London software testing news UK


Succeeding with Offshore software testing

Posted in Software testing by testing in London on August 26, 2007

From Information Week

Offshore software testing projects fail primarily because of a lack of expertise and experience working with an offshore company. This is true for both customers as well as offshore companies. Most offshore vendors have very little experience delivering successful projects.

This problem is compounded by the antiquated and overly simplistic approach that most customers and vendors use to measure success – “on time and on budget.”

When an offshore testing project fails, the cost is typically greater than the cost of the original problem. Software development slows down, customer satisfaction declines, costs increase and frustration mounts.

When do offshore testing projects fail? Right from the start – when you select a partner, not when they actually begin the work.

It can be difficult to detect failure early if you haven’t completed many successful offshore testing projects. Failures typically become evident as deliverables come rolling in (or not) and issues go unresolved. This can take weeks or months.

UK test management job

CA Enhances Application Quality and Testing Solutions

Posted in Software testing,testing tool by testing in London on August 25, 2007

From Business Wire

At SHARE in San Diego, CA  announced new versions of three mainframe application quality assurance and software testing solutions that help organizations more effectively leverage their legacy application investments within a Service Oriented Architecture (SOA) environment.

By enabling customers to more effectively test and troubleshoot mainframe applications — both as self-contained applications and as components of end-to-end Web service-based infrastructures — the newly-enhanced solutions speed problem resolution, improve service levels and promote application quality.

CA InterTest for CICS r8 and CA SymDump for CICS r8 exploit the capabilities of IBM CICS Transaction Server release 3.2, which enables both traditional mainframe transactions and Web-based applications to be efficiently integrated into a SOA environment. CA InterTest for CICS r8 and CA SymDump for CICS r8 provide debugging capabilities and fault analysis for IBM CICS Transaction Server features that support Web services such as Channels and Containers for 64-bit storage. The capabilities enable application developers to leverage traditional mainframe applications in new environments.

In addition, the new support that CA InterTest Batch provides for nested DB2 stored procedures helps developers debug code that is critical for extending legacy applications into Web services and SOA environments. By providing insight into simple as well as nested DB2 stored procedures — which are increasingly used by Web services providers and their customers — CA InterTest Batch r8 makes it much easier for application developers to ensure the quality of applications before they are rolled out into production.

All three solutions also provide program source-level integration with other CA management products, including CA Software Change Manager for Mainframe (formerly CA Endevor). This integration enables developers to speed problem resolution and debugging by ensuring that they are working with the latest versions of whatever program they are testing. Powerful and intuitive menu-driven components make all three solutions easy to use for developers with all levels of mainframe experience.

Approaching SOA Testing

Posted in Software testing by testing in London on August 24, 2007

From SOA world

So, does testing change with SOA? You bet it does. Unless you’re willing to act now, you may find yourself behind the curve as SOA becomes systemic to all that is enterprise architecture, and we add more complexity to get to an agile and reusable state.

If you’re willing to take the risk, the return on your SOA investment will come back three fold…that is, if it is a well-tested SOA. Untested SOA could cost you millions.

Truth be told, testing SOAs is a complex, disturbed computing problem. You have to learn how to isolate, check, and integrate, assuring that things work at the service, persistence, and process layers. The foundation of SOA testing is to select the right tool for the job, having a well-thought-out plan, and spare no expense in testing cycles or else risk that your SOA will lay an egg, and have no creditability.

Organisations are beginning to roll out their first instances of SOA, typically as smaller projects. While many are working fine, some are not living up to expectations due to quality issues that could have been prevented with adequate testing. You need to take these lessons, hard learned by others, and make sure that testing is on your priority list if you’re diving into SOA.

How Do You Test Architecture? The answer is, you don’t.

uk job software testing

Overhauling procurement and UAT requirements

Posted in Acceptance testing by testing in London on August 23, 2007

From IT WOrld Canada

A much-needed overhaul to provincial government guidelines surrounding technology vendor procurement contracts may do away with an often convoluted and frustrating negotiation process.

The government of Ontario will start using new terms and conditions for small, low-risk technology contracts that are intended to reduce the time required to arrive at a contract agreement, increase vendor participation and produce better vendor pricing due to increased competition.

The guideline facelift was the result of collaboration, over the past 14 months, with the IT industry, via the Information Technology Association of Canada (ITAC), a Toronto, Ont.-based organization that champions initiatives pertaining to the IT industry.

