Path: csiph.com!x330-a1.tempe.blueboxinc.net!newsfeed.hal-mli.net!feeder3.hal-mli.net!newsfeed.hal-mli.net!feeder1.hal-mli.net!eternal-september.org!feeder.eternal-september.org!mx04.eternal-september.org!.POSTED!not-for-mail From: Gene Wirchenko Newsgroups: comp.lang.java.programmer Subject: Re: comparing two test files Date: Fri, 23 Dec 2011 11:13:44 -0800 Organization: A noiseless patient Spider Lines: 38 Message-ID: References: <5fdf65a2-052b-4c0b-b6bf-bc4dc8a7dd02@v24g2000prn.googlegroups.com> <1IGdnbzxbYSzD2nTnZ2dnUVZ_rednZ2d@earthlink.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Injection-Info: mx04.eternal-september.org; posting-host="JCn9+dUdCQH+ycRezr2Sxw"; logging-data="31069"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX188MBWF+3Z5iJh3gL/KDQcQ29fDLZvKlbg=" X-Newsreader: Forte Agent 4.2/32.1118 Cancel-Lock: sha1:8aSHLR0luUvN24qwvob7mzVmo3c= Xref: x330-a1.tempe.blueboxinc.net comp.lang.java.programmer:10978 On Fri, 23 Dec 2011 10:25:12 -0800, Patricia Shanahan wrote: >On 12/23/2011 9:18 AM, Gene Wirchenko wrote: [snip] >> How did you come up with that example? It is very good. The >> problem with cooking up examples like that is that one already knows >> what the problem is so any "detective work" is suspect. > >I took a common student problem that needs a little attention to detail, >and wrote a Java implementation as fast as I could type, making minimal >changes to fix compile-time errors. I then went straight to test and >debug, writing up the debug steps as I went along, without doing any >desk checking. > >Normally, my programming keystroke rate is a lot slower than my raw >typing rate. Suppressing self-checking and thinking about details in >favor of typing speed made a good collection of mistakes statistically >likely, but without giving me any knowledge of what the mistakes were. > >I hoped the result would be a more realistic example and debug process >than if I had consciously inserted errors in an otherwise carefully >written program, and then pretended to find them. It is more reasonable that consciously inserting errors since you actually did not design the errors. You appear to have hit some good ones, too. That you had more than one is even better, because it helps make the point that one should not stop after finding the first error. I have bookmarked the page, and I will likely point beginners to it. It really is a good treatment. Sincerely, Gene Wirchenko