Re: Foundation classes status, PHPUnit and pear script

From: Date: Thu, 13 Mar 2003 22:49:51 +0000
Subject: Re: Foundation classes status, PHPUnit and pear script
References: 1 2  Groups: php.pear.dev 
Request: Send a blank email to pear-dev+get-14278@lists.php.net to get a copy of this message
Le Thu, 13 Mar 2003 21:38:03 +0100, Martin Jansen a écrit : > On Thu Mar 13, 2003 at 08:1247PM +0100, laurent bedubourg wrote: >> I've read some messages in the mailing list archive about Foundation >> classes, i've been afraid when i read that these classes should be php 4.2 >> compliant > > What do you mean with "php 4.2 compliant"? Where was that stated? > > Currently there is only Christian's RFC: > > http://news.php.net/article.php?group=php.pear.dev&article=13847. > There > will be a face-to-face/online meeting of the core PEAR developers next > month, after which there will hopefully be more details available. In the link you provided : "- PFCs have to run on each released PHP-version since 4.2. If PHP 5 has matured, they have to run on PHP 5 as well" I'll just put my humble opinion (no need to flame me, it's just my last thinking before going to bed) lolo quote When ZendEngine2 will be out of the box, PEAR class and its error handling system will have to disapear slowly (the time for transition to library and php upgrade). There's no point in preventing PEAR developpers from using ZendEngine2 improvements, it would just be a global project slow down. I'm sure there will be new usage for a PEAR class, but it will be different, storing const, getting PEAR information, packages import, etc... You don't design the same way a library that can use exceptions, namespaces, destructors, setters, getters, consts, ... -- the scenario -- Continuing emulating error handling, destructors, extending ever and ever the PEAR class, making undigest namespaces with 128 characters whereas the language have everything featured in, will just make PFCs un-used by most php OO coders. I bet new libraries will emerge with their own FCs and they'll win the tos because new php comers will look for traditional OO programming (private, namespace/package, const, exceptions), and old php coders will want some fresh air as they will be aware of language improvements, they'll never accept to code in the 'Old and "slow" PEAR system that try to be compliant with php 0.1 version', (yes, they will say that !:) As well, they'll never accept to use PFCs if new alternative classes exists and they'll look for new classes to replace each PFC one by one. So we'll have less and less libraries turnover and PEAR will slowly die... or in the best case PEAR will loose bits of its public image and the PFC project will ends with a new PEAR Foundation Classes version 2 which will allow try/catch, destructors, namespaces, etc... but too late as other libraries will be used in stead of PEAR an users won't change those (why replacing a library X with quite the same library X that cames from 'old fashion thinking guys' ?) -- end of scenario -- I think our work as PEAR supporters (for php5 at least) will be to ensure web server administrators take the time to upgrade their php installation. We'll have to diffuse the information that php5 is THE version to upgrade to, explaining why. ("Hey sys admin, look what i can do with php 5, i want it now !") Meanwhile we'll be able to make improved version of our libraries with latest features (a special note on the last version that was supported on php 4.X (may be a gold pear package:)) and the PEAR project will continue to grow. For me PFCs can only exists with ZendEngine2 unless you want to loose time coding things other people will reimplement in no time. I hope that doesn't hurt anybody, i write that for the health of PEAR. lolo quote Best regards, laurent -- Laurent Bédubourg laurent.bedubourg@free.fr PHPTAL status 0.4.0 - beta - testers welcome PHPTAL home http://laurent.bedubourg.free.fr/

« previous php.pear.dev (#14278) next »