1. 8 January 2006 - 1 - © 731c[charts-mml] Assumptions in Software Development and Evolution Meir (Manny) Lehman School of Computing Middlesex University The Burroughs] London NW4 4BE, UK mml@mdx.ac.uk http://www.cs.mdx.ac.uk/staffpages/mml Abstract: A brief summary of the results of many years of study of E- type system evolution leads to a focus on the inevitability and crucial role of assumptions and their potential impact. The societal and economic consequences of failure in this respect clearly indicates that their adequate management is increasingly vital. January 2006

2. 8 January 2006 - 2 - © 731c[charts-mml] Facts of Life Effective planning, management, control of software evolution a major challenge but potential for significant returns Long known that 50% to 95% of E-type software life cycle costs incurred on maintenance after introduction into use, that is on evolution Recent RAEng/BCS report quoted study predicting UK 2003/4 £B22.6 software expenditure - that of other countries will be similar though modified by size of economy Will increase as problem and environmental complexity, inter-system integration and communication, operational, functional, data detail grow Human and indirect costs in addition Hence UK expenditure in software evolution that year between £B11.3 and £B21.5 - probably (O) £B15 Other economies similar, though dependant on technological, economic levels

3. 8 January 2006 - 3 - © 731c[charts-mml] Opportunities for Process Improvement First, brief historical overview of basic observations Major potential benefit from development of understanding of, discipline, methods and tools for evolution planning and management Potential UK annual saving from improvement of only 10% could yield up to £b2 benefit on direct software costs alone Human, social, economic saving potential many times that

4. 8 January 2006 - 4 - © 731c[charts-mml] Practical implications - Rules and Tools for Software Evolution Planning, Management and Control, Annals of Softw. Eng. Spec. Iss. on Softw. Maint., v. 11, 2001, pps. 16 - 44 Historical Review - 1968 - 69IBM Programming Process study the seed -1969 - 71First recognition of the role of feedback -1969 - 02Collection, analysis of empirical data on evolution of industry-developed systems -1974 - 86Laws of Software Evolution -1979, 04SPE program classification schema leading to -1988Key role of assumptions -1989Principle of Software Uncertainty -1996 - 01FEAST/1 and /2 projects -2004SPE+ Laws were first step in formulating theory Basic observation: evolution, particularly of software, inherent to real world use of computers - not an indicator of weakness or incompetence Significant elements of an emerging theory of software evolution with feedback key mechanism, role and impact of assumptions key factor

5. 8 January 2006 - 5 - © 731c[charts-mml] Status of Laws Formulation of feedback observation as eighth law suggested that first seven are related to feedback properties Subject to extension and generalisation Today's focus: presence and effect of assumptions in real world computing Originally derived individually by observation; relationships between them not investigated That is, feedback-systems theory significant component of software evolution theory

6. 8 January 2006 - 6 - © 731c[charts-mml] Computing in the Real World System acceptability depends on stakeholder satisfaction with results of and behaviour in execution E-type software, systems, applications - address problems, activities and/or operate in real world domains Implication: E-type systems can never be guaranteed fault free - that is, entirely satisfactory or utterly reliable Correctness second level issue

7. 8 January 2006 - 7 - © 731c[charts-mml] Fundamental Requirements To produce acceptable results, system must reflect every property of application, domains to level of precision, detail as at time of concern Must continue to hold as domains, application properties and expectations change Evolution inescapable for provision of continuing useful service Software must be continually adapted to maintain stakeholder satisfaction Such change inevitable since real world undergoes continuing change, in part a consequence of system installation and use

8. 8 January 2006 - 8 - © 731c[charts-mml] Intermediate Summary In any event, software evolution inescapable phenomenon intrinsic to E-type systems Real world dynamic Without evolution there will be inevitable decline in validity, functionality, utility, performance Application and domain changes maintains system in tune with real world needs, states, opportunities, ambitions More generally evolution exploited to increase them

9. 8 January 2006 - 9 - © 731c[charts-mml] Human Development Properties of E-type domains and applications are unbounded in number Moreover, humans can consider only bounded number of product properties in design and construction Essential uncertainty whether any artefact is complete or correct Since human knowledge, understanding, mastery of real world properties is bounded they cannot consider, determine, even know relevance of all these

