ebook img

Man+Machine Review Of Maple Crisis: 1993-2004 PDF

366 Pages·2004·15.433 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Man+Machine Review Of Maple Crisis: 1993-2004

D R A F T Pre-beta release 0.1 December 7, 2004 Maple 9.5.1> int(z-z+z^2+z^3, z=0..1); Maple 9.5.1 > limit(sin(z)^2+cos(z)^2, z=infinity); int(z-z+z^2+z^3, z=0..1); 0 .. 2 int(z-z+z^2+z^3, z=0..1); int(z-z+z^2+z^3, z=0..1); Maple 9.5.1 > int((1+z)^z, z); int(z-z+z^2+z^3, z=0..1); z*hypergeom([1, -z],[2],-z) int(z-z+z^2+z^3, z=0..1); int(z-z+z^2+z^3, z=0..1); Maple 9.5.1 > sum(1/n, n= -infinity..-2); int(z-z+z^2+z^3, z=0..1); Error, (in cot) numeric exception: division by zero int(z-z+z^2+z^3, z=0..1); int(z-z+z^2+z^3, z=0..1); Maple 9.5.1 > evalf(Int(exp(z), z= 0..707)); int(z-z+z^2+z^3, z=0..1); Float(infinity) int(z-z+z^2+z^3, z=0..1); Maple 9.5.1 > maximize(tan(z)^2), z= 0..1); 7/12 f 7/6 7/12 Maple 9.5.1 > evalf(Sum(1, n= 1..infinity)); 7/6 0. 7/12 7/6 Maple 9.5.1 > evalf(EllipticE(I,I)); 7/12 Error, (in evalf/Elliptic/E_M) invalid input: expected 7/6 evalf[]index to be of type posint but received 22+Dm 7/12 7/6 Maple 9.5.1 > evalf(MeijerG([[0], []],[[0], []],0)); 7/12 Error, (in ln) numeric exception: division by zero 7/6 Man+Machine Review Of Maple Crisis: 1993-2004. Part I Failing to deliver an environment of quality, Maplesoft has brought its customers in a sad plight. Can this be fixed? © 2004 Vladimir Bondarenko www.cybertester.com Maple 9.5.1 > with(plots): Maple 9.5.1 > plot(HermiteH(1/2, implicitplot(x^111/(x+y)^17=1 z), z=0..6); , x=-1..1, y=-1..1); Vladimir Bondarenko, Man+Machine Review Of Maple Crisis: 1993-2004. Part I Cyber Tester, LLC www.cybertester.com 2 Vladimir Bondarenko, Man+Machine Review Of Maple Crisis: 1993-2004. Part I Main Maple’s Quality Results 1. Maple 9.5 is an unstable, inconsistent, non-linear, non-uniform, randomized, self-incompatible environment where fundamental math properties (uniqueness of the answer for a good-defined problem commutativity & linearity property etc) now hold, now fail making Maple breaking down grotesquely. 2. Surprisingly, success achieved by Waterloo Maple/Maplesoft came rather by chance than by planning. 3. Up to this time, a ridiculously low fraction of Maple bugs, 2-3% at best, has been reported ever publicly. 4. Waterloo Maple/Maplesoft does not fix publicly known bugs (e.g. in residue)since 1989, for 15 years (!) 5. Systematically, to fix a bug for Maplesoft means only to return the input unevaluated, suppress a wrong answer instead of providing a correct one. This cannot be considered as a true fixing approach. 6. Officially introduced features, systematically, do NOT work as it is stated in the Help. Also, Maple is raw, it often cannot follow the given task to its logical end. 7. In the version x+1 Maplesoft invariably breaks tangibly the functionality that works well in the version x. 8. There exists something like the swing of the tides in the propagation of Maple bugs over versions. 9. Maple 9.03/9.02/9.01/9 are in fact beta release candidates sold by Maplesoft under the statement that they are and at the price of regular commercial releases. 10. There are at least thousands regression bugs in Maple. 11. There are at least hundreds recurring regression bugs in Maple. 12. Reproducible instability exists in Maple at least since 1994. 13. Since Maple 8, Maple is getting more and more instable reproducibly. 14. There is irreproducible documented/recorded instability. 15. The found solutions are often partial; that is to say, unlike its competitors, Maple fails to find all solutions. 16. Maple cannot use its own resources efficiently. 17. Maple uses obsolete, inefficient algorithms. 18. Command line and the GUI versions produce different mathematical results since at least 1997. 