When they describe performance improvements of HHVM over the existing PHP interpreter, are they comparing HHVM to PHP 5.6 or PHP 7? There were some incredible performance improvements going from 5.6 to 7, and most of the HHVM hype that I see doesn't acknowledge this.
I'm on the move do don't have any sources for you, but from what I understand php 7.x benchmarks pretty close to HHVM. I wouldn't be surprised if 7.2/3 surpasses it.
As someone who's dabbled in PHP for various ancient frameworks such as Mediawiki and Wordpress, I'm unsure why people knock on 'php' itself. The frameworks I just referenced may not be that great, and may require a lot of work to make it acceptable. But outside of that, there's a huge amount of development theory in the frameworks used to actually build your own websites, such as Symfony. I mean, take a look at the quick guide http://symfony.com/doc/current/quick_tour/the_big_picture.ht...
Or any other pieces of documentation. Its very well made, and demonstrates proper creation of websites using seperation of logics and configurations. And, you don't have to use symfony, there's tons of others that impliment the same kind of theories, which is why i think PHP has stuck around for so long and is improving like every other language. It has the ability to be formed into these seperations and logics. Much like how C# is able to give you a much better separation of data and objects than say C++. Perhaps thats something that could be done in any per-request response language like PHP, but it has withstood its test of time when properly done.
Symfony is indeed a very well crafted framework. For those who are interested to have a bit of an introduction to what modern PHP web development looks like today, it might be worth having a look through this tutorial that basically walks you through building your own web framework based on Symfony components:
The problem of PHP is it's a old & narrowed language. It been designed to run in as a Web Service dedicatedly, and of course limits it's potential.
It's just like no one had motive to make HTML turing-complete. A language dedicated as a web script is good enough as long as it output web page, right? :(
BTW:
> I claim that PHP’s simpler “think; edit; reload the page” cycle makes developers more productive.
Does this can be simulated by automatically recompile and restart the application? I don't think it can be an serious advantage here.
PHP being largely targeted st web app development is a bonus: how many times do ruby apps need to rely on a swath of gems to achieve functionality PHP has in its standard distributed extensions?
You just notice "Old"? Old usually are not cause any problem, but when you combine old and narrow, that is not a good thing.
It just like Morse Code which designed for telegram operators to send and receive message.
By that time when telegram was new, it was a good invention, everyone who wants to use telegram must know Morse Code. And because of it's designed for telegram, it come with every each bonus to make send and receive telegram faster.
However, one day, a new technology been invented, it's called Telephone. Not long after that, telegram started to fade away.
What fade away with it, is of course, the Morse Code. Because it only good at sending and receiving telegrams which not many people still do today.
Now the PHP. It has a lot of old burdens (To make it good at outputting HTTP respond) and poor design decisions. They need to find a way to organize those things and carefully redesign that language. Then, maybe it can become respectable.
It's almost the opposite of what you're saying. As PHP was designed for web development, "newer" languages often need a lot more (configuration, modules, etc.) to produce what PHP can easily do by default.
13 comments
[ 3.5 ms ] story [ 41.7 ms ] threadhttps://wiki.php.net/phpng
Or any other pieces of documentation. Its very well made, and demonstrates proper creation of websites using seperation of logics and configurations. And, you don't have to use symfony, there's tons of others that impliment the same kind of theories, which is why i think PHP has stuck around for so long and is improving like every other language. It has the ability to be formed into these seperations and logics. Much like how C# is able to give you a much better separation of data and objects than say C++. Perhaps thats something that could be done in any per-request response language like PHP, but it has withstood its test of time when properly done.
http://symfony.com/doc/current/create_framework/index.html
It's just like no one had motive to make HTML turing-complete. A language dedicated as a web script is good enough as long as it output web page, right? :(
BTW:
> I claim that PHP’s simpler “think; edit; reload the page” cycle makes developers more productive.
Does this can be simulated by automatically recompile and restart the application? I don't think it can be an serious advantage here.
PHP being largely targeted st web app development is a bonus: how many times do ruby apps need to rely on a swath of gems to achieve functionality PHP has in its standard distributed extensions?
It just like Morse Code which designed for telegram operators to send and receive message.
By that time when telegram was new, it was a good invention, everyone who wants to use telegram must know Morse Code. And because of it's designed for telegram, it come with every each bonus to make send and receive telegram faster.
However, one day, a new technology been invented, it's called Telephone. Not long after that, telegram started to fade away.
What fade away with it, is of course, the Morse Code. Because it only good at sending and receiving telegrams which not many people still do today.
Now the PHP. It has a lot of old burdens (To make it good at outputting HTTP respond) and poor design decisions. They need to find a way to organize those things and carefully redesign that language. Then, maybe it can become respectable.
We haven't moved on from http for web applications.