1st Day -Track I - 11:30 ~ 12:50

Critical Test Processes: Plan, Prepare, Perform, Perfect
Å©¸®Æ¼Äà Å×½ºÆ® ÇÁ·Î¼¼½º: °èȹ, Áغñ, ¼öÇà, ¿Ï¼º

RexBlack, RBCS (USA)

Users, customers, and stakeholders want systems that provide the needed functions, delivered on time and for a reasonable price. They also want quality systems that work correctly, reliably, securely, and quickly.

Everybody knows that to deliver quality systems, solid testing is a must, right? If everyone knows that, then why did a recent study by the US National Institute of Standards and Technology estimate that inadequate testing cost the US economy $60 billion?

In his book, Critical Testing Processes, Rex Black identifies twelve processes that testers, test managers, and development organization must master to go from knowing they need solid testing to actually doing solid testing. In this talk, Rex will summarize the key ideas of the book for you and give you some ideas and tools you can put to work on your next day back in the office.

Rex will show you why testing matters, in dollars, euros, pounds, or whatever currency you use. Far beyond a hypothetical discussion of how bugs are cheaper when fixed earlier, Rex will show you a real case study from a real project. He'll show you a worksheet you can use to measure what your test process currently saves and how it could save your company even more.

Next, Rex will discuss each of the twelve critical processes. He'll explain what the process should do and why it matters. He'll tell you where to find twelve short, simple checklists you can tailor to your critical testing processes.

Finally, Rex will give you some more ideas and suggestions for taking the next steps to perfect your testing. He'll give you five hard-hitting and in many cases quantifiable questions to ask yourself about your test process. He'll describe the typical challenges that probably inhibit your current test process. He'll close by describing an action plan for transcending those challenges and achieving test process excellence.

1st Day -Track I - 14:10 ~ 15:30

Real Life Stories in Testing
Å×½ºÆ® ¾÷°èÀÇ ½ÇÁ¦ °æÇè´ã
Alon Linetzki, SELA Group (Israel)

In many cases, you remember a presentation from the stories that were told. That is why I have chosen to give only the stories in my presentation, thus leaving a better mark for people to implement. Some call it lessons learned, some call it tips from real life ?at the end of the day, if you learn something that can help you it is adding value to you as a testing professional.

In my presentation I shall share with you short stories about lessons I got while exercising testing as a young engineer and as a test manager:

• In order to add value, listen and learn before you speak
• Getting excellent cooperation from the development managers and team
• Convincing you can test a system without being an expert in that field
• Talk with your facts first

In addition, if possible, we shall spend a short time sharing stories about testing lessons you may have learned.

1st Day -Track I - 15:40 ~ 17:00

Planning and Managing Acceptance Testing
ÀμöÅ×½ºÆÃÀÇ °èȹ ¼ö¸³°ú °ü¸®
Hans Schaefer, Software Test Consulting (Norway)

To put new software systems into service is often a frustrating experience. Important functions may not exist, performance may be low and general system quality lousy. Or it might even be the wrong system and solve your problems. The system may be hard to use and documentation may be bad or non-existing. There are few ways out of this misery once it has happened. Still, it is possible to prevent this. The method is called customer involvement and systematic acceptance testing.

1st Day -Track II - 11:30 ~ 12:50

Case Study: How to Test Web-based Applications?
»ç·Ê¿¬±¸: À¥±â¹Ý ¾îÇ÷¹ÄÉÀÌ¼Ç Å×½ºÆ® ¹æ¹ý
Janet Jing XU, Nankai University (China)

There are three pillars of web-based application testing:

reliability, performance and security. In this presentation three cases are discussed to show how to test web-based applications.

Web-based applications usually provide service for 7*24 hrs, so the system should be reliable enough. In our testing case, we use automation tool to do reliability testing. We design the operation steps including every click buttons and input data.

Performance shows the quality of service. Performance testing is usually executed after reliability testing. The following index should be measured: response time of every transaction; server traffic; workstation traffic; http get-requests/sec; etc. In this testing case, b ased on system response time from sever and client, we can identify where the transaction delay exists. Based on response time and traffic, Nimda Virus is founded.

Web applications are for common service, so everyone can access it. Security is important. The common vulnerability is Parameter Tampering , Cross-sites Scripting (XSS) Flaws , Remote Administration Flaws , Command Injection Flaws , etc. In this presentation we provide two demo tools to detect Parameter Tampering , Cross-sites Scripting (XSS) Flaws .

