Thomas E. Murphy
Research Director
4 years at Gartner
27 years IT industry
Thomas Murphy is a research director with Gartner, where he is part of the Application Strategies and Governance group. Mr. Murphy has more than 25 years of experience in IT as a developer, product manager, technical editor and industry analyst. Read Full Bio
by Thomas Murphy | November 30, 2011 | Submit a Comment
Gartner’s Application Architecture, Development and Integration Summit conference began yesterday (http://bit.ly/rAazgw) with over 1000 attendees and an agenda full of great presentations. This year there Cloud Computing is a major theme and I will be presenting along with Jim Duggan today on the affect that Cloud is having on AD and ALM. We will talk about more mature areas like testing and lab management, and more emergent areas such as the shift of ALM tools from SaaS to PaaS and how Cloud AD platforms are evolving to support both traditional IT as well as the Consumerization of IT.
Jim and I are also taking the time together to work on completing the Magic Quadrant for ALM. This has been a challenging process over the last 6 months given the ever growing size of the market and the evolving nature of ALM. One of the key things we have noted is that in general most tools aiming at enterprise use have gained strong integration facilities recognizing that they won’t be the single stack provider. There are single provider solutions and a few vendors still oriented toward these solutions but in general most organizations won’t have all ALM tasks managed in one single tool. We have noted this in two notes recently: Selection Criteria for Success in Choosing ALM Products and Application Life Cycle Management Matters Where Diversity Persists. In the near term, it would appear that Cloud delivery of solutions will only better enable users to choose governance tools that fit the various roles in an organization and this is a good thing.
Category: ALM Cloud MQ Tags:
by Thomas Murphy | June 23, 2011 | 4 Comments
The rapid movement to build and deploy applications to devices is creating a challenge for testing teams. Solutions from the leading testing vendors are missing, or provided by partner extensions, and the pace of technology evolution will keep them chasing for several more years. Organizations should recognize that testing choices (tools, services, etc) will be tactical during this time period. During this time consider the following:
- Scenarios under test: customer facing, App Store, location, motion
- Breadth of devices: phones, tablets, carriers
- Development technology: MEAP, native, cross platform
- Load vs. Functionality
- Skill set – devices will require more technical testers and more usability experts
There are a growing number of options based on these needs including crowdsourced testing with Utest, outsourced with companies like InfoStretch, tools that extend existing frameworks and stand-alone device testing solutions such as (DeviceAnywhere, eggPlant, Jamo Solutions, Perfecto Mobile) but this market will continue to evolve, we expect acquisitions and general disruptive behavior for the next 3-5 years.
Category: Mobile Testing Tags:
by Thomas Murphy | June 17, 2011 | 1 Comment
I have been to lots of daily stand-up meetings in many different companies. It is great to see how different Scrum teams work and fit agile practices into their culture and to their needs. I saw a more unique instance this week on a visit to Rally Software. This wasn’t the traditional development team stand-up or even a product manager oriented scrum of scrums. No, this was the executive team daily stand-up. I thought it was great to see that outside of development the company taking the concept and putting it to use and the insight that this provides to the company about one of the core agile processes that the company espouses to its customers. Quick, simple, and efficient the meeting keeps the team on the same page and enables them to communicate with greater authenticity to customers and prospects. I think it is a great pattern for IT teams to socialize within their companies. For agile development to scale it has to work not just at the development layer but across the organization. Get them a Scrum spreadsheet and introduce the daily standup so they can experience the value directly.
Category: Agile ALM Tags:
by Thomas Murphy | April 7, 2011 | Submit a Comment
Companies often wonder about the economic value of testing. For functional testing this generally boils down to defect containment and reducing the cost of fixing defects by finding them earlier but putting performance in economic terms has been more of a challenge. However we continue to see high numbers of sites with performance failures and often during critical times especially in the e-commerce arena. Recently Strangeloop (a site optimization company) built an interesting infographic pointing to the growing performance problem and putting it in economic terms http://bit.ly/e4WCxT
The case for testing is never one dimensional it must be built to the specific needs of the organization e.g. risk, regulation, agility, cost, can all be focus factors as well as understanding the types of issues being produced currently. But seeing the cost to your ecommerce efforts of slow page response is quickly eye opening.
Category: Load/Stress Testing Tags:
by Thomas Murphy | March 16, 2011 | 1 Comment
As organizations increasingly adopt Agile development practices they often do this on top of open source tools for life-cycle (e.g. Git, Subversion, Hudson, Eclipse, Mylin, ApacheJira) and while price is a factor (especially when the Agile team is an unofficial skunk-works) it is the fact the many open source projects are themselves Agile. This means the “development team” is clearly connected to the challenges of distributed agile development and also is attuned to the concept of frequent releases and continuous innovation. These teams don’t have a hierarchy, they believe in transparency, developer productivity is paramount.