10. 8 January 2006 - 10 - © 731c[charts-mml] The Consequence Principle of Software Uncertainty follows Hence every E-type program reflects unbounded number of assumptions - that rejected properties are irrelevant to achieving satisfactory execution Design and evolution activity involves abstraction, consciously rejecting or implicitly ignoring unbounded number of application and domain properties Unbounded number of real world properties rejected or ignored in system development, evolution

11. 8 January 2006 - 11 - © 731c[charts-mml] However often a system has executed satisfactorily, satisfaction on its next execution is uncertain Principle

12. 8 January 2006 - 12 - © 731c[charts-mml] Source of Uncertainty Hypothesis: assumptions that are or become invalid major source of project or computer failure during development and use Unbounded set of excluded properties may include such as should have been initially included or which such changes eventually make relevant Application and domain changes will falsify adopted assumptions set

13. 8 January 2006 - 13 - © 731c[charts-mml] Examples of Initial or Eventual Invalid Assumptions that Led to Major Problems CERN Accelerator - unconscious, but valid, assumptions during development eventually became invalid London Ambulance System - invalid design assumption First launch of Ariane V - conscious and valid operational application and domain (hardware) assumptions Sinking of the Sheffield (18 lives) - initially valid, predictably invalid, application assumption Daily examples on Google Alerts for Software Failure provide major source of material for further investigation

14. 8 January 2006 - 14 - © 731c[charts-mml] Assumptions and their Implications Real world changes during use cause invalidation of assumptions Assumptions embedding in system during abstraction, reification, bounding intrinsic to E-type computing system development, evolution, usage Assumption management vital but much neglected development and maintenance responsibility Assumptions may be explicit or implicit, conscious or unconscious, by commission or omission, recorded or unrecorded, (initially) valid or invalid Continual search for and review of assumptions to ensure satisfactory, that is safe, reliable, efficient, controlled computer usage

15. 8 January 2006 - 15 - © 731c[charts-mml] What is Software Maintenance? Software maintenance maintains validity of assumption set and stakeholder satisfaction

16. 8 January 2006 - 16 - © 731c[charts-mml] Theory of Software Evolution Its development non-trivial but development could have major impact Have outlined base and framework for development of theory of software evolution Theory a powerful tool to reduce cost and threat implied by need for continual evolution; need with major technical, economic, social implications Theory can provide direction and discipline to conception, development, integration of methods and tools; software evolution process improvement

17. 8 January 2006 - 17 - © 731c[charts-mml] Reservations Observations, hypotheses, generalisations such as laws and principle all based on data from conventional industrial processes Urgent need for widespread R & D to determine evolution characteristics as function of these and other variables Some studies of Open Source processes appear to suggest other growth patterns but support conclusion that software evolution is predictable phenomenon Not aware of studies of evolutionary behaviour of systems developed using other approaches such as Object Oriented, Open Source, Agile programming But universal and process independent domain concepts, behaviours, in particular feedback, assumptions, suggest that phenomenon is universal - independent in principle, though not necessarily detail, of application, domain, process, degree of system integration, development organisation and so on

18. 8 January 2006 - 18 - © 731c[charts-mml] Areas Ripe for R & D, Industrial Application Likelihood analysis of changes and their impact on application and domains Areas and extent of reviews required to ensure timely changes in advance of failure Search for guidelines to minimise likelihood of assumption invalidation, maximise structural flexibility, increase changeability, reduce cost of system changes Structured, indexed, mechanised data-base system for assumption management Periodic and event-triggered reviews to update assumption sets at all stages of system definition, development, evolution Examples

19. 8 January 2006 - 19 - © 731c[charts-mml] Key R & D Challenges in Assumption Management, Control In general, assumption management and control at every and all stages of development, evolution Examples in specific process activities Determine, develop and, where possible, formalise method and tool support for:- Periodic and, event-triggered reviews to update assumption set, its validity and evaluation of its exposure to and likelyhood of change Identifying, questioning assumptions in problem and requirements definition architecture, design, reviews, implementations, inspections, validation etc.

