5 ms·

I'd be more convinced if the author linked a bunch of software they'd written before writing yet another haskell tutorial. 10x? We're way above 10x in the rati

by harry8 6y ago

I'd be more convinced if the author linked a bunch of software they'd written before writing yet another haskell tutorial. 10x? We're way above 10x in the ratio of haskell/functional/monad tutorials to running applications. If touting that yours is "the one" and we should skip over the tens of thousands of others out there, some supporting evidence might be useful. Applications written in haskell you can run on your computer for doing something other than programming is still a list that you can count on your fingers. Please let me be wrong about that and list them if I am. I'll be delighted and wildly enthused.

leephillips 6y ago
The one I know about is Pandoc, which I use every day.
- harry8 6y ago
  One? Ok we'll do the full list again: xmonad. Some websites usable as applications using yesod maybe. (Shellcheck - great but doesn't count, you use it for programming a computer). What else are we missing. The list is not to bash Haskell, it's is instructive to be aware of a pretty significant data point and understand /why/ edit: Please note that libraries are /not/ applications but the two are conflated here (accidentally?) obscuring the point: https://wiki.haskell.org/Applications_and_libraries https://wiki.haskell.org/Applications_and_libraries
  - burkaman 6y ago
    BazQux is unfortunately not entirely open source, but it is my favorite RSS reader, is a profitable user-facing application, and is written partly in Haskell: https://bazqux.com/faq https://bazqux.com/faq https://github.com/bazqux/bazqux-urweb/tree/master/crawler https://github.com/bazqux/bazqux-urweb/tree/master/crawler Off the top of my head I also know of Freckle (https://www.freckle.com/ https://www.freckle.com/) and Co-Star (https://www.costarastrology.com/ https://www.costarastrology.com/). More importantly, this is not a Haskell book! I'm sure the author would be happy to see these ideas applied using Javascript or Scala or Java 14 or anything else. > Haskell is considered a difficult language to learn, which is certainly true if you come from traditional procedural languages. But rest assured, this is not a Haskell book. You won’t need more than a passing understanding of the language’s syntax and a few high-level idioms. Everything you’ll need to know will be shown later in this section. The ideas present in these pages are adaptable to any piece of technology you’d like, though they might require strictly more discipline to maintain than is necessary in Haskell. - Chapter 1.3.1, page 6 I'm not going to try to list all the user-facing applications written in any functional language, but I'm sure we can agree it is a much longer list.
    - harry8 6y ago
      "they might require strictly more discipline" If it's /hard/ to use the language you have chosen for your book. Which is also the most prominent language in the space to write useful stuff then you need to address that right there on the landing page. Yeah? Is that not reasonable? The Haskell list is very short. Why? Why is this a solution to that problem? Why is it not just yet another burito-like analogy for monads? Frozen-bubble was great, slick and fun at least 15 years ago and is written in functional perl. He's clearly trying to solve the Haskell problem of "We can't write apps" here. He says so.
      - burkaman 6y ago
        I don't really know what to tell you, to me it's clear that this book is basically "lessons I've learned from Haskell that can be applied to any language". I think if it were trying to answer the questions you ask, it would say so. Do you have the same issues with SICP using Scheme? Finally, since this is Hacker News, I'll appeal to authority with Paul Graham: http://www.paulgraham.com/avg.html http://www.paulgraham.com/avg.html > Another unusual thing about this software was that it was written primarily in a programming language called Lisp. It was one of the first big end-user applications to be written in Lisp, which up till then had been used mostly in universities and research labs. At the time, Lisp was 40 years old.
        harry8 6y ago
