5 ms·

Great initiative. I often wonder why Racket isn't used more in the industry, especially when you see much more niche languages being used... One note about you

Great initiative. I often wonder why Racket isn't used more in the industry, especially when you see much more niche languages being used... One note about your site: You are subscribing people to a mailing list without sending a confirmation email?

ctas 8y ago
Thank you. I like the idea that Racket tackles each domain specific problem with its own language. I'd also be happy to see it used more often in the industry, especially since there's a growing ecosystem of libraries, tutorials, and other resources that ease the start into Racket. Regarding the subscription: I'm using Mailchimp. A confirmation mail should've been sent out, if you subscribed. I will look into it. Thanks for pointing that out.
hamandcheese 8y ago
I think by definition Racket not being used in industry makes it more niche than those "much more niche languages being used"
chongli 8y ago
Racket is too powerful to be used by big corporations. The last thing a company wants is to have their rockstar programmer leave behind a terse, arcane, impenetrable walnut of an application. Languages like Java make it much easier to treat your programmers as interchangeable parts.
- squirrelicus 8y ago
  This is true issue irrespective of the company's bigness.
- WalterGR 8y ago
  Racket is too powerful to be used by big corporations. The last thing a company wants... Ah yes, The Slashdot Cynicism. Guaranteed to get a +5 Insightful. What is your experience with Racket? Or failing that, Scheme? Or, failing that, Lisps in general?
  - kamaal 8y ago
    Not exactly Racket. But with Perl. And the same argument. Early 2010's, I remember my company had a fairly fast, stable and small(LOC wise) data feed processing application in Perl. It just worked, barely any bugs, maintained by 2 - 3 programmers. Time came around, and your Director level people realized the teams were too small, programmers were having too much fun and more importantly it made programmers felt valuable and hard to be treated as kleenex. They decided to give some Architect Astronaut love to the application. Then came in Java programmers. Code base size exploded by 500x. Design pattern floods followed. DI to manage complexity which shouldn't have even existed. Hundreds of classes to do things that could be done in hundreds of lines in Perl. Managers routinely boasted how they managed to grow their teams, about their culture, and how big and great their teams were. They could never make it work as well as it did in Perl. I too moved on. But eventually I saw they just shut down the product. No manager likes to see their work less valuable or better than that of a programmer. Languages like Perl, Scala, Clojure, Racket, Common Lisp are never going to see decent adoption in the industry.
  - chongli 8y ago
    Used it in first year into to functional programming class. We didn't even get to see half the features of full Racket because the class used the student versions. Plenty of my classmates who had been programming in Java for years were getting overwhelmed by it. I know from what I've read outside of that class that Racket is powerful enough to implement any other programming language as a library. A single programmer leveraging all these features can write an application that uses bespoke domain specific languages for every component. This is wonderful for a solo project but good luck passing that on to the new recruit. P.S. guilty as charged. I am an old Slashdot reader.
- fulafel 8y ago
  There is life and interesting work in the industry outside the bigcorp java trenches, don't lose hope :)
  - chongli 8y ago
    I'm actually having a lot of fun at my job because I'm not actually a developer, I'm in change management. A lot of our processes are old, clunky, and involve a lot of manual steps. I've been writing Perl scripts to automate this stuff and my boss is loving it.
    - quickthrower2 8y ago
      You are an "X who programs". Nice position to be in. Well done!
truncate 8y ago
Just to add, there is also some kind of bias against Lispy languages because it looks different. People I meet who know about Racket (or other lisp dialects) just hate the "parens" or the prefix notation. Personally, I felt the same until I wrote reasonable amount of code, and that's why I think its a bias that one can get over with eventually.
oxryly1 8y ago
Racket has a few issues that make it hard to put into a production environment with a team of devs. * Poor editor support. Apart from emacs and vi, many widely used editors support syntax coloring, but few do any proper indentation, and none (AFAIK) will do anything like intellisense. Only DrRacket really goes the distance on all those points and it hasn't kept up with the likes of Visual Studio Code or Sublime etc. You may say, just use Emacs! But if another dev doesn't use it they'll just ask for the language plugin for their current editor and then be mystified and frustrated at why they can't be productive. * Poor performance. Out-of-the-box new Racket code will be slow. You can "precompile" it using Racket's built-in tools, which is great. But the process is complex and error-prone, and deploying your "precompiled" Racket code across a team in heterogeneous environment is not straightforward. My teammates are regularly stumped by how to keep our code running quickly. Racket also lacks effective profiling tools to address this. That said, as Bjarne Strousup opined: "There are languages people complain about, and there are languages no one uses." I have used Racket in production environment for many years.

