At 08:26 PM 6/6/2002, brad lafountain wrote: >--- Zeev Suraski <[EMAIL PROTECTED]> wrote: > > At 07:01 PM 6/6/2002, brad lafountain wrote: > > >Please don't reply to this email saying Use Java... Because php is > different > > >than java and always will be even with these new features. > > > > Brad, I beg you, there's nothing anybody can say on this list that would > > lead this to closure. Nothing. I believe that adding the things you > > mentioned does indeed turn PHP into Java, just a messy Java, Java which is > > worse at being Java than the real one is, and torn apart when compared > > against it. > > Why do you think it would be messy.
See Kristian's letters to php-dev. I really don't want to get into it at this point, mental exhaustion :) > I do believe that making a "fork" or patches for php is a bad thing. It > would >lead into a big mess. But at the same time i believe more strongly that cs >is a >good thing, types are a good thing and stronger oo support is a good thing. To >me these are more important. I don't think I can add anything about CS that I haven't already said. Adding type hints is something that we have talked about in the past, and haven't ruled out - we need to think much more about the implications. I believe the OO level we have in ZE2 is the upper limit of what a scripting language should have. There's no doubt in my mind that going beyond that is going to complicate the language beyond what our average users want. > > There's one thing that is clear to me - there's no way to 'find a > > solution', because we don't, at all, agree about the existence of the > > problem. If you believe these features belong in PHP and that it should > > import all (or most) of Java's features, we (and many others) have a > > fundamental gap in our perception of what PHP should be, and how it can > > stay competitive. > > This is exactly true. I really feel that php/zend engine could be a alot > more >than you must think it can be. No, it's not a matter of me settling for little, and you thinking we can do much better. Not at all. While I don't think we can become a better Java than Java, this is not the reason I'm so much against going down this path. If that was it, I would have said 'let's give it a try, what's the worse that can happen?' But that's not the case. I believe that by going down that route we're going to ruin PHP where it is already established as one of the most popular web platforms out there, and that's a price I'm not willing to pay. >For the most part the only argument against these is "do we need it" and "it >will make things slower". Maybe somepeople do maybe somepeople don't but if >someone "thinks" they need it then probally others do too, how much slower >will >some of these changes make the engine? Maybe not much at all. True these kinda >things will make the languge more complicated and some people don't think that >php should get complicated but I do, I think that making these kinda changes >can only make php better. FWIW, I think that performance only plays a second role in such decisions. It can indeed rule out certain features if they really slow things down without giving a significant benefit, but generally, functionality is more important than performance (good functionality does not necessarily mean more features but a good, usable platform, even if it means less features). For me, the key questions are "Does that belong in the language?", i.e., would adding this feature to PHP make it a stronger/more popular/easier to use web platform than it already is, and "Is it worth the price (if any)?". For the CS argument, my gut feeling was 'yes' for the first question (a long time ago) but when I tried to quantify it, it occurred to me that it doesn't really make it stronger/more popular/easier, and it was a clear 'no' for the second question. In my opinion, we should ask ourselves these questions about every new language-level feature. Of course, even if we agree on the questions, it doesn't mean we'll agree about the answers - but it would at least be a good start, and a good change from "Why not add it?", which many people use today. Zeev -- PHP Development Mailing List <http://www.php.net/> To unsubscribe, visit: http://www.php.net/unsub.php
