Re: Re: one abstraction layer (my proposal for)

Hello, Paul Cooper wrote: > > Anyway, things like this are part of the reason why I proposed wrapping > > PEAR-DB around Metabase API. It would certainly be a solution that would > > be feasable and reliable in much less time than your proposal takes to > > accomplish. More on this below. Now for other comments... > > > > Manuel, this PR problem will exist whether or not PEAR includes Metabase > or not. As I understand it Metabase currently is a fully functional > database independent API + much more which no other language has and has > been around for x years (where x > 2 right?). Even if Metabase were > magically ported to C tomorrow and the native APIs done away with so > that the only way to access a DB was through Metabase, this nonsense > would still be spouted - and is easily refuted. There is bad PR like > this about linux, apache, mysql, postgres, etc, etc yet AFIK no of these > projects based a technical decision on PR (leave that to the proprietary > software vendors). You are not yet getting the point. The main goal is not to port Metabase to C but rather provide an officially sanctioned database abstraction package that provide portability to database applications. The only officially sanctioned database abstraction layer is PEAR-DB, but it does not provide database portability now nor anytime soon. Wrapping PEAR-DB around PEAR-DB would provide such portability. Of course that is not all, it needs to clearly advertised wherever you can that PHP would then have a portable database abstraction layer that is distributed with PHP which for the users would still be PEAR-DB. The problem that I am telling that should be attacked is not the lies sales people that push languages of products that compete with PHP. They use half-truths to make their points and drag customers that PHP looses because customers have no way to verify if sales people are telling them complete truths. For instance, in that same site they claim that PHP is slow and was made to be slow by Zend people because they would have a market to sell their accellerators software. Part of this is true. Zend people sell accelarator software, but that does not make PHP a slow solution. The problem is that less informed customers, which is the majority of people that does not know PHP but could be using it with advantage, do not know that for instance PHP optimizer is free and even PHP 4 is reasonably fast in comparision with other languages. Likewise, it half true that PHP does not provide an abstraction layer for developing portable applications. The full truth is that PEAR-DB which is the only officially sanctioned database abstraction layer that is distributed with PHP does not provide database application portability, but there are other free database abstraction layers, like Metabase that provide database portability. My proposal goes in the direction of fixing this mis-information so that less informed customers do not fall for that apparent weakness of PHP of lacking of a portable database abstraction layer. And the goal is not do it now, not in more two years if you do not agree with the merit of what I proposed. Regards, Manuel Lemos