, It seems that the billion dollar piece of machinery has yet again gone caput, this time due to a bird dropping a piece of bread on a section of the accelerator!
Software Testing Hall of Shame: Big Blue Red Faced over poorly tested congestion charging debacle
This week IBM step up to the podium in our Software Testing Hall of Shame. Computerworld UKyesterday reported that London’s congestion charging payments system crashed after IBM took over the contract.
London drivers were unable to pay the congestion charge online on Monday, following a system glitch after IBM took over the work from Capita. The problem emerged after IBM began a “significant upgrade” to the transport agency’s computer systems. IBM had spent the weekend – when no congestion charge is applicable – migrating data from Capita’s databases to its own systems, as it took over the work.
It looks as though the IT giant was not very ‘rational’ in its approach and did not fully test the system before go-live on Monday, as Computerworld reported “there were issues with the new systems and some of the data matching between different programs.”
A spokesperson at Transport for London, which operates the congestion charge, apologised for the online problem and insisted no registered drivers would be penalised if they had had problems paying. She added that the glitch was “resolved by the afternoon and all of the new IBM systems are up and running as they should be”.
The IBM system upgrade is aimed at saving TfL £200 million on running costs by 2018, but so far IBM has only cost the company, in terms of lost revenue.
The Seven Deadly Sins of Software Test Automation
Just to get into the spirit of Halloween next week, take a look at our new whitepaper, “The Seven Deadly Sins of Software Test Automation,” (cue evil laugh). We do enjoy being original and I am sure you will have a little chuckle to yourselves with this one!
In this whitepaper we explore each of the seven deadly sins as they relate to software test automation. Instances that Original Software comes across time and time again and traps clients have often fallen into because of their earthly vices.
I am quite eager to see what you think of this piece, so please write back with any comments! If you have a great sense of humour, why not reply back with your own software testing deadly sins!
Great Places to network with other Test and QA professionals
(Part 1 – Linked-In)
Linked-In is fast becoming one of the best platforms to pick up industry related news and network with other professionals. There are literally thousands of Linked-In groups, so much so, that finding the most relevant, biggest, best or most active groups can be a bit of a challenge!
To help you on your way, here is a selection of the groups we’re familiar with and can heartily recommend.
