About choosing the right programming language for a software startup...

About choosing the right programming language for a software startup... (vettyofficer.blogspot.com)

12 points by Pratheeswaran 15 years ago | 8 comments

badmash69 15 years ago |

OP -- you sound like an MBA : bland and completely devoid of any semblance of originality.

I think hackers should be opinionated ( right or worong doesn't matter) and have something original to say.

So let me answer the question for you.

Drum roll ....

The right programming language for a start up is Scala.

Programmers who have taught themselves Scala , with being virtually very little job requirements for Scala , have self selected themselves into an elite category.

It is statically typed.

It runs on JVM that you can instrument and monitor.

Java ecosystem is vast.

See .. having an opinion isn't that hard.

astrodust 15 years ago | |

The OP, to their credit, did manage to avoid igniting a language war. Now you're standing in the town square with a sandwich board strapped to yourself and shaking a noisy bell while crowing about Java.

Scala, awesome or not, is a pretty crazy thing to advocate. People complain constantly about the lack of developer support for something as common-place as Ruby and instead want to go with PHP or C# simply because they can't throw a rock without hitting one of those developers.

The best language for a project is the one that fits the requirements and can be supported. That short-lists things a bit.

xiongchiamiov 15 years ago | | |

Ah, but you don't need to hire 10,000 mediocre programmers; you just need 5 good ones.

Jabbles 15 years ago |

Google is doing this, Facebook is doing this and so even your startup can do this

Whilst using more than one language may be acceptable, this is a very bad justification for it.

abyssknight 15 years ago | |

I think this all comes down to using the right tool for the right job. I've worked in polyglot environments where the decision to use multiple languages and platforms was a complete disaster. There's just no need to use two tools, which perform the same tasks at the same speed, for one job. It makes hiring a nightmare, and means the team may become conversant in 10 languages, but never master any of them.

That said, sometimes (and I do mean sometimes), there is a justification and reason for this sort of thing. Facebook chat was written in Erlang for speed and scalability purposes. That makes sense. The application needed more than the standard platform (PHP) could handle.

moshezadka 15 years ago |

Maaaan.

This says absolutely nothing, except for the ever-present "just start-up, it doesn't matter." It doesn't matter only if you're a Blub programmer (see http://www.paulgraham.com/avg.html though I imagine most people have already seen it), or if you're a non-technical manager needing to convince himself that it doesn't matter.

So here is some real advice: if you are doing a typical "mostly-web-but-a-lot-of-little-elses" choose Python. Python has good frameworks like Django (that will bring up you to speed clearly for a CRUD-based web site) and Twisted (for the small "everything elses"). If you find performance bottlenecks, well, the C-API is not a "pleasure" to work with, but it's reasonable at least. It is easy to learn, so the whole "find an engineer familiar with it" is a bit of a red herring.

Sure, this advice will not fit all start-ups, but if you think it doesn't apply to yours, you better have fairly solid arguments.

(Some arguments that I think are solid: "We already know Ruby well, and the difference is small enough that this tips the scales", "we need to write in Davlik for Android". "But I know C++" isn't, for example.)

edw519 15 years ago |

Sorry to say, but this is 370 words that say nothing.

choose a programming language that best suits and can easily accomplish your product's requirements

Such as? I would love to hear OP's thoughts about the relationship between product requirements and language choice. Unfortunately, he doesn't offer any.

Frameworks are very important for startups to keep up their pace.

Why? I don't mean to be snarky, but I really want to understand what makes frameworks so important, especially if we've already selected the best language for the product. An explanation is better than a declaration.

Give more preference to the programming language which the team members are familiar with

What happens when this choice is the exact opposite of the best language for the product requirements? I imagine this happens quite often. Then what do you do?

Have an eye on the availability and cost of hiring a programmer in the language you choose.

Again, conflicting advice. What do you do when this is totally different from what the team members already know?

Do not blindly follow the trend

A startup is not about the language, it's all about the people.

don't worry too much, just startup!!!

These last 3 items sound more like fortune cookie content than blog advice.

It's hard to argue with much here because there isn't much here. If I turned this report into any competent manager or professor, it would probably come back with one word on it, "Obvious"

OP, please put some meat on these bones. You sound like you have something important to add. So add it.

jhrobert 15 years ago |

Well, if you are doing anything related to the web... chances are that you cannot escape JavaScript

This reduces the question to: JavaScript or CoffeeScript? Problem solved.

OTOH, who enjoy living in a prison?