20. 8 January 2006 - 20 - © 731c[charts-mml] Assumptions Management Develop and use tool support for the above and for system adaptation In general: search for, examine, control assumptions in all activities Good practice requires one to:- Identify, capture, structure, document, update rationale, assumptions, decisions underlying them with reviews and revalidation whenever changes occur in application, domains Institute periodic and event-triggered assessments to anticipate or identify changes in assumption set and system changes they may require Extend process to make search for and questioning of assumptions in reviews, inspections, validations etc. at each step a specific and required activity Improve questioning of assumptions by, for example, using independent specification, implementation, validation teams

21. 8 January 2006 - 21 - © 731c[charts-mml] Further Example: Requirements Whatever process is followed, all activities must include disciplined examination of possible impact on assumptions, existing or implied, potential consequences General conclusions Explicit and implicit assumptions at each stage of entire process Must explicitly address and rule on issues such as: Continually determine timings for review of all aspects of system to ensure timely changes, forestall undesirable behaviour Linkage of architecture, process, design, code, documentation, procedures, etc to requirements to simplify change and minimise assumption dependent errors Full structured, indexed documentation to facilitate organised, timely searches for missing or invalidated assumption statements Likelihood and potential impact of changes in problem and domains

22. 8 January 2006 - 22 - © 731c[charts-mml] In Summary Relate general and specific behavioural patterns and characteristics to application areas, domains, processes, organisation, cultures etc. Must recognise inevitability of evolution and accept need for related social, economic, operational effort and expenditure Develop and apply understanding of evolution to support development of processes, methods, tools for, for example, its and assumption management

23. 8 January 2006 - 23 - © 731c[charts-mml] Relevant Papers Complete listing provided at http://www.cs.mdx.ac.uk/staffpages/mml/http://www.cs.mdx.ac.uk/staffpages/mml/ Rules, Tools and Guidelines MM Lehman, Rules and Tools for Software Evolution Planning and Management, pos. paper, FEAST 2000 Workshop, Imp. Col., 10 - 12 Jul. 2000, DoC, Res. Rep. Nov 2000, a revised version to appear in Annals of Software Engineering, Spec. Issue on Software Management, vol. 11., 2001 MM Lehman, The Future of Software - Managing Evolution, inv. contr., IEEE Software, Jan-Feb. 1998, pp. 40-44 Theory MM Lehman and JF Ramil, Towards a Theory of Software Evolution - And Its Practical Impact, invited talk, ISPSE 2000, Intl. Symposium on the Principles of Software Evolution, Kanazawa, Japan, Nov 1-2, 2000 Process Modelling G Kahen, MM Lehman, JF Ramil and PD Wernick, Dynamic Modelling in the Investigation of Policies for E-type Software Evolution, ProSim 2000, International Workshop on Software Process Simulation and Modelling, 12 - 14 Jul. 2000, London, UK BW Chatters, MM Lehman, JF Ramil, P Wernick, Modelling a Software Evolution Process, ProSim'99, Softw. Process Modelling and Simulation Workshop, Silver Falls, Oregon, 28-30 Jun. 1999, also as Modelling a Long Term Software Evolution Process in J. of Softw. Proc.: Improv. and Practice, 2000 v. 5, iss. 2/3, Jul. 2000, pps. 95-102 MM Lehman, DE Perry and JF Ramil, On Evidence Supporting the FEAST Hypothesis and the Laws of Software Evolution, Proc. Metrics'98, Bethesda, MD, 20-21 Nov. 1998, pp. 84-88 P Wernick and MM Lehman, Software Process White Box Modelling for FEAST/1, ProSim '98 Workshop, Silver Falls, OR, 23 Jun. 1998. As a revised version in Journal of Systems and Software, Vol. 46, Numbers 2/3, 15 Apr. 1999 MM Lehman and P Wernick, System Dynamics Models of Software Evolution Processes, Proc. Int. Wrkshp. on the Principles fo Software Evolution IWPSE-98, ICSE-20, 20-21 Apr. 1998, Kyoto, Japan, pp. 6-10 MM Lehman, DE Perry, JF Ramil, WM Turski and P Wernick, Metrics and Laws of Software Evolution - The Nineties View, Proc. Fourth International Symposium on Software Metrics, Metrics 97, Albuquerque, New Mexico, 5-7 Nov. 97, IEEE Comp. Soc. or. n. PR08093, pp 20-32. Also as in K El Eman and N H Madhavji (eds.), Elements of Software Process Assessment and Improvement, IEEE CS Press, 1999 WM Turski, A Reference Model for the Smooth Growth of Software Systems, IEEE Trans. Softw. Eng., v. 22, n. 8, Aug. 1996, pp. 599 - 600 Estimation JF Ramil and MM Lehman, Exploring Cost Estimation Models in the Context of Continuing Software Evolution, FESMA-AEMES Software Measurement Conference 2000 "Management Excellence through IT Measurement", Madrid, Spain Oct. 18-20, 2000 Lessons Learnt (e.g., Modelling Methodology) JF Ramil and MM Lehman, Metrics of Software Evolution as Effort Predictors - A Case Study, Proc. ICSM 2000, Int. Conference on Software Maintenance, 11-14 Oct. 2000, San Jose, CA Component-based and Integration-intensive Processes Other MM Lehman and JF Ramil, Software Evolution Phenomenology and Component Based Software Engineering, IEE Proc. Softw., sp. issue on Component Based Software Engineering, scheduled for Dec. 2000, earlier version as Tech. Rep. 98/8, Imperial College, London, Jun. 1998 JF Ramil (ed.), Preprints of FEAST 2000 International Workshop on Feedback and Evolution in Software and Business Process, Imp. Col., London, 10 - 12 Jul. 2000, 124 pp. JF Ramil, MM Lehman and G Kahen, The FEAST Approach to Quantitative Process Modelling of Software Evolution Processes, Proc. PROFES'2000 2nd International Conference on Product Focused Software Process Improvement, Oulu, Finland, 20 - 22 Jun. 2000, in Frank Bomarius and Markku Oivo (eds.) LNCS 1840, Springer V, Berlin, 2000, pp. 311 - 325. MM Lehman, Feedback in the Software Evolution Process, Keynote Address, CSR Eleventh Annual Workshop on Software Evolution: Models and Metrics. Dublin, 7-9 Sept. 1994, Workshop Proc., Information and Software Technology, sp. is. on Software Maintenance, v. 38, n. 11, 1996, Elsevier, 1996, pp. 681 - 686