ctas 8y ago

Thank you. I like the idea that Racket tackles each domain specific problem with its own language. I'd also be happy to see it used more often in the industry, especially since there's a growing ecosystem of libraries, tutorials, and other resources that ease the start into Racket. Regarding the subscription: I'm using Mailchimp. A confirmation mail should've been sent out, if you subscribed. I will look into it. Thanks for pointing that out.

hamandcheese 8y ago

I think by definition Racket not being used in industry makes it more niche than those "much more niche languages being used"

chongli 8y ago

Racket is too powerful to be used by big corporations. The last thing a company wants is to have their rockstar programmer leave behind a terse, arcane, impenetrable walnut of an application. Languages like Java make it much easier to treat your programmers as interchangeable parts.

squirrelicus 8y ago
This is true issue irrespective of the company's bigness.
WalterGR 8y ago
Racket is too powerful to be used by big corporations. The last thing a company wants... Ah yes, The Slashdot Cynicism. Guaranteed to get a +5 Insightful. What is your experience with Racket? Or failing that, Scheme? Or, failing that, Lisps in general?
- kamaal 8y ago
  Not exactly Racket. But with Perl. And the same argument. Early 2010's, I remember my company had a fairly fast, stable and small(LOC wise) data feed processing application in Perl. It just worked, barely any bugs, maintained by 2 - 3 programmers. Time came around, and your Director level people realized the teams were too small, programmers were having too much fun and more importantly it made programmers felt valuable and hard to be treated as kleenex. They decided to give some Architect Astronaut love to the application. Then came in Java programmers. Code base size exploded by 500x. Design pattern floods followed. DI to manage complexity which shouldn't have even existed. Hundreds of classes to do things that could be done in hundreds of lines in Perl. Managers routinely boasted how they managed to grow their teams, about their culture, and how big and great their teams were. They could never make it work as well as it did in Perl. I too moved on. But eventually I saw they just shut down the product. No manager likes to see their work less valuable or better than that of a programmer. Languages like Perl, Scala, Clojure, Racket, Common Lisp are never going to see decent adoption in the industry.
- chongli 8y ago
  Used it in first year into to functional programming class. We didn't even get to see half the features of full Racket because the class used the student versions. Plenty of my classmates who had been programming in Java for years were getting overwhelmed by it. I know from what I've read outside of that class that Racket is powerful enough to implement any other programming language as a library. A single programmer leveraging all these features can write an application that uses bespoke domain specific languages for every component. This is wonderful for a solo project but good luck passing that on to the new recruit. P.S. guilty as charged. I am an old Slashdot reader.
fulafel 8y ago
There is life and interesting work in the industry outside the bigcorp java trenches, don't lose hope :)
- chongli 8y ago
  I'm actually having a lot of fun at my job because I'm not actually a developer, I'm in change management. A lot of our processes are old, clunky, and involve a lot of manual steps. I've been writing Perl scripts to automate this stuff and my boss is loving it.
  - quickthrower2 8y ago
    You are an "X who programs". Nice position to be in. Well done!

truncate 8y ago

Just to add, there is also some kind of bias against Lispy languages because it looks different. People I meet who know about Racket (or other lisp dialects) just hate the "parens" or the prefix notation. Personally, I felt the same until I wrote reasonable amount of code, and that's why I think its a bias that one can get over with eventually.

oxryly1 8y ago

Racket has a few issues that make it hard to put into a production environment with a team of devs. * Poor editor support. Apart from emacs and vi, many widely used editors support syntax coloring, but few do any proper indentation, and none (AFAIK) will do anything like intellisense. Only DrRacket really goes the distance on all those points and it hasn't kept up with the likes of Visual Studio Code or Sublime etc. You may say, just use Emacs! But if another dev doesn't use it they'll just ask for the language plugin for their current editor and then be mystified and frustrated at why they can't be productive. * Poor performance. Out-of-the-box new Racket code will be slow. You can "precompile" it using Racket's built-in tools, which is great. But the process is complex and error-prone, and deploying your "precompiled" Racket code across a team in heterogeneous environment is not straightforward. My teammates are regularly stumped by how to keep our code running quickly. Racket also lacks effective profiling tools to address this. That said, as Bjarne Strousup opined: "There are languages people complain about, and there are languages no one uses." I have used Racket in production environment for many years.