        And they were pushing scheme as the future of everything. "We need a lambda key on the keyboard, maybe you people at HP can help us with that." Which wasn't entirely in jest. I'm pretty sure Abelson and Sussman would have had an honest, will reasoned and nuanced answer as to why lisp had not replaced fortran and whether it will or even should. Those guys didn't go for rank boosterism. Why no lisp? People didn't /own/ computers they could do what they liked with. You needed permission. Ask bill gates about that, he'll tell you... How could you hire lisp hackers? Not an issue for Paul. Done. There are tens of thousands of us who learnt Haskell (and then maybe wrote a monad tutorial). How many applications? This tells us something. What? The author has a theory. The author has written a tutorial. Show me the application!
  - wtetzner 6y ago
    > Shellcheck - great but doesn't count, you use it for programming a computer Why does that make it not count? Not every language needs to be used to write enterprise software.
    - harry8 6y ago
      For this metric removing "things used to program a computer" is useful as it cuts out the circularity of cithing the Glasgow compiler and libraries. The Learn Haskell so you can program haskell well enough to use a tool used for programming haskell... This is a useful metric for someone considering programming technology to get something done rather than to learn something. (The "learn something" case here is fine. If you're thinking of it, definitely do it. Learn FP. Learn Haskell. It's useful for your mind and fun!)
aszen 6y ago
Note that the book iss not a Haskell tutorial, it's rather a design book that can be applied to any language. The author merely choose Haskell for its expressivity. Haskell is used in a lot of places, as an example Facebook's Sigma blocks spam, phishing attacks, and malware and was migrated to Haskell.
- harry8 6y ago
  And yet he calls out haskell by name and only haskell and claims to seek to solve the problem: "Functional programming hasn't taken market share because we collectively don't yet know how to write real applications with it." Someone claiming to have individually solved that problem should probably point at some applications by way of evidence, no? But I do acknowledge the self-awareness of the author that the problem /exists/ and is need of solution. But the actual lack of applications is a problem in need of a solution for anyone claiming haskell is generally useful. Dash off a few decent ones and suddenly I am ALL EARS. Aren't you?
arianvanp 6y ago
Things that are strictly not programming related and are user-facing product software, which are primarily written in Haskell that I could come up from the top of my head: https://www.juspay.in/ https://www.juspay.in/ https://wire.com https://wire.com https://chordify.net/ https://chordify.net/ https://circuithub.com/ https://circuithub.com/ https://channable.com https://channable.com https://www.habito.com/ https://www.habito.com/ https://www.adjoint.io/ https://www.adjoint.io/ https://www.costarastrology.com/ https://www.costarastrology.com/ https://www.kittyhawk.aero/ https://www.kittyhawk.aero/ https://lumiguide.eu/ https://lumiguide.eu/ https://cardano.org https://cardano.org More than I can count on my hand :)
- harry8 6y ago
  Are they all websites? It's this the answer? Use can use Haskell if a website fits the needs? How do we know they're Haskell? Or is it a small Haskell plugin being overhyped? I WANT to see Haskell success. Real success but marketing con-job boosterism. Let's say all your examples are terrific examples and maybe they are. Let's say you've only noticed 20% of then. Still orders of magnitude less than Monday tutorials. We need to face that and understand it it progress isn't going to happen.
  - whateveracct 6y ago
    As I said in my reply, more than "websites" https://news.ycombinator.com/item?id=24347370 https://news.ycombinator.com/item?id=24347370
  - arianvanp 6y ago
