Shmuel Ur - Academia.edu (original) (raw)
Papers by Shmuel Ur
It is our great pleasure to welcome you to the 5th PADTAD Parallel and Distributed Systems: Testi... more It is our great pleasure to welcome you to the 5th PADTAD Parallel and Distributed Systems: Testing and Debugging -- PADTAD07. When we started PADTAD, only four years ago, it was the only workshop dedicated to the testing of concurrent and distributed software. This field is taking center stage as can be seen by the growing number of workshops that are dedicated to it or give it high visibility. The fact that PADTAD topics are of interest to the industry is indicated by our distinguished list of sponsors that includes IBM, Intel, Microsoft, and TotalView Technologies. In addition, companies like IBM, Intel and Microsoft now have dedicated groups to threading testing tools. The mission of PADTAD is to share the advances in the field of testing/debugging concurrency between the industry and the academic world. This sharing is achieved both through the almost even mix of papers from industry and universities, as well as the many attendees from both backgrounds. Our field is such that the impact of research can go both ways. The academic works go very quickly from theory to practice, and the industrial work is cutting edge and publishable. Our distinguished keynote speaker, Prof. Bill Scherlis, Director, Institute for Software Research (SCS/ISR), Carnegie Mellon University, career shows that one can get to the highest levels both in the industry and academic (as well as government agency) working on problems that addresses.
This paper is about t he ex perience gained in IBM Haifa Research Lab (HRL) in creating regressio... more This paper is about t he ex perience gained in IBM Haifa Research Lab (HRL) in creating regression-suites and minimizing their size, while maintaining high quality as measured by coverage. The problem that we solve, while similar to the one addressed in the literature, has a key difference; the c ompaction a lgorithm has to b e implemented on -line
We compare the benefits and costs of using off-the-shelf coverage tools vs. application specific ... more We compare the benefits and costs of using off-the-shelf coverage tools vs. application specific coverage tools. We start with an overview of coverage, its benefits and its risks. We elaborate on relatively unfamiliar functional coverage as a way to create custom made coverage models. We explain what functional coverage models are, how to create them and their great benefits with an example from our experience. We provide guidelines on how to decide if coverage should be used at all and whether code based or functional coverage (or both) should be used. 1 Introduction Testing is one of the biggest problems of the software industry. The cost of testing is usually between 4080 % of the development process (70% for Microsoft) as compared with less than 20% for the coding itself [4]. The practice of letting the users find the bugs and fixing them in the next release is becoming dangerous and costly for three main reasons: reputation and brand-name are harmed, replacing the software...
Computer Aided Systems Theory – EUROCAST 2017
We compare the benefits and costs of using off-the-shelf coverage tools vs. application specific ... more We compare the benefits and costs of using off-the-shelf coverage tools vs. application specific coverage tools. We start with an overview of coverage, its benefits and its risks. We elaborate on relatively unfamiliar functional coverage as a way to create custom made coverage models. We explain what functional coverage models are, how to create them and their great benefits with an example from our experience. We provide guidelines on how to decide if coverage should be used at all and whether code based or functional coverage (or both) should be used.
Testing of concurrent programs is difficult since the scheduling non-determinism requires one to ... more Testing of concurrent programs is difficult since the scheduling non-determinism requires one to test a huge number of different thread interleavings. Moreover, a simple repetition of test executions will typically examine similar interleavings only. One popular way how to deal with this problem is to use the noise injection approach, which is, however, parametrized with many parameters whose suitable values are difficult to find. To find such values, one needs to run many experiments and use some metric to evaluate them. Measuring the achieved coverage can, however, slow down the experiments. To minimize this problem, we show that there are correlations between metrics of different cost and that one can find a suitable test and noise setting to maximize coverage under a costly metrics by experiments with a cheaper metrics.
Concurrency and Computation: Practice and Experience
실제 상황으로부터 시뮬레이션을 시작하기 위한 기술이 개시된다. 다양한 실시형태들에서, 기술은 가상 콘텍스트에서의 물리적 엔티티의 시뮬레이션에 대응하는 파라미터를 수신하고, 물... more 실제 상황으로부터 시뮬레이션을 시작하기 위한 기술이 개시된다. 다양한 실시형태들에서, 기술은 가상 콘텍스트에서의 물리적 엔티티의 시뮬레이션에 대응하는 파라미터를 수신하고, 물리적 콘텍스트 내의 물리적 엔티티에 대응하는 속성을 수신하며, 자극으로서 속성을 이용하여 가상 콘텍스트에서 물리적 엔티티를 시뮬레이팅한다. 속성은 물리적 콘텍스트 동안 실제 발생을 식별할 수도 있다.
A system and method for enabling social dancing. The system comprising a movement registration mo... more A system and method for enabling social dancing. The system comprising a movement registration module configured to monitor and identify movements by a participant; and a tactile feedback actuator configured to supply tactile feedback to the participant based on a determination by a logic module. The method performed by a computerized device, the method comprising: monitoring movements by a participant; determining feedback in response to the movements; and instructing a tactile feedback actuator to supply the feedback to the participant.
