6

Compile is troublesome. There exists many strange behaviors of Compile that can't be reasoned with any universal rules but being unclear if they're bugs, for example, this one. They're frustrating. They're not mentioned in the document. They're hard to notice by oneself. Should we start a collection for this? For example, a community wiki? Or a new tag like strange-behavior-of-compile (of course this one is not good)? Any ideas?

| |
  • I don't know if it's a bug or not, but I think they're not hard to notice, most of the time there will be a warning, and we have CompilePrint. – Silvia Apr 9 '14 at 8:50
  • @Silvia Only for experienced users, I think……OK, I admit that those who are interested in Compile are more likely to be experienced users 囧. Anyway, I think a collection will be kind of useful for late-comers. – xzczd Apr 9 '14 at 9:16
  • 2
    IMO, most important information about Compile is either undocumented, or not properly documented. So your question essentially reduces to "should we start to document Compile?". It's a nice idea, but the main problem is that undocumented behavior can change frequently and without notice, which is bound to frustrate our efforts. The fact that all this stuff is undocumented and that, as such, there are no expectations for what is proper behavior and what is a bug, probably contributes significantly to the inconsistencies and oddities that we observe. – Oleksandr R. Apr 13 '14 at 2:43
  • Has this started yet nor not? – matheorem Sep 7 '16 at 12:27
  • @matheorem No, it hasn't yet started, at least in this site... I collected some here in Chinese (partly as a draft for a post in this site in the future): tieba.baidu.com/p/2668939110 but it's still quite incomplete. – xzczd Sep 7 '16 at 12:37
  • @xzczd Haha, you know I can read Chinese :D. I just leave a comment here mathematica.stackexchange.com/a/61597/4742 about vector operation in compile, so disappointed : ( – matheorem Sep 7 '16 at 12:53

You must log in to answer this question.

Browse other questions tagged .