# Questions about Web Scraping [duplicate]

This has definitely been asked before, but what are the rules for web scraping in a challenge? From what I can tell, the standard rule is that your program cannot get anything from the web unless the challenge explicitly requires it. But what if I design a language where the interpreter takes the source code of the program, gets the file at pastebin.com/......, and interprets it as a BF program? Then we now have a language where every program is 8 characters long (Even shorter if we encode it in base 96 or our own custom encoding) but can accomplish any task.

What exactly are the rules for:

• Where a program is allowed to take data from the internet
• Where an implementation is allowed to take data from the internet

This is , so the shortest answer in bytes wins This is a serious question about what kinds of programs and languages are allowed on this website.

• This is somewhat covered by certain standard loopholes, which forbid obtaining the output or the source code itself from an external source. As I understand it, if a solution does download data from the internet and the challenge doesn't specify otherwise, the byte count of the data downloaded should be included in the byte count of the solution. I could be wrong on that front, though, since I can't find that documented anywhere. – Doorknob Nov 6 '16 at 1:28
• Some may argue that implementations themselves cannot fetch the data from the internet but even if you disput this, it wont end well. I tried a similar thing with OEIS and result was -5 in around 3 minutes IIRC – Downgoat Nov 6 '16 at 2:09
• In addition to other loopholes, it would also run foul of the "newer than the challenge" rule. Each new remote code would "update" the language imo, no different than adding a new operator to some other language. – Geobits Nov 6 '16 at 2:37
• This is similar to the MetaGolfScript-X loophole as well. – mbomb007 Nov 8 '16 at 17:39