modify x1f4_jack_lx???? functions so they return non zero if not all the
    dependencies have been identified
describe the x1f4_jack_lx???? functions
function pointers:
    print a descriptive error message when failing for the pointer is null
xshell document:
    .lx
	.if files are specified for xs execution (for input and output) they
	     are closed upon xs execution - document such feature
src/aime does not wait for children to complete (say, xshell children)
    -- should it?
sshell pipes (also) the input of the run process - it does not use the input
    pipe.  dubious useful feature and one to document - or revise.
document _mode_ method interface for _trans_ allocator:
    no special semantics for mode(NULL) or mode(0) - data to be reallocated
	is assumed previously allocated and the reallocation is not to be nil
--no-dd option for at least express17 and aime
nodetype section for each library exported data type
the info for nodetype.free is not describing some argument - it is now, should
    expand into an article - update example for decq
closeexec in lxport.xshell - only after fork, in main process
study _move_list use of _mind_list
reenable configure search for libx1f4i0 once aime is removed from libx1f4i0
    -- it seems that some files looking to include libx1f4i0.h do not include
       config.h
fix the `swap' method in 1f4/xalist.c (as for e2list)
hack_expression does not detail error, shamefully
improve on copy.h so no tests are done for fixed size collections such as rates
e4 is not using c1_empty_string.  maybe it should.
it looks as if lxbail is expecting the (extended?) libaime c3? layout
    (ln.h.x1f4_leadnode_type).  as such, it requires an undocumented object
    layout.
improve the layout of structures such as a1 shuffle, c1 program
describe the new textport autolink feature (done), explain how having e4less
    created by fast_indexset is not suited with using the executive for
    multiple programs
infix+postfix binary operators (indexing operators)?