    No. did you click on any of the links? Non of the things I linked are websites as the main product; they're all very much product-software. My point wasnt to point out websites that are written in haskell. but _PRODUCTS_ written in Haskell; and giving you links so you can read about them :) If you're not in the mood for clicking: Juspay: Payment processor in india Wire: End to end encrypted chat Chordify: Automatic analysis of music to generate tabs and sheet music Circuithub: Tool for part picking and placing for PCBs Channable: Datafeed management tool Habito : Mortage broker Adjoint: Treasury software Costar: Astrology app Kitthawk: Airplanes Lumiguide: Smart parking management system using Computer Vision Cardano: A blockchain > Let's say all your examples are terrific examples and maybe they are. Let's say you've only noticed 20% of then. You're moving goal-posts again. You asked for more than 10 product software solutions success stories in Haskell. I gave more than 10 and now you're telling me any number I give you I need to divide by five before it is sufficient. I'm not sure what your point is that you're trying to make here. Many product software companies _outside_ of programming language technology use Haskell successfully and I gave some examples for you to look at.
whateveracct 6y ago
Millions of people have received & participated in Starbucks Rewards offers powered by Haskell. Millions of people have bought things from Target stores whose inventories were planned by Haskell.
codygman 6y ago
> Applications written in haskell you can run on your computer for doing something other than programming is still a list that you can count on your fingers. Before you make such strong claims, you should at least try and validate they're true. Please stop spreading this lie.
- harry8 6y ago
  As i said if it's outdated I'll be DELIGHTED. Let's see the list again. It's always been very disappointing in the past but less so than pretence.
  - harry8 6y ago
    This actually is pretty ugly. Accusing someone of lying. Hang your head in shame. I'll change my mind with evidence. You have provided precisely none. Seems you'd rather abuse than show evidence? Make a list. See if it convinces you. Do it without rancour and abuse. Or just move on to a different thread. This is the bit you cut from your edit of my claim: "Please let me be wrong about that and list then if I am. I'll be delighted and wildly enthused."
    - codygman 6y ago
      > This actually is pretty ugly. What precisely is "this"? From my point of view at best it's calling out FUD and an unsupported claim sold as obviously true. It's not abuse to challenge unsupported claims and call them what they are. That sort of attitude cultivates horrible horrible things... One example being science denial. > I'll change my mind with evidence. You have provided precisely none. Your initial claim had previously no evidence. Others provided evidence and your responses have me no confidence that spending energy changing your mind was/is worth it.
      - harry8 6y ago
        I am not a corporation marketing something. I am a person. I am raising a point. There are more monad tutorials than applications by an order of magnitude. This seems interesting to me. If it doesn't meet your marketing neeeds and you need to start calling people names maybe HN isn't for you. The evidence of many and varied haskell applications is very, very, very thin. I wish it wasn't. You clearly do too. Great, we agree on that. Learn haskell it's good fun. It's interesting. Do not expect to write successful applications in haskell because the evidence points elsewhere. But sure, link your repos of actual applications. Please. Not marketing pages. Actual applications. Code. The goalposts are: Why is haskell so very much less successful in applications than any other language with as many people who have learnt it. Including me and you. Git Annexe. There's another good one. There are some. Just not many. Where are the rest? This author has stated clearly he seeks to address this. So it seems like something sensible people can discuss. This is apparently more difficult than it should be. Calling people names suggests you know there's a real problem here that you wish wasn't. I'd suggest working on that problem. For a successful technology reality must take precedence over public relations. I will change my mind with evidence. collate the list, see how compelling it is or isn't. Or just call people names so we can be clear on what isn't.