24. 8 January 2006 - 24 - © 731c[charts-mml] Bibliography - Lehman et al A listing with additional papers and other material may be accessed via http://www.cs.mdx.ac.uk/staffpages/mml/. References indicated with an ‘ * ’ reprinted in Lehman MM & Belady LA, Program Evolution—Processes of Software Change, Acad. Pr., London 1985 and now available on webhttp://www.cs.mdx.ac.uk/staffpages/mml/ Lehman MM, Kahen G and Ramil JFl,, Simulation Process Modelling for Managing Software Evolution,in Juristo N. and Acuna S (eds.), Software Process Modelling, Kluwer Ac. Pr. Lehman MM and Ramil JF, Theory of Software Evolution - Background, Theory, Practice, Seventh World Conf.on Integrated Design and Process Tech., Austin, TX, Dec. 3 rd – 6 th, 2003, id., Towards a Theory of Software Evolution - Background, Theory, Practice, Info. Proc. Letters, v. 88, 2003, pp. 33 - 44 id., Software Evolution - Background, Theory, Practice, IDPT, TX, Dec 2003 Lehman MM, A Formal Theory of Software and Artificial Systems Evolution, Pos. Pap., Worksh. on Grand Challenges for Computing Research, EPSRC proposal, 17 Oct. 2002, 2 ps id., Software Evolution,, Encyclopaedia of Software Engineering, J Marciniak (ed), 2 nd ed. Wiley and Co, 2 vols., 2002, pp.1507–1513, revised and extended version of mml507 Lehman MM and Ramil JF, Some Lessons Learnt in and from the FEAST Project, WESS 2002, Montreal, 2 Oct 2002 id., Software Uncertainty in General and in KBS Applications in Particular, 9th Int. Conf. on Info. Proc. and Man.of Uncertainty in KBS IPMU 2002, Annecy, France, 1-5 Jul. 2002 Lehman MM, Software Uncertainty, Proc. Soft-Ware, 1st Int. Conf. on Comp. in an Imperfect World, Interactive Science and Tech. Conf. Centre, Belfast, Northern Ireland, 8-10 Apr. 2002 Lehman MM, Kahen G and Ramil JF,, Behavioural Modelling of Softw. Proc. of Long Lived Evol.Proc. –Issues and Exa. ICSTM, J. Softw. Maint. and Evol., v 14, 2002, pps. 335 – 351 Lehman MM,, Software Evolution and Software Evolution Processes, Inv. Contr. sp. iss. on Proce.-based Softw. Eng, Ann., Softw. Eng. v. 14, 2002, pps. 275 – 309 Nov. 2002, Lehman MM and Ramil JF,, SEPiCS – Evolution Phenomenology in Component-intensive Software, WESS 2001, Florence, 9 Nov. 2001 id., Is Evolution Intrinsic to All Real World Software?, EPSRC Network on Evolvability in Biological & Software Systems, Proc. Symp. "Software Evolution and Evolutionary Computation", (SEEC), U. of Herts. Workshop, 7 – 8 Feb. 2001 Lehman, JF Ramil and U Sandler,, An Approach to Modelling Long-term Growth Trends in Software Systems,, Proc. ICSM 2001, 6-10 Nov., Florence, Italy, Nov. 2001 Lehman MM, Kahen G and Ramil JF,, Thoughts on the Role of Formalisms in Studying Software