Specifically, he said, the changes pertained to the clarification of the contract language; and amendments to policy, including those that vendors were complaining had been difficult to implement.

Other affected policies included intellectual property, limitation of liability, and user acceptance testing.

Part of the revamp entailed removing certain time-consuming terms and conditions that were not likely to have practical application anyway, “yet for the bidders they make life so complicated that some of them refuse to bid or have to try and factor in the extra risk,” said Bernard Courtois, president and CEO of ITAC.

Mobile Java Testing Services Expanded

Posted in Software testing by testing in London on August 22, 2007

From 3G

Complimentary to conformance testing according to the new GSMNA (GSM North America) test specifications CETECOM now offers quality tests for the Java platform in mobile communications devices.

These tests go beyond the conformance requirements of PTCRB and serve to test integration and performance of a Java Virtual Machine (JVM) and its Application Programming Interfaces (API) in a mobile communications device.

In order to meet the individual needs of manufacturers, network operators and developers for additional quality tests as well as the new PTCRB requirements for mobile Java CETECOM has invested in Communology’s test platform mte (Mobile Test Edition).

This platform has been developed according to the PTCRB specifications as well as to the functional requirements of the carrier industry for mobile applications based on Java. The test environment for J2ME-enabled mobile phones covers five different fields: basic tests, security tests, network tests, performance tests and OEM specific API.

mte is in equal measure a conformance test system and a quality test tool. Thanks to this, the conformance tests which are required for entering the North American market can be combined with the additional carrier-specific tests on one platform so that the customer benefits from an optimised testing process.

A high level of automation of the user-friendly test environment combined with clearly arranged test reports allow the fast analysis of potential errors and an efficient defect management. As a result of this, the customer benefits from savings in terms of time, costs and resources

Project Boat in user acceptance testing (UAT)

Posted in Acceptance testing by testing in London on August 21, 2007

From Financial News

Project Boat, the trade reporting consortium owned by 15 of the largest investment banks, spent most of the past 12 months quietly going about its business because attention has been focused on trading initiative Project Turquoise. But all may not be plain sailing as Boat nears its destination.

The project, which was originally led by Merrill Lynch with the backing of eight of its largest rivals, including Deutsche Bank, Goldman Sachs, Morgan Stanley and UBS, aims to take advantage of new European Commission rules coming into effect on November 1.The Markets in financial instruments directive (Mifid) outlaws concentration rules, mandating brokers to report trades to a recognized trading venue, whether or not the trade is executed OTC or on the exchange, effectively opening the exchanges’ trade reporting businesses to competition.

Project Boat was overshadowed last November when seven of its banks announced they were also working on a parallel platform for trading, a move that challenged the main trading revenues of Europe’s top stock exchanges.

Hindsight suggests Boat may have benefited from this. Whereas Turquoise has found neither a technology provider nor a chief executive, despite an exhaustive search, Boat in January announced its technology partners – Markit Group, a data vendor, and Cinnober, an exchange system specialist – and in June confirmed another six banks – Barclays Capital, BNP Paribas, Dresdner Kleinwort, JP Morgan, Lehman Brothers and the Royal Bank of Scotland – had joined as shareholders on an equal footing to the nine founding partners.

The consortium has been careful never to commit to a live date but sources close to the banks said last week they have circled November 5 for full live running, a date that gives the banks the preceding weekend to finalize their internal system tests. User acceptance testing began as planned.

A consortium member said last week he was “delighted” with the progress made by Boat this year. He said: “We are slightly ahead of schedule in achieving our objective and there is no doubt we will meet the deadline. The platform provided by Cinnober is well within target, with the final release having been received in July. Testing is progressing well and parallel running can be expected to start about the end of August, as projected. All the major market data vendors are signing up agreements to take and distribute the data provided through Boat.”

Whether the consortium banks can resist the temptation to subjugate their rivals in the same way the exchanges lorded it over them is a question only they can answer. But it is easy to understand why the industry, particularly the banks’ fund management clients, are skeptical.

It’s stress testing, but not as we know it

Posted in General by testing in London on August 20, 2007

From Startribue

There’s a difference between reacting to short-term market wiggles and overreacting to them.

So while experts routinely tell fund investors to ignore day-to-day deviations, what happens over a month or two can be telling, particularly in bond funds. And based on what has been going on in the mortgage and credit markets, this is a time for many bond fund investors to react and recheck their position.

That’s not a call to sell bond funds, but rather to examine performance more closely, because the wild conditions that the market has lived through since the end of June have created the perfect crucible for stress-testing a bond fund’s portfolio and your tolerance for what the fund holds.

