3 ms·
Today's wildly unpopular reality check for language booster-ism reminds you that there are next to no applications written in Haskell for a purpose that isn't p
by harry8 7y ago
Today's wildly unpopular reality check for language booster-ism reminds you that there are next to no applications written in Haskell for a purpose that isn't programming a computer. When we do this exercise we can't get to double figures and nobody had a meaningful explanation as to why that is the case for a language that has been around with claims of being useful as a general purpose programming language for more than 2 decades.
- philzook 7y agoI do not have a satisfactory answer for this. But, nevertheless I want to use it. Why can't I use it? I can't get my friends on board. Have they tried Haskell and seen that it doesn't work? No. Have I had success with it? Yes. Why is there so much distrust of Haskell exactly? "Nobody else uses it. It's weird and hard." I can extrapolate this battle to one with business people with money on the line and see it falling flat a lot. But perhaps Haskell is fundamentally ill suited to applications. I hope not. Other possibilities: A language's success may be more sociological than technical, Business factors dominate, success is totally random, who got there first, holdover opinions from eras of more constrained computational resources. More answers for the same question: https://news.ycombinator.com/item?id=11907839 https://news.ycombinator.com/item?id=11907839
- harperlee 7y agoI think it is mainly sociological. Are basque, irish or croatian less useful to convey ideas than chinese or english? Not at all. Perhaps they might even be better. It doesn’t matter. What matters is who talks the language, what is written in the language, where can you be understood in the language. In programming language terms, the important questions are: do I have a library for X? Is bug Y solved quickly? How many answers do I have in stack overflow? Do I get job interviews due to that word in linkedin? Can I pair with a random valley programmer to work a quick pilot over a weekend?
- anoncake 7y agoThis is just completely irrelevant. Quality and popularity are independent concerns.
- mcv 7y agoIndependent from each other, but both relevant to adoption. For any meaningful project, it's better to use a popular language than an obscure one, no matter how good, because it's easier to find developers for the popular language, and easier to find people to support your application in the future.
- philzook 7y agoBut to be fair, they should be positively correlated. Unless the world has beaten us down into nihilism and we think merit doesn't even tend to be rewarded.
- anoncake 7y agoI'm sure they are although for programming languages that correlation seems to be particularly weak. But judging the quality of a programming language by its popularity causes a kind of distributed circular argument: People use a programming language (=it is popular), so we assume it is good. We use it which means it becomes even more popular, convincing other people it is good etc. When making a decision, using popularity as an easy but noisy indicator for quality can still make sense. If you kept your choice secret, you wouldn't even cause any negative externalities. But here, we don't make a decision so I don't think popularity is an argument worth considering (when evaluating the merits of the language itself, popularity does affect ease of hiring of course).
- King-Aaron 7y agoAnecdotally, my father develops offshore process control interfaces for a large international oil and gas technology provider, written in Haskell. Granted, I don't know a great deal about that work directly, but I often hear about use cases for Haskell programming.