19. Maple bugs are ubiquitous; the customer can encounter a bug even in ‘simple’ functions: /(division), type(), define(), whattype(), print(), printf(), lprint(), version(), cpulimit(), expand(), euler(), plot(), DEplot(), subs(), residue(), pade(), densityplot(), ellipse(), inequal(), coeff(), coeffs(), convert(), changevar(), eliminate(), combinat(), charpoly(), contourplot() etc 20. There are thousands undocumented features, according to Maple experts. 21. Maple makers do not know how to or in any case do not want to use a spellchecker. 22. Summing up, since some 2002, Maple is becoming more and more user-hostile. 23. Maplesoft quality assurance group’s activity is blocked or inadequate, in any case, it is inefficient. 24. Maplesoft is increasingly loosing control over Maple development. 25. There are lasting severe decision making problems in the top Maplesoft’s management staff. 26. During the last decade, a user-centered Maple development model evolved gradually into a revenue- centered Maple development model, which represents a direct threat to the very existence of Maplesoft. 27. If immediate actions are not taken, Maplesoft is running a risk of loosing its market share increasingly. Cyber Tester, LLC www.cybertester.com 3 Vladimir Bondarenko, Man+Machine Review Of Maple Crisis: 1993-2004. Part I Cyber Tester, LLC www.cybertester.com 4 Vladimir Bondarenko, Man+Machine Review Of Maple Crisis: 1993-2004. Part I Caveat The stuff you are reading relates to the final version of this Review like Michelangelo’s study does to the David. Thus, if you look for an impressive document, please stop reading, and wait for the final version of the Review to be released in about 10 months. Cyber Tester, LLC www.cybertester.com 5 Vladimir Bondarenko, Man+Machine Review Of Maple Crisis: 1993-2004. Part I Cyber Tester, LLC www.cybertester.com 6 Vladimir Bondarenko, Man+Machine Review Of Maple Crisis: 1993-2004. Part I CONTENT Abstract Key Words Motivation Warm-up Intended Readership Cyber Tester’s Visitors Forewords (cid:127) Foreword For Impatient Readers (cid:127) Foreword For Regular Readers (cid:127) Foreword For Math Instructors (cid:127) Foreword For Maple Experts (cid:127) Foreword For Math Environment Manufacturers (cid:127) Foreword For Software Testers (cid:127) Foreword For James Cooper (cid:127) Foreword For Laurent Bernardin (cid:127) Foreword For Critics (cid:127) Foreword For Software Historians (cid:127) Foreword For CAS Architects Overall Idea Introduction Maple Successes How Exactly Maple Is Degrading, Details? Does Maplesoft Force Upgrades To Get Bugs Fixed? Review Of Maple Reviews User Reported Bugs That Were Never Corrected First Registered Case GEMM Gains Lead A Live Top Maple Expert Are Maple Experts Always Infallible? Cyber Tester, LLC www.cybertester.com 7 Vladimir Bondarenko, Man+Machine Review Of Maple Crisis: 1993-2004. Part I Testing and Assessment with MapleT.A. Myths ## 1J10 About Maplesoft Selected Most Dangerous Maple Defects (cid:127) Instability (cid:127) Regression Bugs (cid:127) Recurring Regression Bugs (cid:127) Sand-Solid Numerics (cid:127) Slow-downs & Timing (cid:127) Plotting Problems (cid:127) Incompatibility (cid:127) User’s Comfort And Maple Syntax (cid:127) Simple Integrals Maple Cannot Calculate But Competitors Can (cid:127) 10Years Bugs The Long-Livers Fixed in Maple 9.5.1 (cid:127) Raw Functionality (cid:127) GUI and Command Line Math Results Differ (cid:127) Standard Worksheets: Persist To Be Raging Headache Maple Application Limits? GEMM: Genuine Error Monitor for Maple GEMM: Maple Bug Manifestations Validation Maplesoft Claims Vs Reality 1993-2004: User-Centered Model To Revenue-Centered Model Transition Howlers: Average Fixtime Some Bug Facts (ABD) Maplesoft QA