On 14/11/2007, Steph Fox <[EMAIL PROTECTED]> wrote: > Hi Richard, > > > It would be useful to see some stats about actual performance > > increases from using the new runtime. If it is minimal, then > > cost/benifit isn't great and we are probably going to have to > > "make-do" for a while on VC6. > > Agree. > > > But, if MSVC2005EE (Microsoft Visual C++ 2005 Express Edition - > > specifically chosen because of the "free" nature of the product - > > giving more opportunity to us unfortunates who have to normally pay > > for all of our development tools to "have a go" at building our own > > PHP binaries), offers a significant performance increase, then this is > > the marketing tool we use. > > 'marketing tool'???
"PHPx will run with these Microsoft libraries and provide a 25% on average performance increase in real world tests" sort of thing. If the new libraries REALLY are that much better we owe it to the users to give it to them! > > > Sure, we will have to explain why you need to install a runtime > > library, but this is windows. > > And these are Windows users. Ha. Yes. I see your point, but for those INSTALLING PHP, they would prefer an MSI which just "did it". So, bundle the libs and move on. > > Windows users unknowningly install the > > latest runtime all the time. Many MSI installers have them there > > simply because that is the safest way to install it and have you app > > run. > > > > Sure it makes the binary package bigger, so you have 2 packages - one > > with and one without. You make the recommendation that you have the > > "with" package if you are not sure of the difference. > > > > Even at this current time, PHP is relying on the presence of a runtime > > library. It just so happens that it is so ancient that it is not > > possible to be on windows without it. > > 'Relies on the presence of' is not the same thing as 'distributes'. The > runtime library we rely on the presence of has been shipped by MS since > Windows 98. > > > So, give 'Liz the space. Please. So those of us who will benefit from > > her expertise can do so. > > Where did anyone say she should have no space? Ok, support then. Rather than a no we are not, but an OK, yes, good idea, let's do it and see what the problems are. > > > > I really feel that the core developers who only deal with *nix really > > should allow those that deal with windows be allowed to do so. It has > > no impact on *nix development other than making code compliant across > > multiple platforms (a good thing, surely?) > > Couldn't agree more. Although actually most of the *nix guys have ended up > having to load doze at some point because otherwise Windows bugs don't get > tested/fixed. Some of us doze users submit patches which are overlooked or ignored! > > As a windows user I feel that we need to move forward and offer > > another binary which takes advantage of a more modern run time - if > > the benefit can be expressed. > > Are you assuming a Windows user is a rarity? lol No, I would guess that most people who INTERACT with PHP are windows users. Most people who CODE PHP with are probably NOT windows users. So, more *nix webservers, but more IE browsers. "Prove the benefit! Enhance the world!" (to suckingly paraphrase Heroes). > > - Steph > > Richard. -- ----- Richard Quadling Zend Certified Engineer : http://zend.com/zce.php?c=ZEND002498&r=213474731 "Standing on the shoulders of some very clever giants!" -- PHP Internals - PHP Runtime Development Mailing List To unsubscribe, visit: http://www.php.net/unsub.php