It is our great pleasure to welcome you to the 5th PADTAD Parallel and Distributed Systems: Testi... more It is our great pleasure to welcome you to the 5th PADTAD Parallel and Distributed Systems: Testing and Debugging -- PADTAD07. When we started PADTAD, only four years ago, it was the only workshop dedicated to the testing of concurrent and distributed software. This field is taking center stage as can be seen by the growing number of workshops that are dedicated to it or give it high visibility. The fact that PADTAD topics are of interest to the industry is indicated by our distinguished list of sponsors that includes IBM, Intel, Microsoft, and TotalView Technologies. In addition, companies like IBM, Intel and Microsoft now have dedicated groups to threading testing tools. The mission of PADTAD is to share the advances in the field of testing/debugging concurrency between the industry and the academic world. This sharing is achieved both through the almost even mix of papers from industry and universities, as well as the many attendees from both backgrounds. Our field is such that the impact of research can go both ways. The academic works go very quickly from theory to practice, and the industrial work is cutting edge and publishable. Our distinguished keynote speaker, Prof. Bill Scherlis, Director, Institute for Software Research (SCS/ISR), Carnegie Mellon University, career shows that one can get to the highest levels both in the industry and academic (as well as government agency) working on problems that addresses.
This paper is about t he ex perience gained in IBM Haifa Research Lab (HRL) in creating regressio... more This paper is about t he ex perience gained in IBM Haifa Research Lab (HRL) in creating regression-suites and minimizing their size, while maintaining high quality as measured by coverage. The problem that we solve, while similar to the one addressed in the literature, has a key difference; the c ompaction a lgorithm has to b e implemented on -line
We compare the benefits and costs of using off-the-shelf coverage tools vs. application specific ... more We compare the benefits and costs of using off-the-shelf coverage tools vs. application specific coverage tools. We start with an overview of coverage, its benefits and its risks. We elaborate on relatively unfamiliar functional coverage as a way to create custom made coverage models. We explain what functional coverage models are, how to create them and their great benefits with an example from our experience. We provide guidelines on how to decide if coverage should be used at all and whether code based or functional coverage (or both) should be used. 1 Introduction Testing is one of the biggest problems of the software industry. The cost of testing is usually between 4080 % of the development process (70% for Microsoft) as compared with less than 20% for the coding itself [4]. The practice of letting the users find the bugs and fixing them in the next release is becoming dangerous and costly for three main reasons: reputation and brand-name are harmed, replacing the software...
Computer Aided Systems Theory – EUROCAST 2017
We compare the benefits and costs of using off-the-shelf coverage tools vs. application specific ... more We compare the benefits and costs of using off-the-shelf coverage tools vs. application specific coverage tools. We start with an overview of coverage, its benefits and its risks. We elaborate on relatively unfamiliar functional coverage as a way to create custom made coverage models. We explain what functional coverage models are, how to create them and their great benefits with an example from our experience. We provide guidelines on how to decide if coverage should be used at all and whether code based or functional coverage (or both) should be used.
Testing of concurrent programs is difficult since the scheduling non-determinism requires one to ... more Testing of concurrent programs is difficult since the scheduling non-determinism requires one to test a huge number of different thread interleavings. Moreover, a simple repetition of test executions will typically examine similar interleavings only. One popular way how to deal with this problem is to use the noise injection approach, which is, however, parametrized with many parameters whose suitable values are difficult to find. To find such values, one needs to run many experiments and use some metric to evaluate them. Measuring the achieved coverage can, however, slow down the experiments. To minimize this problem, we show that there are correlations between metrics of different cost and that one can find a suitable test and noise setting to maximize coverage under a costly metrics by experiments with a cheaper metrics.
Concurrency and Computation: Practice and Experience
실제 상황으로부터 시뮬레이션을 시작하기 위한 기술이 개시된다. 다양한 실시형태들에서, 기술은 가상 콘텍스트에서의 물리적 엔티티의 시뮬레이션에 대응하는 파라미터를 수신하고, 물... more 실제 상황으로부터 시뮬레이션을 시작하기 위한 기술이 개시된다. 다양한 실시형태들에서, 기술은 가상 콘텍스트에서의 물리적 엔티티의 시뮬레이션에 대응하는 파라미터를 수신하고, 물리적 콘텍스트 내의 물리적 엔티티에 대응하는 속성을 수신하며, 자극으로서 속성을 이용하여 가상 콘텍스트에서 물리적 엔티티를 시뮬레이팅한다. 속성은 물리적 콘텍스트 동안 실제 발생을 식별할 수도 있다.
A system and method for enabling social dancing. The system comprising a movement registration mo... more A system and method for enabling social dancing. The system comprising a movement registration module configured to monitor and identify movements by a participant; and a tactile feedback actuator configured to supply tactile feedback to the participant based on a determination by a logic module. The method performed by a computerized device, the method comprising: monitoring movements by a participant; determining feedback in response to the movements; and instructing a tactile feedback actuator to supply the feedback to the participant.