This however points out how tools may or may not fit for your company. Open source tools evolve “naturally” without the rigor of planning and there are few open source planning facilities, and the tools themselves are often a patchwork that has to be carefully assembled by the user.
As companies mature in their use of Agile and it moves from dark corners to mainstream development, open source will still be an important driver of innovation but I expect that commercial tools will adapt: more frequent product releases, published defect repositories, greater community involvement to shape directions and ways for the community to extend and often freemium type offerings. These behaviors or the ability to effectively utilize them will separate traditional structured ALM from Agile ALM.
Category: Uncategorized Tags:
by Thomas Murphy | February 24, 2011 | 3 Comments
As we worked through the MQ for Integrated Software Quality Suites one of the key issues we are seeing is that tools are lagging behind the pace of innovation and are often lacking support for platforms and technologies that are becoming popular leaving organizations to perform a lot of manual testing. Key in these areas are lack of support for Apple and its Safari browser (best option is TestPlant) and generally poor support for devices (one good option is InfoStretch). With the continued fast pace of change in client technologies we expect this disparity to continue and most organizations will continue to see a need to have testing tools from multiple vendors.
Category: Mobile Testing Tags:
by Thomas Murphy | January 31, 2011 | 2 Comments
We published the updated MQ for Integrated Software Quality Suites today (see http://bit.ly/hd4reJ ) after several months of work with vendors, users, and community review. This document focuses on vendors who provide (or are working to provide) a complete set of tools for primarily driving the software testing process including test case management, defect management, functional automation, and load/stress testing. The testing tool market continues to be dynamic with many new entrants each year. Most of these are new solutions for a specific area: functional automation, device testing, bug tracking but there are also new players entering the market each year going after more comprehensive solutions.
Software testing continues to be hard, costly, and embattled to explain why it is both of these. It is hard to measure the ROI of testing. However, we will next turn our attention to the ALM market which plays a key element into Quality because “Quality can’t be tested in” some of the players in that document will match this but just as their is growth in quality and testing tools, the ALM market is exploding and our biggest challenge will be focusing in on a reasonable subset of these products. It is great that options are expanding, this means better pricing, improving functionality, and support for a wider variety of technologies. It also means that a decision isn’t as simple as picking from a grid. So if you read the research and have questions about how to apply it to your needs, give me a call.
Category: ALM Quality Testing Tags: Magic Quadrant
by Thomas Murphy | January 7, 2011 | 1 Comment
A few relatively new books have been filling my reading time lately and they offer great content on how to improve the development process and team especially if you are making a shift towards Agile:
Continuous Delivery – this is a great new book on what it really takes to make Continuous Integration work and beyond. If you want to go faster, you have to automate and remove friction from the cycles and this book helps explain how.
The RSpec Book – Solid text that introduces the concepts of Behavior Driven Development using RSpec and Cucumber and drives a refocus of unit testing as a design method – a method designed to create well crafted code and to drive productive interaction between users, developers, and testers in a virtuous cycle of acceptance criteria and unit tests.
Agile Testing – There are many Agile books out there but they are about developer practices, project management and sometimes requirements. Testers have been left hanging in the wind. This book helps define terminology, practices and the shifts a team will go through in the transition from waterfall to agile.
Category: Agile Testing Tags:
by Thomas Murphy | December 29, 2010 | 1 Comment
During his GTAC 2010 talk, Robert Binder covers a lot of interesting material. He ranges through white box and black box tests, the value of automation and movement toward model-based testing paradigms. One of the most interesting elements comes near the end of the talk, around 59 minutes in when we presents a chart on the factors behind testability and who controls those factors (see slide 38 – I suggest downloading to view as Google Docs PPT rendering is poor). If testing productivity is low, it is generally an issue outside the control of the tester as their overall performance is controlled by the work of the architects and developers: size and complexity of the system, structure of the code, testability is driven in before the tester gets his or her hands on the system. If you want to increase testing success therefore, don’t just look at the testers but pay close attention to the guys up-stream.
Category: Testing Tags:
by Thomas Murphy | December 22, 2010 | 1 Comment
I always enjoy the presentations given at Google’s annual Test Automation Conference. One of the presentations I have gone through recently is by Matt Evans of Mozilla discussing their evolving use of crowd sourced testing. He talks about what they have learned to make it successful. I especially appreciated his comments about their use of continuous integration being the foundation of community testing success. Every code change results in approximately 3.3 million tests being automatically executed. This provides the required stability validation that is needed for the community to then pick up each build and hammer on new functionality. Get continuous integration in place and it will hold the arch of development and testing together.
Category: Uncategorized Tags:
