3 ms·

I find it curious that this kind of thing still happens "oh we wrote it in the wrong language". Shouldn't every mainstream language by now be good enough to bui

I find it curious that this kind of thing still happens "oh we wrote it in the wrong language". Shouldn't every mainstream language by now be good enough to build most (if not all) applications?

xeeeeeeeeeeenu 6y ago
Most languages are good enough, but they rarely have good enough ecosystems.
- galaxyLogic 6y ago
  I think that's a great criterion to use when choosing a language
gameswithgo 6y ago
Probably yes, but OCaml isn't mainstream.
gleenn 6y ago
The author stated many reasons this just wasn't the case. Lots of missing libraries to common services or libraries that weren't mature enough like the saddeningly reported Postges driver. That seems particularly damning, if I can't reasonably access one of the most popular and general purpose DBs then how am I going to expect a litany of other software I might want to use and expect after having worked with Java or Ruby or Puthon. Personally this drives me to Clojure more, which they said they used at CircleCI and had success with given it's interoperability with the large Java ecosystem. That meshes well with my experience too.
- dwohnitmok 6y ago
  It's perhaps worth pointing out that the author of this blog post explicitly decided against using Clojure a second time for Dark after cofounding CircleCI, mainly due to a lack of sum types.
  - pbiggar 6y ago
    Right, Clojure is dynamically typed, and I'm not interested in dealing with the challenges of that again (esp nulls).
    - elwell 6y ago
      I, for one, find nil-punning useful in Clojure.
ojnabieoot 6y ago
There’s a big gap between “can theoretically build this application” and “can efficiently and reliably build and maintain this application.”
yawaramin 6y ago
Maybe that happens, but the author has stated that wasn't the case here--that OCaml was absolutely the right choice to write Dark in, given the constraints they faced at the time.
chubot 6y ago
Reminds me of Joel Spolsky's 2006 post on languages. And it even mentions OCaml !! https://www.joelonsoftware.com/2006/09/01/language-wars/ https://www.joelonsoftware.com/2006/09/01/language-wars/ FWIW I can see this igniting more arguments. And it is for sure interesting that he mentioned as Python the 0.5 that's on the edge, and Ruby on Rails that's a little further on the edge. And Ruby on Rails turned out to be phenomentally successful for many billion dollar companies since 2006. https://news.ycombinator.com/item?id=24279611 https://news.ycombinator.com/item?id=24279611 But that doesn't make Joel wrong. If you adjust 2006 to 2020, the same point can be made with different languages/platforms. I know that typically on new projects there’s a long evaluation period where you decide what technology to use, along with lots of debates that include some crazy person actually wasting quite a lot of time evaluating Squeak and Lisp and OCaml and lots of other languages which are totally, truly brilliant programming languages worthy of great praise, but just don’t have the gigantic ecosystem you need around them if you want to develop web software. These debates are enormously fun and a total and utter waste of time, because the bottom line is that there are three and a half platforms (C#, Java, PHP, and a half Python) that are all equally likely to make you successful, an infinity of platforms where you’re pretty much guaranteed to fail spectacularly when it’s too late to change anything (Lisp, ISAPI DLLs written in C, Perl), and a handful of platforms where The Jury Is Not In, So Why Take The Risk When Your Job Is On The Line? (Ruby on Rails).

I find it curious that this kind of thing still happens "oh we wrote it in the wrong language". Shouldn't every mainstream language by now be good enough to build most (if not all) applications?

xeeeeeeeeeeenu 6y ago

Most languages are good enough, but they rarely have good enough ecosystems.

galaxyLogic 6y ago
I think that's a great criterion to use when choosing a language

gameswithgo 6y ago

Probably yes, but OCaml isn't mainstream.

gleenn 6y ago

The author stated many reasons this just wasn't the case. Lots of missing libraries to common services or libraries that weren't mature enough like the saddeningly reported Postges driver. That seems particularly damning, if I can't reasonably access one of the most popular and general purpose DBs then how am I going to expect a litany of other software I might want to use and expect after having worked with Java or Ruby or Puthon. Personally this drives me to Clojure more, which they said they used at CircleCI and had success with given it's interoperability with the large Java ecosystem. That meshes well with my experience too.

dwohnitmok 6y ago
It's perhaps worth pointing out that the author of this blog post explicitly decided against using Clojure a second time for Dark after cofounding CircleCI, mainly due to a lack of sum types.
- pbiggar 6y ago
  Right, Clojure is dynamically typed, and I'm not interested in dealing with the challenges of that again (esp nulls).
  - elwell 6y ago
    I, for one, find nil-punning useful in Clojure.

ojnabieoot 6y ago

There’s a big gap between “can theoretically build this application” and “can efficiently and reliably build and maintain this application.”

yawaramin 6y ago

Maybe that happens, but the author has stated that wasn't the case here--that OCaml was absolutely the right choice to write Dark in, given the constraints they faced at the time.

chubot 6y ago

Reminds me of Joel Spolsky's 2006 post on languages. And it even mentions OCaml !! https://www.joelonsoftware.com/2006/09/01/language-wars/ https://www.joelonsoftware.com/2006/09/01/language-wars/ FWIW I can see this igniting more arguments. And it is for sure interesting that he mentioned as Python the 0.5 that's on the edge, and Ruby on Rails that's a little further on the edge. And Ruby on Rails turned out to be phenomentally successful for many billion dollar companies since 2006. https://news.ycombinator.com/item?id=24279611 https://news.ycombinator.com/item?id=24279611 But that doesn't make Joel wrong. If you adjust 2006 to 2020, the same point can be made with different languages/platforms. I know that typically on new projects there’s a long evaluation period where you decide what technology to use, along with lots of debates that include some crazy person actually wasting quite a lot of time evaluating Squeak and Lisp and OCaml and lots of other languages which are totally, truly brilliant programming languages worthy of great praise, but just don’t have the gigantic ecosystem you need around them if you want to develop web software. These debates are enormously fun and a total and utter waste of time, because the bottom line is that there are three and a half platforms (C#, Java, PHP, and a half Python) that are all equally likely to make you successful, an infinity of platforms where you’re pretty much guaranteed to fail spectacularly when it’s too late to change anything (Lisp, ISAPI DLLs written in C, Perl), and a handful of platforms where The Jury Is Not In, So Why Take The Risk When Your Job Is On The Line? (Ruby on Rails).