If you set aside the headlines tormenting fixed-income land, you can sum up the current situation this way: It’s a great time to be a tortoise, the bond markets were so favorable that bond fund managers had an easy time putting up decent results. The hares — the funds that stretch for extra yield — were rewarded for taking high risks.

While that appeared to be changing in the early part of 2007, there’s no doubt it changed by June 30. That’s when the market effectively repriced risk, and a four-year trend of tightening credit spreads effectively reversed itself over the course of a month. As a result, Treasury yields are down dramatically, and when yields fall, bond prices rise.

Cisco and Texas Instruments carry out Interoperability Testing

Posted in Software testing by testing in London on August 19, 2007

From TMCNet

Cisco  has  performed interoperability testing of the Cisco uBR10012 Cable Modem Termination System (CMTS) platform with Texas Instruments’ Puma 5 DOCSIS 3.0 CPE development platform. During the interoperability testing, the Cisco uBR10012 CMTS solution showed upstream channel bonding – a key feature of CableLabs’ DOCSIS 3.0 specifications.

The interoperability testing at CableLabs revealed that the Cisco uBR10012 CMTS platform and Texas Instruments’ Puma 5 DOCSIS 3.0 CPE development platform can be used jointly as part of a DOCSIS 3.0 network.

Cisco’s uBR10012 CMTS solution offers competitive density, performance and features. The upstream channel bonding, which integrates several radio frequency channels, will definitely make it more resourceful and also proves that the Cisco uBR10012 CMTS platform is likely to support cable operators’ high-speed cable-modem services by offering faster upload speeds.

Testing Tool for Agile Software Development

Posted in Software testing,testing tool by testing in London on August 18, 2007

From Linux Magazine

SolutionsIQ Inc., a leading provider of information technology services, consulting and staffing, has announced the launch of StoryTestIQ (STIQ) 2.0, an open source testing tool designed to create executable requirements for Agile software development projects. As a leading provider of Agile based software development services, STIQ complements the company’s service offerings and is now available to developers for broader use.

StoryTestIQ was developed based on two other open source testing tools, Selenium and FitNesse, combining the web browser-based testing facilities and command language of Selenium with the wiki-based editing and pluggable FIT fixture loading of FitNesse. STIQ 2.0 advances testing beyond user-interface scripting with new features including domain-specific language support, enabling increased productivity, and better team-to-client collaboration.

As part of the SolutionsIQ mantra “Build Integrity In,” and a dedication to Test Driven Development, StoryTestIQ is uniquely designed for cross-functional teams that are developing web applications for enterprise customers. With STIQ, developers have a definitive and executable view of the requirements for a specific iteration. StoryTestIQ can be used to record customer needs and negotiate acceptance criteria before a single line of code is developed. StoryTestIQ also can be used by traditional development teams to rapidly create acceptance and functional tests.

StoryTestIQ is a free, downloadable tool available online by visiting (http://storytestiq.sourceforge.net/gettingstarted.html)

Red Hot Testing Model

Posted in General by testing in London on August 16, 2007

From Indian Muslims

Testers will no longer have to torture their taste buds now that testing chilli’s for hotness has gone digital. A team of chemists has developed a mathematical computer model to measure the “heat” inside a chili pepper, a process that could provide quicker and cheaper information to some segments of the food and drug industries and law enforcement.

Peppers are hot because of a family of chemicals named capsaicinoids. Two members of the group (capsaicin and dihydrocapsaicin) make up 90 percent of the capsaicinoids. The spiciness of each pepper depends on the relative amounts of these hot compounds hidden inside the pepper’s flesh.

“Capsaicinoids are the active ingredient in pepper spray, tear gas and some arthritis medications, not to mention spices and foods like salsa, so a wide range of industries could find this new approach useful,” said lead scientist Kenneth Busch, a chemist at Baylor University in Texas.

Currently, the standard test of pepper hotness relies on a process called high-performance liquid chromatography. This method is expensive and time consuming (each sample can take 10 minutes to run), because the machine chemically separates all the different compounds before spitting out a reading.

Busch and his colleagues developed a computer model that can calculate the hotness of a pepper based on information gleaned from spectroscopy, which measures how a substance absorbs light. The program is set up so that when a scientist feeds a “spectra” into it, the program disregards components uninvolved in spicy heat and focuses only on the capsaicinoid factors.

job software testing

Next Page »