1st Day -Track II - 14:10 ~ 15:30

Using MindMap for Software Testing Activities
Mind MapÀ» »ç¿ëÇÑ ¼ÒÇÁÆ®¿þ¾î Å×½ºÆà Ȱµ¿
Micky Suzuki, TIS co. & Akira Ikeda, Hitachi Co. (Japan)

Software testing is creative activities. In design phase, the design document may be included a lack of design, error of design. Testers are supposed to find those design bugs, and these finding activities are important.

1st Day -Track II - 15:40 ~ 17:00

Code review and inspection practice in Japan
ÀϺ» Å×½ºÆ®¾÷°è¿¡¼­ÀÇ ÄÚµå ¸®ºä¿Í ÀνºÆå¼Ç »ç·Ê
Nobuhiro Hosokawa, IBM (Japan)

The normal Software Inspection?that generally applied to the software system development was originally advocated by Michael E. Fagan IN 1976, as the defect detection or defect removal technique for software source code or pseudo code.

Despite its fame and popularity in the Japanese market, there has not been enough research or investigation about Software Inspection conducted and published. In general, other studies have concluded that Software Inspection techniques brings so much kind of benefits to practitioners, but little is known about a problem its economy and difficulty of evaluation.

Also, in other aspect, quality of inspection depends on the inspector's skill and experience.

To solve these problems, IBM Japan tried to launch the Quality Inspection (QI) activities for IT services project. QI activities for minimizing project risk and improving deliverables' quality is so cost effective. QI with optimized tools and process brings you the way to prevent defects and trouble project as “Smoke Sensor System?

The purpose of this presentation in ASTA 2007 is to show the importance of quality management in development project through the practical demonstration of our quality inspection tools that measuring metrics from source code, visualize them after then making hypothesis before starting the formal inspection.

1st Day -Track III - 11:30 ~ 12:50

TMM(Test Mutriry Model) in Practice
½ÇÀü¿¡¼­ÀÇ TMM(Å×½ºÆ® ¼º¼÷µµ ¸ðµ¨)À» ÀÌ¿ëÇÑ Å×½ºÆ® ÇÁ·Î¼¼½º Æò°¡
Byunggul Lee, Seoul Woman's University (Korea)

No, Description

1st Day -Track III - 14:10 ~ 15:30

Testing practice on large system with TMap life-cycle model.
¹é¸»ºÒÂ¥¸® °áÇÔÀ» ¹æÁöÇϱâ À§ÇØ TMapÀ» »ç¿ëÇÑ Å×½ºÆ® »ç·Ê
Yongtae Jung, STAV (Korea)

This presentation will present a specialized domain in Port operating system industry. In this field, most of equipments are valuable including vessels. When critical defects are found, impaction of business is great like a loss of million dollar or even billion dollar when it¡¯s on vessel balance program, Crain control program, Rail module with trains. Therefore the testing is one of most essential activity in this field. Presentation covers introduction of domain knowledge and TMap which is widely used life-cycle model.

1st Day -Track III - 15:40 ~ 17:00

Nightly incremental static analysis and report system
¾ß°£¿¡ Á¡ÁøÀûÀ¸·Î ¼öÇàÇÏ´Â Á¤ÀûºÐ¼® ¹× ¸®Æ÷Æà ½Ã½ºÅÛ
Jaehwang Kim, LG Electronic (Korea)

We made automatic defect detection and reporting system by integrating a configuration management tool, a static analysis tool and a issue tracking system. By analyzing daily changed set of source code we can find bugs earlier so that we can provide developers more time to fix those bugs.

he automatic defect detection and reporting system is highly efficient in dealing with millions lines of code our target programs have. It supports incremental analysis which means we analyze files in the changed set. It reuses the result of previous analyses so as to make incremental analysis have the same effect of whole program analysis. We could save more than 95% of analysis time by incremental analysis. It used to take at least one day to analyze one program and report defects to developers. But now it takes less than 30 minutes to analyze a program and report defects.

2nd Day -Track I - 10:00 ~ 11:00

Risk based Testing by IBM
IBM ¸®½ºÅ© ±â¹Ý Å×½ºÆÃ
Dongyoung Kim, IBM (Korea)

Structured approach to software testing is critical for the success of the development project. Risk based testing is located at the core of the structured testing and guide the overall testing effort and practice. It gives the big-pictured guidance almost whenever the test organizations need to decide on testing issues. Especially risk based testing is tightly related to the test case derivation (use of test design technique). Test case derivation without considering the risk based testing strategy may deteriorate the completeness of the testing practice.

In this presentation we will look into the risk analysis and testing strategy practically and show how the test case design technique is used in the aligned context based on our test consulting services offered to our client.

