On Mon, Nov 2, 2015, at 01:43 AM, Björn Helgason wrote: > We have a great programming language. > It has a potential to do a lot. > There is a learning barrier. > Very many potential users give up.
I think J's biggest problem is not lack of demos but lack of full-scale apps/libs for developers to build upon. My impression is that J users are mostly using it for one-off tools in financial analysis or signal processing projects. I considered all of the APL family languages for a text mining project, and I didn't feel that the tutorial support for J was particularly deficient. Forum support is also better than for other APL languages. I eventually decided to go with K/Q because Arthur Whitney has done an incredible job of simplifying the language while keeping and even expanding the facilities for data manipulation. If J had a pre-existing library for this, I would have used it. Consider the Gensim project as an example of what I mean. It is popular but very slow, due to being written in Python. An implementation in J would be competitive with C for speed, but more concise and hence easier to maintain. https://en.wikipedia.org/wiki/Gensim -- http://www.fastmail.com - Accessible with your email software or over the web ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
