Most puzzles and questions asked on CodeGolf are interesting and challenging. However, usually the answers that win are created from pre-fabricated languages designed for golfing code. This seems unfair since it is easy to make ANY language you like for any situation. Languages used to answer these questions are not "Real" languages. They are made for a specific purpose to reduce code size. A symbol used in a golfing language needs to represent a built-in function. At university, we made an artificial language using Java. I believe answers in CodeGolf should be from mainstream/real languages...this would show real skill.

  • 9
    \$\begingroup\$ I'm not exactly sure what your question is, but this seems like a duplicate of either meta.codegolf.stackexchange.com/q/6983/8478 or meta.codegolf.stackexchange.com/q/286/8478. \$\endgroup\$ May 26, 2016 at 14:35
  • 6
    \$\begingroup\$ The gist is, once you disallow golfing languages (which begs the question of what exactly is a golfing language), J and APL will start winning all challenges, because they are "real languages" but still extremely terse, and then people complain about those. There is no good solution to this. Admittedly, there are some boring golfing languages, but designing a language that is short on a wide variety of tasks is an interesting meta game that is simply an integral part of this community. The best solution is for people who don't like them to ignore them. The green checkmark is only 15 rep. \$\endgroup\$ May 26, 2016 at 14:38
  • 2
    \$\begingroup\$ Arguably, it takes "real skill" to create and implement a golfing language in the first place ... if someone chooses to spend X amount of time to create a golfing language to save Y amount of time when answering challenges with that language, seems fine to me. \$\endgroup\$ May 26, 2016 at 14:59
  • 1
    \$\begingroup\$ It also takes skill to win in a golfing language - I often see what looks like a shockingly short solution later beaten by a completely different approach in the same golfing language. \$\endgroup\$ May 26, 2016 at 15:04
  • \$\begingroup\$ I don't see this as a duplicate, since it is saying golfing languages are not real languages, rather than just saying they are not fun. \$\endgroup\$ May 26, 2016 at 15:06

2 Answers 2


You're not allowed, to make a random language, to solve a specific challenge

We have things called "forbidden loopholes". One of them is, that a language can't be made to only include one specific builtin, for a specific challenge, and it must had a compiler before the challenge was posted. This makes it impossible to create artificial languages for specific challenges.

  • \$\begingroup\$ Just a nitpick, I don't think there are any compilers for any of the golfing languages (although I may be wrong). They're all interpreted. \$\endgroup\$
    – Liam
    May 30, 2016 at 6:33
  • \$\begingroup\$ @Liam I didn't felt the need to actually copy the rule from the loopholes page \$\endgroup\$
    – Bálint
    May 30, 2016 at 7:45

Your suggestion has already been implemented. Only real programming languages are permitted to enter challenges on this site. What constitutes a real programming language for this site is defined here on meta.

Note that this rules out some languages which are designed only for a small number of specific tasks. Languages which are designed to be terse still need to meet all of the requirements for being a real programming language.

Most of the golfing languages you see on this site are real programming languages, some of which are actually used outside of the golfing context (several of our users use CJam for scripts unrelated to PPCG). If you spot a language that does not meet our definition, raise it on meta so the community can verify whether or not it counts as real.


Not the answer you're looking for? Browse other questions tagged .