Empirical Project Monitor and Results from 100 OSS Development Projects Masao Ohira Empirical Software Engineering Research Laboratory, Nara Institute of Science and Technology ohira@empirical.jp
collection EASE Project analysis improvement � Empirical software development environment for tens of thousands of projects � Massive data collection � Intensive data analysis � Feedback for software process improvement in organizations/communities (not only a single developer/project) 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 2
Empirical Environment Code clone Component Metrics Project Cooperative detection search measurement categorization filtering EPM （ developing ） Product data archive Process data archive (CVS format) (XML format) Managers Format Format Format Format Translator Translator Translator Translator Issue Versioning Mailing Other tool GUI tracking (CVS) (Mailman) data (GNATS) Project x Project y Project z . . . Developers Widely used development support tools
EPM: Empirical Project Monitor � A partial implementation of Empirical Environment � Collect, measure, and show various data for project control � Data source from tools used in software development � Versioning system (e.g. CVS) � Mailing list manager (e.g. Mailman) � Issue tracking tool (e.g. GNATS) 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 4
Architecture of EPM analysis tools developer Manager measurement of intra and inter projects PostgreSQL(Repository) Standardized empirical SE data （ in XML ） Standardized empirical SE data （ in XML ） developer prediction/schedule manager metrics value versioning problem mail history other tool data history history etc. CVS, Mailman, GNATS (ShareSource TM ) 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 5
Characteristics of EPM � Use open source development tools → Easy to introduce � Small overhead of data collection � Most data from versioning history � Communication through e-mail, and recoding issues by tracking tool � Easy to transform other data format to the standardized empirical SE data format 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 6
Application Area of EPM � Large project � Share project status immediately � Reduce project management load � Reduce risk for tampering data � Small project � Apply with small cost � Apply to various projects, including XP and distributed development 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 7
collection Data collection from analysis improvement OSS Development Projects � SourceForge.net � hosted projects: 72,853 (Dec. 15) � registered Users: 753,428 (Dec. 15) � A variety of collaboration tools � SourceForge Collaborative Development System (CDS) web tools � Project Web Server � Tracker: Tools for Managing Support Available � Mailing lists and discussion forums data source � MySQL Database Services for EPM � Project CVS Services � etc. 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 8
collection Overview of analysis improvement Collected Data � 100 Active projects @ SF.net � Data sources for EPM • CVS data (only 40 projects) • Mailing Lists data • Issue (Bug) reports data � Project info. in a summary page • number of developers • programming language • number of bugs • period of a project • number of CVS commits • development status • etc. • intended audience 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 9
collection SourceForge.net analysis improvement links to available data source for EPM information related to the project 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 10
collection Summary of 100 OSS analysis improvement projects@SF.net: Evolution? 5 0 4 5 4 0 3 5 ? s r e 3 0 p o l e v e D 2 5 t n e r r u 2 0 C 1 5 1 0 5 0 A u g - 9 9 M a r - 0 0 O c t - 0 0 A p r - 0 1 N o v - 0 1 M a y - 0 2 D e c - 0 2 J u n - 0 3 J a n - 0 4 R e g i s t e r e d D a y o f P r o j e c t s 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 11
collection Result of CVS Product Data: analysis improvement Lines of Code (history of software growth) 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 12
collection Result of CVS Process analysis improvement Data: Check in/out (history of developer’s activities) 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 13
collection How can we use such analysis improvement a lot of data? 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 14
collection Gross Classification analysis improvement using EVIDII � EVIDII: Interactive interfaces that visualize relationships among three sets of data (original application domain: face-to-face communication support between clients and designers) 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 15
collection Demo: organizing dynamic analysis improvement community? Project X Project info. numbers of developers, LOC, development terms, etc. 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 16
collection Scenario: organizing a dynamic analysis improvement community / providing feedback for improvement Comparing other projects with a target 1. project Finding similarities and differences 2. between them DynC approach EASE approach 3-a. Notifying to related 3-b. Identifying factors of project leaders of the the similarities and existence of communities differences 4-a. Asking them help/ 4-b. Providing suggestions advices for improvement for improvement 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 17
Summary and Future Work � EPM: Empirical Project Monitor � Data Collection from 100 OSS projects (only 40 CVS data… ) � Two scenarios using EVIDII � More data collection (mails and bug issues) and analysis using EPM/EVIDII 12/21/2003 CDEKA Workshop@UC Irvine, December 16-18, 2003 18
Recommend
More recommend