2nd Day -Track I - 11:30 ~ 12:50

Using SCRUM as a Test Management method
SCRUMÀ» »ç¿ëÇÑ Å×½ºÆ® °ü¸®±â¹ý
Klaus Olsen, Softwaretest.dk (Denmark)

Using Scrum is a very strong test management method to be aware of what your test team is working on, and at the same time, get the best of you team members.

Scrum uses a Product Backlog, Daily Meeting, and a Burndown charters as some of the vehicles to an improvement performance when delivering software.

In this presentation Klaus Olsen will explain how these tools also are of value when we work as Test Managers.

2nd Day -Track I - 14:10 ~ 15:30

Test Plan 101 using IEEE 829
IEEE829¸¦ »ç¿ëÇÑ Å×½ºÆ® °èȹ 101
Bernard Homes, Tessco (France)

This presentation first describes the different types of documents used in the IEEE 829 standard and their relations to one another.

Building on this information it details the different steps and aspects of the test plan as described in the standard and the relations between the different chapters, how they complement each other in order to provide a complete solution.

From there it shows how to use the plan for different types of tests such as functional / regression testing, performance testing and configuration tests.

The presentation continues with the application of sub-plans to cover the different aspects of testing for a complex application.

The presentation also covers some of the evolutions in the future version of IEEE 829, so that the audience can understand where this standard is heading to in terms of mandatory and informational aspects.

This presentation is based on the experience of the presenter in different test plans evaluation and development, both using the IEEE 829 standard and other standards, thus ensuring the audience acquired practical and useful knowledge;

This presentation also builds upon the participation of the presenter in the IEEE 829 standard workgroup, enabling the audience to have a first look at the evolutions of the standard.

2nd Day -Track I - 15:40 ~ 17:00

eXtreme Testing meets Risk Based Testing
ÀͽºÆ®¸² Å×½ºÆðú ¸®½ºÅ© ±â¹Ý Å×½ºÆÃÀÇ ¸¸³²
Yaron Tsubery, Comverse (Israel)

Many (if not all of us) are searching for the best ways to do their job in order to improve the performance and achieve more. Yes, this is the era which we are leaving in ? many results (which always mush be the best result) shell be presented in a very short time ? what business persons call TMM (Time to Market). One of our task as test engineers or as test managers are to find the path(s) that will lead to effective result based on efficient processes; in the 1 st case we are searching for results in terms of better and improved test coverage and in the 2 nd case we are searching for large scale, efficient and effective defect detection.

Many was suggested and visible for implementation; few methods developed for that and my objective is to present the combined process which will show a way of implementation that accelerates critical defects detection. This process developed due to customers demands; in fact it is based on real projects practice. Main methods which will be sentenced in the presentation are: risk-based testing, quality risk analysis and exploratory testing; all of them combined into one suggested process that will enable the test engineer and the manager to accelerate defects detection. The practice is focused on projects of complex systems, delivered to telecommunication companies under restricted rules and stiff exit criteria elements ? which made this more complex to handle among with tense delivery timelines

2nd Day -Track II - 11:30 ~ 12:50

Noun and Verb Technique
¸í»ç¿Í µ¿»ç¸¦ ÀÌ¿ëÇÑ Å×½ºÆ® ÄÉÀ̽º ±â¹ý
Vipul Kocher, Puretesting (India)

This paper presents an Extension of Noun and Verb Technique. This technique can help create test cases ranging in complexity from simple to very complex. The technique allows creation of scenario based tests as well as individual feature/field validation tests. The technique has successfully been used on various projects with great results.

2nd Day -Track II - 14:10 ~ 15:30

Creating a Company Wide Measurement Program for Verfication
VerificationÀ» À§ÇÑ Àü»çÀûÂ÷¿øÀÇ ÃøÁ¤ÇÁ·Î±×·¥ ÀÛ¼º
Sigrid Eldh, Ericsson AB (Sweden)

Some of our company wide measurements have challenged the testing within Ericsson. Our goal to become 50% more efficient forces you to think in different ways. What does 50% really mean? How can we save time and money and still keep quality in our products? In this presentation, we will show some of our important measurements we use to improve and assess testing within Ericsson. We will address different ways of measuring testing - and quality on software - and how we handle the wide variety of products.

2st Day -Track II - 15:40 ~ 17:00

Forth Generation Test Strategy: A Quicker Test for Death March Projects
4¼¼´ë Å×½ºÆ® Àü·«: Á×À½ÀÇ ÇàÁø ÇÁ·ÎÁ§Æ®¸¦ À§ÇÑ ½Å¼ÓÇÑ Å×½ºÆ® Àü·«
Tsuneo Yamaura, Tokai University (Japan)

Since the year of 2000, or after the Y2K problem passed, due to the wide and rapid dissemination of embedded systems and web-based development, quicker release to the marketplace is a crucial issue for a software development project.

