Why bet365 (with 22m customers) use Elixir and Erlang

[AstonJ]

###Go, Erlang, Elixir; what’s in a programming language anyway?

bet365 has more than 22m customers, making it the world’s largest online gambling company, with reports suggesting that up to three-quarters of its £1.5bn revenues come from international markets. Extreme scalability, concurrency, and reliability is a must to keep bet365’s engines turning. bet365 lives by the mantra the right tool for the job, and runs a mix of Erlang, Elixir, and Go in production.

When you pit Go’s requests per seconds against the likes of Ruby on Rails or Django, Go returns some impressive benchmarks performing three times better. Go can scale to hundreds of thousands with relative ease, in much the same way that Erlang and Elixir scale to millions.

Full story: https://sbcnews.co.uk/features/2017/09/26/erlang-elixir-programming-language

via @rvirding

[brightball]

I really wish they’d post more numbers.

[melmoth]

Definitely…

It would be really great to have some real facts, tests and numbers…

[jmilet]

As much as I like Erlang and Elixir my experience regarding Go’s performance is just the opposite.

I totally agree that it would be nice to have facts to support this statement.

[sztosz]

As much I’d like to use GO it was great disappointment for me. Not the performance wise but rather semanticaly; it’s too damn imperative

[AstonJ]

Not sure if it’s related…

[peerreynders]

FYI:

A clip showing off the responsive bet365 front end. It’s part of a slide show used in

Ben Frain is the author of Enduring CSS (eCSS).

[mnussbaumer]

Once he moves the css and js into a single file he’ll have vue single file components

[peerreynders]

Evaluate approaches against your own requirements

Combining JS and HTML is one thing - keeping the authoring CSS separate probably allows greater freedom when it comes to tooling. And depending on the circumstances you may need two separate CSS files per component - structure and skin (e.g. if the site/application likes to adjust the look with seasons relevant to the business).

[mnussbaumer]

I was just pointing out that they have the same concerns and seemed to have reached similar conclusions on how to solve that - I was not saying that it would be the right choice to use another 3rd party system with a bunch of dependencies instead of using plain css/js/html

[AstonJ]

Why does it say:

The company even went as far as to acquire Erlang in 2016 and contributing the code to the open source community on GitHub, to encourage more enterprise developers to use it.

They haven’t acquired Erlang as far as I know

[peerreynders]

The comparison is written in terms of Go vs C/.NET. The interesting thing is that it isn’t actually stated that they dumped the BEAM (silently implied maybe).

And to be realistic they’ve probably gone mainstream by standardizing on a container management system that lets them shift workloads to external commodity servers. So fault tolerance and reliability have to be approached completely differently.

The question is whether there is a niche for the BEAM in the cloud space where container management systems are the norm. From $erverless to Elixir makes me hopeful but it is a fair question whether the same effect could have been just as easily accomplished with Go.

I still think the BEAM will be relevant in the resource constrained IoT space.

[mikemccall]

That blog posts sounds like they were paid to write it. It provides zero value and sounds like a paper I would write just to get a passing grade in school. To be clear I don’t have any hate towards go. I really just don’t understand blog posts like that. They provide nothing of value. It’s actually pretty frustrating to read because I enjoy when large companies write about how they solve big problems. I hate reading fluff where there genreral theme is “Hey look at us we are using {insert language}. We cool” You could pretty much find and replace “golang” in this post with any other technology (serverless, elixir, node, etc) and the article would read the same.

To their credit they have some other articles that are pretty good.

[OvermindDL1]

That blog posts sounds like they were paid to write it. It provides zero value and sounds like a paper I would write just to get a passing grade in school. To be clear I don’t have any hate towards go. I really just don’t understand blog posts like that. They provide nothing of value. It’s actually pretty frustrating to read because I enjoy when large companies write about how they solve big problems. I hate reading fluff where there genreral theme is “Hey look at us we are using {insert language}. We cool” You could pretty much find and replace “golang” in this post with any other technology (serverless, elixir, node, etc) and the article would read the same.

Wow, you weren’t kidding, that article is the worst one of theirs I’ve read yet, they are usually pretty decent, what happened… o.O

[joaoevangelista]

From bet365 blog about the adoption of go:

Resilience is where the volume really increases. You’ve got to be prepared for the unknown of the live environment. The ‘What If’ scenarios.

Really Go? What ifs?
I think that is just horribly written blog posts that states the same reasons everyone says Go is good (because Google uses it) and you combine a CTO that seems to love hype trains and catch this one to replace .NET applications (maybe developed during the .NET hype?)
The Erlang part seems to still be there otherwise they would not have taken a step as maintainers of Riak

[dwahyudi]

Pretty much what they do.
Gambling.

[rey]

I know this was over a year ago, but thank you SO much for posting the eCSS video from Ben Frain. I’ve been doing CSS for “years” and I have some huge projects (personal and business) that this will most likely affect in a huge way, as I planned on doing them mostly with Tailwind / OOCSS.