Evolution, Proc. of Worksho. on Formal Foundations of Softw. Evolution, 13 Mar. 2001, Lisbon, Portugal, 8 pp., in Mens T. and Wermelinger M (eds.) Tech. Report UNL-DI-1-2001, Departamento de Informática, Universidade Nova de Lisboa, 8pps., Evolution and its Impact on Business Processes, Technology Outreach, EPSRC/DTI workshop, London, 15 Mar. 2001 Lehman MM and Ramil JF, iSoftware Evolution, wProc. IWPSE 2000 Lehman MM, Kahen G and Ramil JF,, Software Evolution: Phenomenon, Metrics, System Dynamics, with JF Ramil and G Kahen, Managing Systems Evolution: Software Systems Lehman MM and Ramil JF, FEAST/2 Workshop - Introduction and Overviews, ICSTM, 19 Dec. 2000 Lehman MM., Models and Modelling in Software Engineering, Encycl. of Softw. Eng., J Marciniak (ed), Wiley and Co, 1994, vol. 1, pp. 698 - 702 (Revised ed. 2002) id., Software Evolution, loc cit, vol. 2, pp. 1202 - 1208 id., Uncertainty in Computer Application, Comm. of the ACM, vol. 33, No. 5, May 1990, pp. 584-586 id., Uncertainty in Computer Application and its Control through the Engineering of Software, J. of Softw. Maint., Res. & Pract., v. 1, 1 Sept 1989, pp. 3 - 27 Lehman MM, Process Models, Process Programs, Programming Support, Inv. Resp. To A Keynote Addr. by Lee Osterweil, Proc. 9th Int. Conf. on Softw. Eng., Monterey, CA, 30 Mar. 2 Apr 1987, IEEE Comp. Soc. pub. n. 767, IEEE Cat. n. 87CH2432-3, pp. 14 - 16 id,. Evolution - The Cause of Iteration, Third Process Workshop, Proc. 3rd Int. Process Wrkshp., Dowson M (ed), IEEE Comp. Soc. Press, Mar. 1987, pp. 29 - 32 Lehman MM & Belady LA, Program Evolution,- Processes of Software Change, Acad. Pr,,London, 1985, 538 p. Lehman MM, Stenning V & Turski WM, (1984). Another Look at Software Design Methodology, Software Engineering Notes, v. 9, no 2, Apr 1984, pp. 38 - 53 *id., Programs, Life Cycles and Laws of Software Evolution, Proc. IEEE Sp. Iss. on Softw. Eng., v. 68, n. 9, Sept 1980, pp. 1060 - 1076 *id., Laws of Program Evolution - Rules and Tools for Programming Management, Proc. Infotech State of the Art Conf, Why Software Projects Fail, - Apr 9 - 11 1978, pp. 11/1 - 11/25 id., Human Thought and Action as an Ingredient of System Behaviour, Contribution to the Encyclopaedia of Ignorance, July 1976, (Pergamon Press 1977) pp. 397-354 *Lehman MM Programs, Cities, Students - Limits to Growth?., Imp. Col. Inaug. Lect. Sers, v. 9, 1970 - 1974, and [GRI78], pp. 42 - 69, Lehman MM & Belady LA, 1985, pp. 133 - 163, *Belady LA & Lehman MM, An Introduction to Program Growth Dynamics, in Stat.Comp. Perf. Eval, W Freiburger (ed), Acad Press, N,Y. 1972, pp. 503 - 511 *Lehman MM, The Programming Process, IBM Res. Rep. RC 2722, IBM Res. C,, Yorktown Heights, NY, Sept. 1969