This presentation presents an example of a quicker test with retaining the verification quality, whose basic idea is (1) debugging and testing are overlapped to reduce the time, (2) a quality development manager (a tester who manages programmers for quality assurance) primarily orchestrates the quality assurance activities in the quality assurance.

1st Day -Track III - 10:00 ~ 11:00

Continuous Integrated Testing (CIT): Test Early, Test Often with Confidence
Áö¼ÓÀû ÅëÇÕ Å×½ºÆÃ: È®½ÅÀ» °¡Áö°í Ãʱ⿡ ÀÚÁÖ Å×½ºÆÃÇÏ´Â Á¢±Ù¹ý
Hyunsoo Kim, Compuware (Korea)

Today, most IT organizations still treat Quality Assurance as an end-of-life-cycle event. Why is this? For one, the developer¡¯s time is precious. So most developers aren¡¯t asked to do any more testing than is necessary. In addition, most QA vendors and solution providers provide testing tools for testers ? not developers. It¡¯s also easier for the QA organization to test everything at one time once they receive the completed code from development.

For many good reasons, this approach remains in place today ? though it results in other challenges for IT. When defects are found late in the development life cycle, the cost of rework becomes significant. This cuts testing time and more importantly, costs IT and the business in terms of lost development time for new innovative projects.

CIT allows more testing to be performed and this testing is performed more quickly and more accurately. This is accomplished through an automated nightly build that performs unit and functional testing. By capturing manual testing efforts into a set of reusable test scripts, tests can be executed ? running unattended whenever needed. When combined with testing tools that provide line-of-code root cause analysis, code can be automatically assessed for memory leaks, poorly designed code, code coverage, poor performing code, and even security vulnerabilities. Pulling these two processes together allows defects to be found quickly ? with minimal developer intervention. Root cause analysis ? augmented with system-recommended corrections ? reduces the time spent on problem diagnosis by both the QA and development teams. Overall, higher quality code is sent to the QA organization.

2nd Day -Track III - 11:30 ~ 12:50

Lesson learned from the survey on Software Process Capability and K-model as the Goverment Certification
Sangeun Lee, KIPA & Wonil Kwon, STA (Korea)

K-model is a guidance to assess the process of software supplier. The major objective of this guidance is to evaluate and certify the capability level of software suppliers which participate in software projects of public sector. But the ultimate goal of this guidance is to support software supplier to establish and improve its own software process. K-model consists of three major parts; the Assessment Model, Measurement Framework, and Certification Process. Assessment Model has 92 measurement items to evaluate each activity in 17 major processes. Measurement Framework consists of Process Capability Level (CAL), Process Certification Level (CEL) and Rating. CAL is defined on three point scale and CEL is defined on two level scale; good and excellent. This assessment model reflects characteristics of Korean software industry based on ¡®the study on process capability level in software industry¡¯ which has been conducted annually. The model is designed to guarantee the fairness and objectivity during the assessment because it can be the guidance for the selection of software supplier. It also provides the check list template to simplify the assessment and reduce the evaluation term and cost. Therefore the model can be fit to small and medium size company especially.

2nd Day -Track III - 14:10 ~ 15:30

Unit testing best practice @TDD (Test Driven Development)
TDD(Å×½ºÆ® ÁÖµµ°³¹ß)¿¡¼­ÀÇ ´ÜÀ§Å×½ºÆà Best Practice
ChangJoon, Kim, Agile Consulting (Korea)

Unit testing in TDD has different roles from its general acceptance in SE. We will mainly focus on the evolutionary aspect of unit tests.

2nd Day -Track III - 15:40 ~ 17:00

Introduction of Product Evaluation Service for Quality Improvement and Case study
Ç°Áú°³¼±À» À§ÇÑ Á¦Ç°Æò°¡ ¼­ºñ½º¿Í »ç·Ê¼Ò°³
Jaewoo Bae, LG CNS (Korea)

We have experienced the downward equalization of the q uality level of product as the turnkey project increases. It was mainly due to the uncontrolled quality assurance. The improved quality from early phase in SDLC is needed to overcome this problem.

Product Evaluation Service evaluates the software product in end-user ' s view to increase customer ' s value, and c onduct objective evaluation by using quantitative indicator with same evaluation factor to verify quality level of project.

Product Evaluation Service provides quantitative quality level of product in end-user ' s view, detects and fixes defect in early phase in SDLC, and reminds the weakness in quality.

Product Evaluation Service uses automation tools that eliminate simple and repeated work to improve test productivity.