Comment by bcheung | Hacker News Reader

Comment by bcheung | Hacker News Reader

Parent Full thread

I worked in the industry as well, pretty much all angels (agent, photographer, webmaster, videographer, programmer). I second that programmers are given a lot of flexibility and that development tends to be much faster. There's a lot less overhead.

I don't agree with the fact that porn is a frontier of tech anymore. Just read some of the forums on GFY (NSFW). They are all using PHP still and are absolutely clueless about anything that was invented after 2004. Also, read through the API docs for some of the billers. It's amazing how bad the docs and APIs are.

And no, I'm not going to give you FTP access to my server to set up your "script" because I haven't used it in over a decade, and second, I'm not giving you access to my server period. Who the hell still uses FTP? LOL

The porn industry is like the Cuba of the tech world.

Lots of fun times. Lots of horror stories too. Very unethical stuff going on though. Funny thing is, the porn sets were probably more ethical than the programming jobs.

All kinds of shady stuff happen at porn tech companies. Making the favicon look like a lock to fool users into think it was a secure site. Creating fake sites en masse just to get credit card approval because the accounts got shut down so fast for deceptive hidden cross sales.

PHP is still the best for that type of website, I've done my fair share and that is an undeniable fact. Handles load perfectly, easy maintenance, works well with Redis/Memcached/MySQL, FPM+Nginx is solid, plenty of competent devs available and very very fast development times. What more do you want? It just work and it is not a clusterfuck of modules and libraries.

Front load everything with something like Varnish (Nginx can do some of it to a certain extent) and you'll be handling 15+ millions users a day on a skeleton server stack. Hell back in '09 Pornhub was running smooth on a similar stack with very few servers (when you consider the traffic).

If you ask me most of what was "invented" after 2004 is stuff invented by Google/Facebook who are realistically the only ones needing it, but they saw an opportunity to scoop up market share in dev so they marketed their stack as "bleeding edge". The only thing bleeding is my eyes when I see something that could be wiped up in a standard PHP/Python/Ruby stack but instead is made with so many dependencies and 3rd party library that you wonder if the dude who wrote it actually knows programming or if he just glued cool techs together because Techcrunch and HackerNews say they are cool.

But yes, the smaller players are usually using outdated stuff, then again 99% of the web is. Hence why Wordpress is still a thing.

And as a former Lead Dev of Pornhub, I can assure you that tech peeps definitely are aware of the bleeding edge of tech, just that most have a tendency to not buy the hype. My most recent experience still prove to me that 95% of the cool tech I see mentioned is pure mental masturbation, it makes life easier I keep hearing yet I've never seen it, always a mess, "Goddamnit Grunt!", "Fucking npm", 'Damn dependencies not resolving!", "npm is down", are so commonly heard nowadays I wonder if people who tout it as modern development practice have actually done any "old" school development and realised "modern development" is mostly vendor lockin and provides very little added value to a competent developer.

/rant

Yeah. It's almost laughable how little you need to process high volumes. Our stack was PHP/memcache/MySQL, a closed-source front-end cache/web server, and a custom mostly-MVCish framework.

I introduced Redis to the company after using it for a distributed system (the web scraper I built to suck down other tube site content). There was initially pushback from other devs on using redis, because it was new at the time. I ignored their concerns and used it anyway, then they started using it for future projects.

At my current company I feel that the leads are largely driven by articles and buzz they glean from Twitter. They'll make a technology choice on hearsay and then shoehorn it into a running stack. Or if they start a new project, they suck in a whole ecosystem of janky libraries and frameworks that are so abstract or unstable that they end up in a rabbit hole wasting time solving problems in their dependencies. "We can commit this to open source!!!" says the lead. No, you won't. That code is shit, your code is shit, and you're not being paid to solve other people's problems.

Development at porn companies is actually more like a sprint; development at most other companies is like old people shuffling down the nursing home corridor to get their weekly enema - you don't really want to get there, so you go slowly, only because you've got nothing else to do.

Good points. I've had similar feelings about node / react ecosystem. So much damn stuff getting in the way. Actually made the decision to move away from Node / React for my personal stuff. Looking into Haskell / Yesod for side projects. We'll see. Seems really fast for development once you get through the learning curve.

> Also, read through the API docs for some of the billers. It's amazing how bad the docs and APIs are.

I tried really hard to use Stripe to bill for an adult project I was working on (which was being built with Django so I could learn Python, which I needed to sharpen up on for my day gig), and Stripe wouldn't budge on it because it was an adult site. I was willing to hold excessive reserves, accept chargeback charges at a higher rate, whatever. Ended up having to use a well-known biller instead.

Some of the tools the adult industry ends up using is solely because there are limited options for that business avenue, and because of that, captive audience.

Yeah, have to use something like CCBill or Epoch but they charge way more because it is "high risk". Can't stand their APIs though.

This sounds like an opportunity, maybe? Stripe for Porn

It would be a welcome addition. Reputation is huge though. Customer must have confidence they are going to be able to cancel and not be charged for other things. Businesses need to know the company is going to be around and pay out in a timely and reliable manner. iBill was a company that stole a bunch of people's money a little over a decade ago and lots of people in the biz are extremely cautious because of it.

Reply on news.ycombinator.com