Re: Re: [PEPr] Comment on RFC::ProtectedMembers

Alan Knowles wrote: >Arnaud Limbourg wrote: > >>>The other langauges that I know of that support private & protected vars >>>don't have any silly naming convention to indicate the priv/protected >>>status of variables. >>> >Most other languages dont have silly ($) to indicate variables, but in >some people view (like me) they make an incredible improvement in >readibility (although in reality they are there to speed up the parser). > >We already have lots of silly naming convensions WhichMayOrMayNot aid >readibility.. - but I dont think having _underscore prefixing protected >vars/methods is going to any more of a hastle than our existing rules.. IIRC, you explained in a recent message that, in DataObject, you use the '_' prefix on public variables in order to have them look different from the variables associated with table fields. In a way, you took the liberty not to follow PEAR CS in order to serve your code readability in a way you were thinking would help the end-users. In regard to what you say above, this does not look very logical. I think you should give others the same level of liberty that you took. Bertrand Mansion Mamasam