Am 15.04.16 um 18:41 schrieb Selva Nair: > > On Fri, Apr 15, 2016 at 8:59 AM, Arne Schwabe <[email protected] > <mailto:[email protected]>> wrote: > > Am 14.04.16 um 19:58 schrieb Jens: > > Hi, > > > > please review the following patch for a compiler warning on MacOS X. > The program logic is unchanged. > Thanks for the patch. > Program logic is unchanged and the patch is fine but reading the fixed > code is more difficult. > > Instead of error of if not A you have now need to stare bit more to get > that meaning. > > > I disagree. > > Although if ( !strcmp() ) is a common usage in C, ( !srtcmp() == 0 ) is > not, and such double negatives are best avoided. strcmp() == 0 and > strcmp() != 0 are easily understood by C-programmers. > > As for the scanf line, the original is > > if (!sscanf (dev_node, "utun%d", &utunnum)==1) > > Here sscanf can potentially return EOF or 0 for failure or 1 for > success, but the test catches only 0. EOF can result if dev_node == "" > or "tun", so probably unlikely (?) because of previous checks. In any > case. the intent appears to be to catch sscanf() != 1. And the latter is > much better and easier to read. > > If its only a question of style, the patch clearly improves the unusual > usage in the original, imo.
Fine. I am not going to nitpick here. Lets just say ACK to the original patch as it fixes the compiler warnings. Arne
