In een eerdere update over onze Dev Summit vroegen we jullie om vragen voor onze keynote-speaker Rasmus Lerdorf, de oervader van PHP. Die vragen hebben we verzameld en aan Rasmus voorgelegd. Onderstaand vind je antwoorden op de vragen van onder andere Barryke, mind123, afraca en Tarabas.
What is your opinion on HHVM and have you used it yourself?
Most popular languages have multiple implementations and variations. It took a while for PHP to get a viable alternative implementation, but I am glad it finally happened. The biggest thing it has done for the PHP project is that we had gotten a bit complacent about performance and it spurred us to focus on performance in PHP 7.0 to the point where there is no longer a performance gap between HHVM and PHP.
What about PHP do you like the least and what the best?
Probably the worst part about PHP, which is generally true for any piece of code that has been around for nearly 25 years, is that there is a lot of legacy complexity internally in the C implementation. It makes it hard for new contributors to learn and to become productive.
The best part of PHP is seeing the incredible things people have done with it and traveling around and meeting people whose lives have been impacted by it.
How deeply are you involved in the PHP community and how would you compare this community to others?
I go to a lot of community-driven conferences all over the world and am still actively contributing to various parts of the PHP project and related projects. The latest one being a static analysis tool for PHP called Phan. As for comparing the PHP community to other communities, I have no idea really. It is a bit like comparing your son to someone else's son, isn't it?
What is your opinion on the current state of PHP, compared to alternatives?
I think the project is quite healthy. PHP 7 has brought a lot of enthusiasm to the project both because of its speed and some of the more modern features that it introduced.
Who makes the final decision on what feature is implemented in a new version of PHP?
The community does. We have an RFC process - https://wiki.php.net/rfc - where people create an RFC for a new feature, we then discuss it on the internals mailing list and usually revise it and the implementation over the course of a couple of weeks and then eventually vote on it.
What are your ideas about the new JIT engine planned for future versions of PHP?
Writing a bad JIT is hard. Writing a really good JIT is nearly impossible. But work is underway and I am excited to see where we end up with it. For certain types of code with well-defined hot spots it can bring dramatic performance improvements. We already see that in artificial benchmarks. The trick is that most PHP applications aren't centered around one big hot spot, like generating a Mandelbrot fractal, for example. Instead they tend to spend their time in many parts of the code. The challenge will be to get the JIT to the point where it can help for common PHP patterns and that is definitely not an easy task.
With so many competing languages where do you see PHP standing out and what separates it from the rest?
PHP has always stood out by being Web-centric and extremely accessible. We have no plans to change that. It is, of course, a challenge to have the language be accessible to weekend-warrior beginners and at the same time run on many of the top-100 sites in the world. When the target audience is that broad, trade-offs need to be made that aren't popular with every segment of this wide audience, but I think we have done a good job making the easy things trivial and the really hard things possible with PHP.
Meer horen van Rasmus?
Wil je meer horen van Rasmus? Dat kan, tijdens de Tweakers Developers Summit zal hij tijdens zijn keynote ingaan op de technische staat waarin PHP anno 2017 verkeert en kijkt hij vooruit naar komende functies in PHP 7.1. Kaarten voor de Summit koop je via onze actiepagina.