The Great Methodologies Debate: Part 1
Total Page:16
File Type:pdf, Size:1020Kb
ACCESS TO THE EXPERTS The Journal of Information Technology Management December 2001 Vol. 14, No. 12 The Great Methodologies Debate: Part 1 Resolved “Today, a new debate rages: agile software Traditional methodologists development versus rigorous software are a bunch of process- development.” dependent stick-in-the-muds who’d rather produce flawless Jim Highsmith, Guest Editor documentation than a working system that meets business needs. Opening Statement Jim Highsmith 2 Rebuttal Lightweight, er, “agile” Agile Can Scale: Inventing and Reinventing methodologists are a bunch of SCRUM in Five Companies glorified hackers who are going to be in for a heck of a surprise Jeff Sutherland 5 when they try to scale up their “toys” into enterprise-level software. Agile Versus Traditional: Make Love, Not War! Robert L. Glass 12 Business Intelligence Methodologies: Agile with Rigor? Larissa T. Moss 19 Agility with the RUP Philippe Kruchten 27 Extreme Requirements Engineering Larry Wagner 34 Exclusion, Assumptions, and Misinterpretation: Foes of Collaboration Lou Russell 39 Opening Statement by Jim Highsmith In the early 1980s, I participated in rigorous software development. others be able to understand the one round of methodology debate. Agile approaches (Extreme similarities and differences and be Structured analysis and design Programming, Crystal Methods, able to apply the right mix to their champions such as Tom DeMarco, Lean Development, Feature-Driven own organization. Both the SEI and Ed Yourdon, and Tim Lister were Development, Adaptive Software Rational have made wonderful on one side of the debate, while Development, SCRUM, and contributions to software develop- data-driven design aficionados like Dynamic Systems Development ment, but it is important to Ken Orr, Jean-Dominique Warnier, Methodology) populate one camp. understand the similarities and Michael Jackson (not the singer), The Software Engineering Institute, differences between them and the and myself were on the other. It with its Capability Maturity Model self-professed agile approaches. was debate, but a collegial one. (not considered a methodology Only by purposeful debate will Although we argued, we did so out per se by the SEI, but so consid- organizations be prepared to the debate rages of mutual respect and the desire to ered by many others), and Rational decide which approach, or mix of further the cause of software Software, with its Rational Unified approaches, matches their partic- development. Tom, Ed, Tim, Ken, Process (RUP), populate the other. ular culture and problem domains. and I now serve together on the But even these lines are obscure. Debates become heated at times, Cutter Business Technology In the next few months, I am but thats an important part of a Council still debating. participating on two panel good debate. If people arent debates, one ostensibly on passionate about what they There have been a variety of agile versus RUP and another believe, why bother? similar debates over the last 25 ostensibly on agile/RUP versus the years data-driven versus CMM. There are articles about light One of the great services that process-driven approaches, infor- or agile RUP. It seems that no Extreme Programming proponents mation engineering versus struc- one is willing to concede agility are performing for software devel- tured development, relational to the other side. opment is their determination to database design versus object push extreme positions for database design, object everything Proponents of XP, SCRUM, and example, no documentation, no versus non-object everything. Each other agile approaches have up-front design, limiting work of these debates was useful in argued endlessly for their values weeks to 40 hours, and communal advancing the cause of better soft- and practices often stridently, as code and really radical positions ware development. Whatever have proponents of the CMM and such as frequent testing of ones ones position, having to listen and the RUP. The point is that spirited own code. By taking these posi- respond to thoughtful discourse debate and even some in-your- tions and clearly defining the from the other side helped each face jabs are good for software extremes, XPers have launched side clarify its position. development. Healthy, collegial this healthy debate within the soft- debate helps everyone think ware development community. If Today, a new debate rages: agile through the issues more clearly. Kent Beck and others had advo- software development versus Only by defining and debating will cated moderate positions, 2 December 2001 ©2001 Cutter Information Corp. everyone could just say, Well, Make Love, Not War. Glass, who Cutter IT Journal® we really do just about the same has a long history of addressing Cutter Business Technology Council: thing. If Extreme Programming the yin and yang of issues in his Rob Austin, James Bach, Tom had been named Moderate work, analyzes each of the Agile DeMarco, Jim Highsmith, Tim Lister, Ken Orr, Dick Nolan, Ed Yourdon Programming, it would not