This paper presents work practice data of the daily activities software engineers. Four separate studies are presented; one looking longitudinally at an individual SE; two a engineering group; and company-wide tool usage statistics. We also discuss advantages in considering practices designing tools for engineers, include some requirements we have developed as result our studies.

As valuable software systems get old, reverse engineering becomes more and important to the companies that have maintain code. Clustering is a key activity in discover better design of or extract significant concepts from an old activity, highly sophisticated, offering many methods answer different needs. Although these been well documented past, discussions may not apply entirely domain. We study some clustering algorithms other parameters establish whether why they could be used for...

Software engineering has been striving for years to improve the practice of software development and maintenance. Documentation long prominent on list recommended practices help Recently however, agile methods started shake this view, arguing that goal game is produce documentation only useful as it helps reach goal.On other hand, in re-engineering field, people wish they could re-document legacy so may continue maintain them or migrate new platform.In these two case, a crucial question...

SUMMARY With the growing need for quality assessment of entire software systems in industry, new issues are emerging. First, because most metrics defined at level individual components, there is a aggregation methods to summarize results system level. Second, evaluation requires use different metrics, with possibly widely varying output ranges, combine these into unified assessment. In this paper we derive, from our experience on real industrial cases and scientific literature, requirements...

Software engineering research now considers that no system is an island, but it part of ecosystem involving other systems, developers, users, hardware, ... When one (e.g., a framework) evolves, its clients often need to adapt. Client developers might adapt functionalities, client systems be adapted new API, users User Interface. The consequences such changes are yet unclear, what proportion the expected react, how long take for change diffuse in ecosystem, do all react same way? This paper...

We discuss how to extract a useful set of subsystems from existing source-code file names. This problem is challenging because many legacy systems use thousands files names, including some that are very short and cryptic. At the same time important software maintainers often find it difficult understand such systems. propose general algorithm cluster based on their alternative methods for implementing algorithm. One key tasks picking candidate words try identify in do this by (a) iteratively...

Decomposing complex software systems into conceptually independent subsystems is a significant engineering activity which received considerable research attention. Most of the in this domain considers body source code; trying to cluster together files are related. We discuss techniques for extracting concepts (abbreviations) from more informal information: file names. The task difficult because nothing indicates where split names substrings. In general, finding abbreviations would require...

Recently, there has been an important progress in applying search-based optimization techniques to the problem of software re-modularization. Yet, a major part existing body work addresses modularizing systems from scratch, regardless packages structure. This paper presents novel multi-objective approach for improving The aims at increasing cohesion and reducing coupling cyclic connectivity packages, by modifying as less possible organization. Moreover, maintainers can specify several...

Software re-modularization is an old preoccupation of reverse engineering research. The advantages a well structured or modularized system are known. Yet after so much time and efforts, the field seems unable to come up with solutions that make clear difference in practice. Recently, some researchers started question whether basic assumptions were not overrated. main one consists evaluating high-cohesion/low-coupling dogma metrics unknown relevance. In this paper, we study real structuring...

Bug prediction is an important challenge for software engineering research. It consist in looking possible early indicators of the presence bugs a software. However, despite relevance issue, most experiments designed to evaluate bug only investigate whether there linear relation between predictor and bugs. it well known that standard regression models cannot filter out spurious relations. Therefore, this paper we describe experiment discover more robust evidences towards causality metrics...

During software evolution, source code is constantly refactored. In real-world migrations, many methods in the newer version are not present old (e.g.,60% of Eclipse 2.0 were 1.0). This requires changes to be consistently applied reflect new API and avoid further maintenance problems. this paper, we propose a tool extract rules by monitoring during system evolution. process, mined at revision level history. Our focuses on mining invocation keep track how they evolving. We also provide three...

As valuable software systems become older, reverse engineering becomes increasingly important to companies that have maintain the code. Clustering is a key activity in used discover improved designs of or extract significant concepts from an old, highly sophisticated, which offers many methods meet different needs. The various been well documented past; however, conclusions general clustering literature may not apply entirely domain. In paper, authors study three decisions need be made when...

To harness the complexity of big legacy software, software engineering tools need more and information on these systems. This may come from analysis source code, study execution traces, computing metrics, etc. One received less attention than code: bugs system. Little is known about evolutionary behavior, lifetime, distribution, stability bugs. In this paper, we propose to consider as first class entities a useful that can answer such topics. Such inherently complex, because are intangible,...

In this paper, we report an experience on using and adapting Semantic Clustering to evaluate software remodularizations. is approach that relies information retrieval clustering techniques extract sets of similar classes in a system, according their vocabularies. We adapted support remodularization analysis. our adaptation six real-world remodularizations four systems. conceptual metrics can be used express explain the intention architects when performing common modularization operators,...