by harry8 6y ago

leephillips 6y ago

The one I know about is Pandoc, which I use every day.

harry8 6y ago
One? Ok we'll do the full list again: xmonad. Some websites usable as applications using yesod maybe. (Shellcheck - great but doesn't count, you use it for programming a computer). What else are we missing. The list is not to bash Haskell, it's is instructive to be aware of a pretty significant data point and understand /why/ edit: Please note that libraries are /not/ applications but the two are conflated here (accidentally?) obscuring the point: https://wiki.haskell.org/Applications_and_libraries https://wiki.haskell.org/Applications_and_libraries
- burkaman 6y ago
  BazQux is unfortunately not entirely open source, but it is my favorite RSS reader, is a profitable user-facing application, and is written partly in Haskell: https://bazqux.com/faq https://bazqux.com/faq https://github.com/bazqux/bazqux-urweb/tree/master/crawler https://github.com/bazqux/bazqux-urweb/tree/master/crawler Off the top of my head I also know of Freckle (https://www.freckle.com/ https://www.freckle.com/) and Co-Star (https://www.costarastrology.com/ https://www.costarastrology.com/). More importantly, this is not a Haskell book! I'm sure the author would be happy to see these ideas applied using Javascript or Scala or Java 14 or anything else. > Haskell is considered a difficult language to learn, which is certainly true if you come from traditional procedural languages. But rest assured, this is not a Haskell book. You won’t need more than a passing understanding of the language’s syntax and a few high-level idioms. Everything you’ll need to know will be shown later in this section. The ideas present in these pages are adaptable to any piece of technology you’d like, though they might require strictly more discipline to maintain than is necessary in Haskell. - Chapter 1.3.1, page 6 I'm not going to try to list all the user-facing applications written in any functional language, but I'm sure we can agree it is a much longer list.
  - harry8 6y ago
    "they might require strictly more discipline" If it's /hard/ to use the language you have chosen for your book. Which is also the most prominent language in the space to write useful stuff then you need to address that right there on the landing page. Yeah? Is that not reasonable? The Haskell list is very short. Why? Why is this a solution to that problem? Why is it not just yet another burito-like analogy for monads? Frozen-bubble was great, slick and fun at least 15 years ago and is written in functional perl. He's clearly trying to solve the Haskell problem of "We can't write apps" here. He says so.
    - burkaman 6y ago
      I don't really know what to tell you, to me it's clear that this book is basically "lessons I've learned from Haskell that can be applied to any language". I think if it were trying to answer the questions you ask, it would say so. Do you have the same issues with SICP using Scheme? Finally, since this is Hacker News, I'll appeal to authority with Paul Graham: http://www.paulgraham.com/avg.html http://www.paulgraham.com/avg.html > Another unusual thing about this software was that it was written primarily in a programming language called Lisp. It was one of the first big end-user applications to be written in Lisp, which up till then had been used mostly in universities and research labs. At the time, Lisp was 40 years old.
      - harry8 6y ago
        And they were pushing scheme as the future of everything. "We need a lambda key on the keyboard, maybe you people at HP can help us with that." Which wasn't entirely in jest. I'm pretty sure Abelson and Sussman would have had an honest, will reasoned and nuanced answer as to why lisp had not replaced fortran and whether it will or even should. Those guys didn't go for rank boosterism. Why no lisp? People didn't /own/ computers they could do what they liked with. You needed permission. Ask bill gates about that, he'll tell you... How could you hire lisp hackers? Not an issue for Paul. Done. There are tens of thousands of us who learnt Haskell (and then maybe wrote a monad tutorial). How many applications? This tells us something. What? The author has a theory. The author has written a tutorial. Show me the application!
- wtetzner 6y ago
  > Shellcheck - great but doesn't count, you use it for programming a computer Why does that make it not count? Not every language needs to be used to write enterprise software.
  - harry8 6y ago
    For this metric removing "things used to program a computer" is useful as it cuts out the circularity of cithing the Glasgow compiler and libraries. The Learn Haskell so you can program haskell well enough to use a tool used for programming haskell... This is a useful metric for someone considering programming technology to get something done rather than to learn something. (The "learn something" case here is fine. If you're thinking of it, definitely do it. Learn FP. Learn Haskell. It's useful for your mind and fun!)

aszen 6y ago

Note that the book iss not a Haskell tutorial, it's rather a design book that can be applied to any language. The author merely choose Haskell for its expressivity. Haskell is used in a lot of places, as an example Facebook's Sigma blocks spam, phishing attacks, and malware and was migrated to Haskell.

harry8 6y ago
And yet he calls out haskell by name and only haskell and claims to seek to solve the problem: "Functional programming hasn't taken market share because we collectively don't yet know how to write real applications with it." Someone claiming to have individually solved that problem should probably point at some applications by way of evidence, no? But I do acknowledge the self-awareness of the author that the problem /exists/ and is need of solution. But the actual lack of applications is a problem in need of a solution for anyone claiming haskell is generally useful. Dash off a few decent ones and suddenly I am ALL EARS. Aren't you?

arianvanp 6y ago

Things that are strictly not programming related and are user-facing product software, which are primarily written in Haskell that I could come up from the top of my head: https://www.juspay.in/ https://www.juspay.in/ https://wire.com https://wire.com https://chordify.net/ https://chordify.net/ https://circuithub.com/ https://circuithub.com/ https://channable.com https://channable.com https://www.habito.com/ https://www.habito.com/ https://www.adjoint.io/ https://www.adjoint.io/ https://www.costarastrology.com/ https://www.costarastrology.com/ https://www.kittyhawk.aero/ https://www.kittyhawk.aero/ https://lumiguide.eu/ https://lumiguide.eu/ https://cardano.org https://cardano.org More than I can count on my hand :)

