-- D. J. <[email protected]> wrote (on Friday, 13 August 2010, 11:11 PM +0800): > I only for myself: > I was the first one replying your email of "General inclinations > regarding prefixing non-public members?" with "No underscore". > But I did not participate your poll for I really don't understand your > intention of creating a poll after you had that email and so many > people already responded.
Because it's one thing to ask contributors, and another entirely to poll the general community. While this decision would primarily affect contributors, it also affects anyone who extends ZF classes, as well as those who adopt the ZF coding standards. As such, it's important to get a broader viewpoint. If you had feedback on the mailing list, I encourage you to also vote in the poll. > On Fri, Aug 13, 2010 at 8:45 PM, Matthew Weier O'Phinney > <[email protected]> wrote: > > > > -- Саша Стаменковић <[email protected]> wrote > > (on Friday, 13 August 2010, 10:14 AM +0200): > > > Can you tell us current score? :) > > > > There are currently 381 responses: > > > > * 57% vote "Yes" (to remove the underscores) > > * 38% vote "No" (to retain underscores) > > * 4% vote "No opinion" > > > > What has been interesting is that the percentages have remained > > consistent from the outset -- I expected more deviation. What is also > > interesting is that there is no real clear majority. Typically, I like > > to see a 2/3 vote to feel comfortable that the change is widely > > accepted, but that is not the case at this time. > > > > > > > On Thu, Aug 12, 2010 at 6:05 PM, Matthew Weier O'Phinney > > > <[email protected]> > > > wrote: > > > > > > ZF Coding Standards are based on PEAR's CS. That standard was > > > developed > > > first by Horde, then expanded by PEAR, during the PHP 4 lifecycle. > > > PHP 4 > > > had no concept of visibility in its object model; to provide some > > > pseudo-visibility, PEAR CS mandated that members considered non-public > > > should be prefixed with an underscore. > > > > > > With the advent of PHP 5, PHP's object model received visibility > > > operators in the form of private, protected, and public. Applying PEAR > > > CS to PHP 5 code meant that if you marked a member as private or > > > protected, you would also prefix with the underscore. Many have felt > > > this is redundant, and also that it makes refactoring more difficult > > > (changes in visibility often mean renaming the members). Proponents of > > > the standard, however, argue that the leading underscore leads to > > > easier > > > maintenance of the code -- you know immediately what the visibility of > > > the member you're dealing with is just by looking at it. > > > > > > PEAR2 has decided to eschew the underscore prefix: > > > > > > http://wiki.pear.php.net/index.php/MeetingMinutes20080824# > > > Underscore_prefix_on_private_.28protected.3F.29 > > > > > > Basically, this rule is no longer required (as it was in PEAR1), > > > though > > > developers may choose to use them. > > > > > > What is YOUR opinion? Should the underscore be dropped in ZF2? > > > Please vote! > > > > > > http://is.gd/eeA6f > > > > > > Please do _not_ reply to this thread -- the arguments for and against > > > are well known at this time -- we're simply trying to decide on > > > whether > > > or not to amend the coding standards for ZF2. > > > > > > Thanks! > > > > > > -- > > > Matthew Weier O'Phinney > > > Project Lead | [email protected] > > > Zend Framework | http://framework.zend.com/ > > > PGP key: http://framework.zend.com/zf-matthew-pgp-key.asc > > > > > > > > > > -- > > Matthew Weier O'Phinney > > Project Lead | [email protected] > > Zend Framework | http://framework.zend.com/ > > PGP key: http://framework.zend.com/zf-matthew-pgp-key.asc > > > > -- > Dev Lead for Xoops Engine > Internet Application R&D @PerfectWorld > -- Matthew Weier O'Phinney Project Lead | [email protected] Zend Framework | http://framework.zend.com/ PGP key: http://framework.zend.com/zf-matthew-pgp-key.asc
