Tracy, I believe that there are a small, well-defined set of concepts that must be presented to the novice reader before tackling even the basic NuVoc pages. Concepts like "J primitives consist of either one or two ASCI symbols", "primitives can have just right, or right and left arguments", "execution is from right to left", and "there is no precedence for primitives" are some of the most basic concepts that a newbie needs to understand, before jumping into the NuVoc pages.
Those concepts, along with some basic primitive usage such as array creation (i. and $), assignment (=. =:), and some pre-built array definitions will be the content of the "Read This First" page I want to define. This start-up page will not go into detail on defining any primitives. It will tell just enough to get the newbie going. With this head start, the newbie should be able to jump right into the NuVoc pages. I don't want to say that we shouldn't repeat some of the start-up page info on the NuVoc pages. Redundancy is a good thing, if judiciously placed. It can significanly improve the learning curve. Veteran APL and J users tend to forget just how revolutionary these concepts are. Experts in traditional scalar language may have a particularly hard time learning some of these concepts. In some ways, true novices can absorb these concepts more easily than the programming expert. In any case, It is these concepts that need to be presented in the "Read This First" page.of the NuVoc. That "Read This First" page is what I will try to define. Skip Cave . . . .Tracy Harms wrote: > Hi, Skip. > > It is not possible to do what you've described, but I trust those who come > to J to have much more modest expectations of the material. > > When in the early phase of encountering J, a reader will be working both on > the most basic aspects of the language and on advanced aspects that fit > their particular interests. They may be naturally drawn to look at C. or p.. > or t. because of the sorts of problems they are interested in solving, yet > be learning things like "parameters don't need to be put in parentheses." > > The thing worth aiming at, in my mind, is making it easier for novices to > structure their exploration. By this I mean that we help them learn the > relationship among the major concepts that apply in J. Hyperlinking provides > a nice tool for this, but the hard work is in extracting the concepts, > communicating them well, and organizing the component explanations well. > > My main point here, though, is just to say that anybody who learns J will do > so by returning again and again to particular examples and explanations. > Assuming this, rather than assuming that the reader needs to be able to walk > away with full understanding from a single reading, may help in deciding > what should be written and how it is best organized. > > -- > T > > On Wed, Feb 24, 2010 at 8:48 AM, Skip Cave <[email protected]> wrote: > > >> ... >> >> That is the primary difficulty with a random-access reference/tutorial. >> How can you make every NuVoc page provide enough information so that it >> can be the first page that a novice sees, and still provide all the >> information they need to understand the specific symbol that they are >> looking up? ... >> >> > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