harry8 6y ago
Are they all websites? It's this the answer? Use can use Haskell if a website fits the needs? How do we know they're Haskell? Or is it a small Haskell plugin being overhyped? I WANT to see Haskell success. Real success but marketing con-job boosterism. Let's say all your examples are terrific examples and maybe they are. Let's say you've only noticed 20% of then. Still orders of magnitude less than Monday tutorials. We need to face that and understand it it progress isn't going to happen.
- whateveracct 6y ago
  As I said in my reply, more than "websites" https://news.ycombinator.com/item?id=24347370 https://news.ycombinator.com/item?id=24347370
- arianvanp 6y ago
  No. did you click on any of the links? Non of the things I linked are websites as the main product; they're all very much product-software. My point wasnt to point out websites that are written in haskell. but _PRODUCTS_ written in Haskell; and giving you links so you can read about them :) If you're not in the mood for clicking: Juspay: Payment processor in india Wire: End to end encrypted chat Chordify: Automatic analysis of music to generate tabs and sheet music Circuithub: Tool for part picking and placing for PCBs Channable: Datafeed management tool Habito : Mortage broker Adjoint: Treasury software Costar: Astrology app Kitthawk: Airplanes Lumiguide: Smart parking management system using Computer Vision Cardano: A blockchain > Let's say all your examples are terrific examples and maybe they are. Let's say you've only noticed 20% of then. You're moving goal-posts again. You asked for more than 10 product software solutions success stories in Haskell. I gave more than 10 and now you're telling me any number I give you I need to divide by five before it is sufficient. I'm not sure what your point is that you're trying to make here. Many product software companies _outside_ of programming language technology use Haskell successfully and I gave some examples for you to look at.

whateveracct 6y ago

Millions of people have received & participated in Starbucks Rewards offers powered by Haskell. Millions of people have bought things from Target stores whose inventories were planned by Haskell.

codygman 6y ago

> Applications written in haskell you can run on your computer for doing something other than programming is still a list that you can count on your fingers. Before you make such strong claims, you should at least try and validate they're true. Please stop spreading this lie.

harry8 6y ago
As i said if it's outdated I'll be DELIGHTED. Let's see the list again. It's always been very disappointing in the past but less so than pretence.
- harry8 6y ago
  This actually is pretty ugly. Accusing someone of lying. Hang your head in shame. I'll change my mind with evidence. You have provided precisely none. Seems you'd rather abuse than show evidence? Make a list. See if it convinces you. Do it without rancour and abuse. Or just move on to a different thread. This is the bit you cut from your edit of my claim: "Please let me be wrong about that and list then if I am. I'll be delighted and wildly enthused."
  - codygman 6y ago
    > This actually is pretty ugly. What precisely is "this"? From my point of view at best it's calling out FUD and an unsupported claim sold as obviously true. It's not abuse to challenge unsupported claims and call them what they are. That sort of attitude cultivates horrible horrible things... One example being science denial. > I'll change my mind with evidence. You have provided precisely none. Your initial claim had previously no evidence. Others provided evidence and your responses have me no confidence that spending energy changing your mind was/is worth it.
    - harry8 6y ago
      I am not a corporation marketing something. I am a person. I am raising a point. There are more monad tutorials than applications by an order of magnitude. This seems interesting to me. If it doesn't meet your marketing neeeds and you need to start calling people names maybe HN isn't for you. The evidence of many and varied haskell applications is very, very, very thin. I wish it wasn't. You clearly do too. Great, we agree on that. Learn haskell it's good fun. It's interesting. Do not expect to write successful applications in haskell because the evidence points elsewhere. But sure, link your repos of actual applications. Please. Not marketing pages. Actual applications. Code. The goalposts are: Why is haskell so very much less successful in applications than any other language with as many people who have learnt it. Including me and you. Git Annexe. There's another good one. There are some. Just not many. Where are the rest? This author has stated clearly he seeks to address this. So it seems like something sensible people can discuss. This is apparently more difficult than it should be. Calling people names suggests you know there's a real problem here that you wish wasn't. I'd suggest working on that problem. For a successful technology reality must take precedence over public relations. I will change my mind with evidence. collate the list, see how compelling it is or isn't. Or just call people names so we can be clear on what isn't.