25. 8 January 2006 - 25 - © 731c[charts-mml] Bibliography - Others A first listing with additional papers and other material is provided at and some may be accessed via http://www.cs.mdx.ac.uk/staffpages/mml/http://www.cs.mdx.ac.uk/staffpages/mml/ Needs to be updated Boehm BW, Software Engineering, IEEE Trans. on Comp., v. C-5, n. 12, Dec. 1976, pp. 1226 - 1241 id., A Spiral Model of Software Development and Enhancement, Computer, v. 21, May 1988, pp. 61 - 72 id., Software Engineering Economics, Englewood Cliffs, N.J, Prentice-Hall, 1981 Brooks FP, No Silver Bullet - Essence and Accidents of Software Engineering, Information Processing 86, Proc. IFIP Congress 1986, Dublin, Sept. 1-5, Elsevier Science Pubs. (BV), (North Holland), pp. 1069 - 1076 Osterweil L, Software Processes are Software Too, Iteration in the Software Process, Proc. of the 3rd Int. Proc. Worksh., Breckenridge, CO, 17 - 19 Nov. 1986, IEEE cat. n. TH0184-2, IEEE Comp. Soc. order n. 709, 1987, pp. 79 - 80 Perry DE, Policy and Product-Directed Process Instantiation, Proc. of the 6th Int. Softw. Process Workshop, 28-31 October 1990, Hakodate, Japan Rajlich VT and Bennet KH, A Staged Model for the Software Life Cycle, Computer, July 2000, pp. 66 - 71 Turski WM, And No Philosophers' Stone Either, Inf. Processing 86, Proc. IFIP Congr., Dublin, Sept. 1 - 5, 1986, Elsevier Sci. Pubs, London, pp. 1077 - 1080 Wilkes M V, Wheeler D J & Gill S, The Preparation of Programs for an Electronic Digital Computer, Addison Wesley Press Inc., 1951, 167 pp. Wirth N, Program Development by Stepwise Refinement, CACM, v.14, n.4, Apr 1971, pp.221-227 *Woodside CM, A Mathematical Model for the Evolution of Software, J. of Sys. and Softw. vol. 1, no. 4, Oct 1980, pp. 337 - 345 Zurcher FW and Randell B, Iterative Multi-Level Modelling - A Methodology for Computer System Design, IBM Res. Rep. RC 1938, Nov. 1967, IBM Res. Centre, Yorktown Heights, NY 10594. Also in Information Processing 67, Proc. IFIP Congr. 1968, Edinburgh, Aug. 1968, pp. D138 - 142

26. 8 January 2006 - 26 - © 731c[charts-mml] Laws of Software Evolution Derived from observation and metrics Provides base for empirical Theory of Software Evolution

27. 8 January 2006 - 27 - © 731c[charts-mml] Example of E-type Evolution Instability and break-up from rsn 20 triggered by excessive positive feedback preceding growth rate decline observed in other systems Underlying linear growth with superimposed ripple to rsn 19 Ripple, indication of feedback stabilisation Initial faster than linear growth First observation Size in modules over release sequence numbers (RSN)

28. 8 January 2006 - 28 - © 731c[charts-mml] Further Example Consistently similar evolution patterns observed on other systems Understanding has gone through three stages, linear, inverse square, S-curve growth patterns Similar faster than linear initial growth and sustained ripple Suggests empirical generalisations such as: Laws, Principle of Software Uncertainty, ∫- growth (S-curve) Large current telephone switch - 44M instructions 4 years ago