-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi Stefan
+1 for all questions. Regards Felix On 10/04/10 10:11, Stefan Seelmann wrote: > Hi dev, > > we are trying to fix remaining checkstyle errors in shared [1] and > have some questions: > > > 1. Inline Conditionals > We have 151 inline conditionals, should we get rid of them or should > we allow them? > > IMO 'simple' inline conditionals are ok: > return oid == null ? "" : oid; > > Such constructs could be simplified > return ( ( byteArray[index] == car ) ? true : false ); > to > return byteArray[index] == car; > > However nested inline conditionals are hard to read and should be avoided: > return ( val < 0 ? -1 : ( val > 0 ? 1 : val ) ); > > So what is our policy regarding inline conditionals? With checkstyle > we can't configure that simple inline conditionals are allowed but > more complex ones are not allowed. My opinion here is to find and > eliminate the complex ones and then to allow inline conditionals. > > > 2. Protected Fields > We have 135 fields with 'protected' modifier. Checkstyle complains > that instead the modifier should be private accessor methods should be > used. The rationale is to enforce encapsulation. Should we configure > checkstyle to allow protected and/or package modifiers? > > > 3. Javadoc for Private Members > Checkstyle complains about missing Javadoc of private fields. I think > we should relax that rule and don't force Javadoc for private fields > because IMO the variable name should be descriptive. Thoughts? > > > Kind Regards, > Stefan > > > [1] https://hudson.apache.org/hudson/view/Directory/job/dir-shared-metrics/ > -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.16 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAkypj20ACgkQ2lZVCB08qHFO/ACfcZmKwf0XgW+8dH0TylLuxK6K PqMAninyyT4/Blbnd7G/n22tkNu45PcI =BNWY -----END PGP SIGNATURE-----
