Hacker News Reader

Comment by berkes | Hacker News Reader

How does this compare to other comparable containers? Because that relative usage would show us if "it is thriving" is acually true.

``` for repo in node python php ruby golang; do

  echo "${repo}: $(curl "https://hub.docker.com/v2/repositories/library/$repo" -H 'Accept: application/json' -s | jq .pull_count)"

done

node: 5457781829

python: 7821948426

php: 1208701378

ruby: 1468243407

golang: 2272339163 ```

Or: it's the last in line for these alternatives.

Also, containers build on other containers. E.g. the highly popular phpmyadmin or nextcloud images build on top of php. So each "nextcloud" pull, will result in a pull for php too. Are those developers? Does that mean hundreds of millions less pulls?

View on HN

Let me share this—I've worked in a large recruitment company where we were hiring developers from all around the world, and our business model was pretty much the same as Crossover's.

PHP was always the most in-demand tech for backend development, and of course, this includes frameworks like Laravel and Symfony.

So yes, PHP is thriving.

It doesn’t mean that other languages are bad or something that’s ridiculous.

this is just to answer the "PHP is dead" folks

Every language has its place, and they all have their strengths in different contexts.

Just because there's demand doesn't mean it's thriving.

Is usage growing?

How many new projects are being started with it relatively to other languages?

I certainly don't think it's dead but it probably peaked 10-15 years ago.

having peaked 15 years ago and still being actively used is success. How many JS frameworks survive more than a few years?

I agree about JS. Although jQuery and Angular are very much alive.

PHP is a mature, established language. There is a large ecosystem of developers that actively use it, and large scale projects built on top of it. It is certainly not dead nor dying. It is just mature.

It is probably past its growth phase, nothing wrong with that. There is a very long list of languages that are arguably "better" that will never have a fraction of the success PHP enjoyed, will never see thr growth PHP saw, and will never be as widely used.

I don't even like PHP btw. But the HN crowd has some very warped view of how things actually are when it comes to programming languages and how popular/successful they are.

I'd even argue that starting a new project in php makes more sense today than it did 10-15 years ago. php now has a stable development cycle and mature frameworks and tools.

Honest question, have there been startups which chose PHP recently?

To be even more honest, I never thought that PHP was dying, but I often thought it should not be used for new projects.

Previous job was in a very PHP-oriented company and it was just a Java alternative, the only advantage being the edit and refresh page dev cycle, and being worse in every other way. Perhaps Laravel could have changed my mind, as most devs there hated Laravel because of the magic (some then proceeded to implement control flows with typed exception throwing so maybe not the greatest bunch to discuss such things)

To answer your question YES

https://wellfound.com/startups/tech/php

I believe a good developer chooses technologies based on use case and requirements, not emotions.

All technologies can coexist effectively.

Personally, I use PHP, Node.js, and other like Python for different scenarios, selecting the best fit for each project's needs.

Plenty of "new projects" are really just WordPress deployments, and companies need PHP developers to customize it to fit their business processes. I'd imagine that's the market for 90% of PHP developers being hired today.

My project uses PHP and it’s pretty successful. Technology doesn’t matter as much as executing quickly on your ideas.

My point was that "PHP is thriving" cannot be deduced from the pull stats from docker. As the TLA does.

You now add an anecdotal evidence to underpin the "PHP is thriving". Which also is not a good argument.

That is not to say "haha php ded" - I'm merely and only making the point that in order to show that PHP is doing good, you need good arguments. Not a download stat, that on closer inspection, shows it's not really thriving, nor a story from your last job.

I also believe PHP still has its place. But haven't seen any solid evidence or arguments that this is the case and will remain the case.

> So each "nextcloud" pull, will result in a pull for PHP too

I think each nextcloud build will pull the base PHP image. Ordinary pulls of nextcloud will not trigger a build but will pull the nextcloud built image. So If I built an image called "A" on top of "B" then 100 person pulled "A" then will not mean that "B" gets 100 pull. This is still one pull for "B" (assuming I did everything right from first time).

> Also, containers build on other containers. E.g. the highly popular phpmyadmin or nextcloud images build on top of php. So each "nextcloud" pull, will result in a pull for php too. Are those developers? Does that mean hundreds of millions less pulls?

When pulling images from dockerhub that are based on other images, the image is rebuilt from scratch each time? That doesn't sound right.

> the image is rebuilt from scratch each time?

That's not what parent is saying.

A Dockerfile usually start off with a base image, like:

    FROM php

This will pull the php base image, which is already built and published, from dockerhub.

I guess this makes sense if you are familiar with Docker.

Yeah but the dockerfile that builds the wordpress image or whatever is not being recreated/rebuilt with every pull of wordpress.

Yes, you seem to be grasping this.

Layers are pulled.

If foolang image has:

    layer c
    layer d
    layer e

And then fooadmin image is FROM foolang:latest (and the more stuff) and builds the layers:

    layer f
    layer g

... then a user pulling fooadmin pulls layers g,f,e,d,c,b,a (b and a being layers provided by the image that foolang was built on top of).

The question is then "does the pull of layer e in docker pull foo admin count as part of docker pull foolang?"

> The question is then "does the pull of layer e in docker pull foo admin count as part of docker pull foolang?"

It doesn't.

Reply on news.ycombinator.com