Process (Beta Testing, Night Testing, etc) Maple-Oriented Groups And Their Impact UCITA And Software Customer Bill Of Rights Maple Customers’ Opinions Nomination: 100 Most Absurd Maple Bugs Free Maple Patches Statistics & Diagrams Maple Wish List Cyber Tester, LLC www.cybertester.com 8 Vladimir Bondarenko, Man+Machine Review Of Maple Crisis: 1993-2004. Part I A Win-Win Strategy: Customer-Focused Approach To Quality Top 10 Reasons to Think of Before Investing in Maple Maplesoft’s Survival Kit Testing: Machine vs Man Efficiency Comparison Cyber Tester: A Customer’s Control Shell Over Maplesoft Conclusions Future Work Acknowledgements References Appendix 1. Test Hardware Specification Appendix 2. Maple Errors List Cyber Tester, LLC www.cybertester.com 9 Vladimir Bondarenko, Man+Machine Review Of Maple Crisis: 1993-2004. Part I Abstract Once a famous Hellenic hero did a favor to a local community via certain actions with a specific type of buildings over there. In view of the Hardware Software Conflict http://www.cybertester.com/problem.php,someone must proceed along this track, Jlet it be us who begin, at the Maple case study. We have reviewed the 22 years long attempt to produce a powerful symbolic-numeric industry-oriented environment, Maple, the lessons it teaches us as well as the status quo and the future of this effort. We present the 1st world's publicly available professional investigation of the Maple computer algebra system quality assurance issues based upon 1,500,000+ test cases and 5.5+ Gb of Maple bug-related logs, produced by our automated bug identification machines, GEMM http://www.CAS-testing.org/ and VM. Trying to predict the Maple’s future and the customer’s prospects, we consider evolution of Maple quality and Maple bugs propagation along the versions over the last decade, 1993-2004. 3types of Maple’s quality decay are exposed. New types of defects in Maple are described. Recurring regression bugs are discovered. Instability propagation along Maple versions are explored. Quality-version diagrams are given. Maple myths, reasons for their occurrence have been conjectured, and their propagations have been studied. Aset of ideas to meet the Maplesoft crisis and to narrow the credibility gap is proposed. Our main technical results to be deployed in this memoir boil down to following statements; we give the reader 1024 most frustrating Maple bug manifestations to support our conclusions. The overwhelming majority of the bug manifestations quoted here are not published elsewhere except our definitive Maple Bugs Encyclopaedia http://maple.bug-list.org. Key Words User's quality viewpoint, Maple bugs, bug taxonomy, bug history, bug signature, bug severity, bug life cycle, bug list, failure prediction, bug fixing cost, black box testing, propagation of Maple bugs, bug manifestation pattern, average fix time, automated testing system, automated AI driven testing, test oracle, behaviour simulator, automated bug sample minimization, automated bug classification, gates, neuron network, symbolic computation, validation, verification, user driven quality control, quality assurance, quality control, software development, quality assurance overheads cost reduction, developer’s version, network, cluster, build, remote automated testing, parallel automated testing, distributed automated testing, computer algebra future. Being involved in several other projects, we propose to your attention at least a small fraction of our initial analysis right away; the rest, some 80%, is coming within the next months. So, with this Review we are starting the promised Maple Bug tour de force. Maple bugs carnival is called to order. Up we go! Cyber Tester, LLC www.cybertester.com 10

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.