Can you please provide a good reason why puredata shouldn't carry the features itself of 'a professional programming language'? It seems
no :-)
i'd like to see a lot of things in pd ... strings, debugging / assertion features, a clean type system (no one-element / zero-element lists or the nested lists), an improved file format (for automatic merging using version control software) to make it a more professional programming language, kernel / gui separation, threading and thread safety to make it a professional audio / dsp software...
somewhat hypocritical to complain about non-portability of patches and then follow up with a comment like this, which would require users to have every possible 3rd party scripting language installed if they want to use strings.
imo, there are two different things
- compatibility from pd to pd
- requirement of externals (i'd guess 99% of pd users are using
externals)
what i criticized of günter's implementation, was not the implementation itself, but the way, the use of it would make patches incompatible to pd versions ... somehow like ansi c vs c99 vs gnu89 vs gnu99 vs whatever. since it's a language extension, we should either agree to include it in the new revision of the language pd (like sinf is not included in ansi c, but in c99) and make both pds (stable & dev[ei]l) compliant to the standard of the language pd or not include it at all... otherwise we risk that patches / code written with dev[ei]l / gcc won't run / compile with stable / msvc. using a scripting language wouldn't change the language pd, though ...
btw, of course, i'm critical with pd, but only because i'd like to show the weak parts of it in order to improve it ...
cheers ... tim