have Manifesto values and principles. generated such debate. His analysis indicates the pros and Editorial Board: Larry L. Constantine, Bill Curtis, cons of each of them, and he then Tom DeMarco, Peter Hruschka, Jeff Sutherland brings up one of indicates which side he thinks is Tomoo Matsubara, Navyug Mohnot, these key extreme points around more correct. The final tally? Roger Pressman, Howard Rubin, Paul A. Strassmann, Rob Thomsett which debate should ensue. In his Agile: 8, Traditional: 5, Tie: 3. article, Sutherland challenges the Editor Emeritus: Ed Yourdon Larissa Moss article describes the assumption that software devel- Publisher: Karen Fine Coburn balancing act that most people and Managing Editor: Karen Pasley opment is a predictable, repeat- Production Editor: Linda Mallon their organizations go through. able process. I think this is one Client Services: Christine Doucette Moss says, The debate about rigor position that helps distinguish agile Cutter IT Journal® (ISSN 1522-7383) versus agility should not be about development from traditional is published 12 times a year the choice between them, but development. It challenges one of by Cutter Information Corp., rather about creating methods and 37 Broadway, Suite 1, Arlington, the most fundamental assump- MA 02474-5552 (Tel: +1 781 648 8700, guidelines for merging just tions upon which many project or, within North America, enough of both. Moss article is +1 800 964 5118; Fax: +1 781 management and software devel- about synthesizing rather than 648 1950 or, within North America, opment methodologies are built. +1 800 888 1816; Web site: compromising and she illus- It attacks the premise that plans, www.cutter.com/consortium/). trates this synthesis through two designs, architectures, and require- Cutter IT Journal® covers the wonderful phases the thrills of ments are predictable and can software scene, with particular chaos and the dregs of struc- emphasis on those events that will therefore be stabilized. It attacks ture. Moss reports that a dot-com impact the careers of information the premise that processes are technology professionals around group she worked with did not repeatable the very foundation the world. reject structure per se, only the upon which traditional, rigorous ©2001 by Cutter Information Corp. rigor that accompanied it and only methodologies are constructed. All rights reserved. Cutter IT Journal® when imposed by other people, is a trademark of Cutter Information It is a fertile area for debate and such as their management. She Corp. No material in this publication discussion. may be reproduced, eaten, or comments, The most valuable distributed without written permission When the arguments are all about lesson learned from my dot-com from the publisher. Unauthorized experience is that agility and rigor reproduction in any form, including mushy, moderate, middle-of-the- photocopying, faxing, and image road positions, then others have not only can complement each scanning, is against the law. other they should. little information from which to Subscription rates are US $485 a year analyze and draw their own in North America, US $585 elsewhere, conclusions. Spice and diversity In his article on the Rational payable to Cutter Information Corp. Unified Process (RUP), Philippe Reprints, bulk purchases, past issues, are vital to useful debate. But so and multiple subscription and site are thoughts from those who Kruchten discusses the idea of license rates are available on request. moderate the debate, who empha- tailoring, arguing that through size the similarities between posi- customization, the RUP can adapt tions, who say, Well, they may not to a wide range of projects. He be as far apart as we first thought. asks which is easier: Starting from In his wonderfully conciliatory a blank slate, with a few key prin- article, Bob Glass suggests that we ciples, and then building the Get the Cutter Edge free: www.cutter.com/consortium/ Vol. 14, No. 12 3 process from the bottom up? Or to discuss when he thinks the XP are much easier to talk about starting from a rich knowledge practice of user stories is appro- than do. base and choosing, shrinking, priate and when he believes that modifying, and evolving existing other techniques would be more Debate can be an energizing and recipes to fit the problem at hand? suitable. Addressing the current positive force in software develop- Kruchten describes the concepts debate, he emphasizes what ment and project management. behind the RUP and characterizes you can leverage from the This debate over agile versus it as an open framework with a controversy. rigorous or traditional method- rigorous underlying process model. ologies is healthy, particularly Some of its detractors call RUP Finally, Lou Russells article when approached from a perspec- a heavyweight process and depict reminds us of the need, whatever tive of collegiality and respect. We it as a behemoth that forces you the debate, to include and should value the ideas of those to do zillions of useless and unnat- embrace diversity whether in who advocate both ends of this ural things, he says.