Re: Adds some info about pure properties to the CG. (issue 5364048)

[mike]

On Mon, 21 Nov 2011 08:02:00 +0000, address@hidden wrote:
> http://codereview.appspot.com/5364048/diff/4001/Documentation/contributor/programming-work.itexi
> File Documentation/contributor/programming-work.itexi (right):
>
>
> http://codereview.appspot.com/5364048/diff/4001/Documentation/contributor/programming-work.itexi#newcode1832
> Documentation/contributor/programming-work.itexi:1832: a @emph{prima
> face}.  If you write a pure-function, you are promising
> What does 'prima face' mean?

oops, I meant `a priori'

> http://codereview.appspot.com/5364048/diff/4001/Documentation/contributor/programming-work.itexi#newcode1858
> Documentation/contributor/programming-work.itexi:1858: @var{start} 
> and
> @var{end} is facultative, as items only occupy one
> 'facultative' is an odd word here.  I guess you mean it in the sense 
> of
> 'optional', where the user of the optional thing can turn to a
> substitute if it is not provided.
> Then just say start and end are optional parameters where the 
> function
> provides the defaults.

ok.  i'll say that the values are thrown away.

> http://codereview.appspot.com/5364048/diff/4001/Documentation/contributor/programming-work.itexi#newcode1861
> Documentation/contributor/programming-work.itexi:1861: @var{end} are
> important, as we may can get a `purer' estimation of a
> I'm getting confused.  "pure(lilypond)" generally means 'crappy' in 
> the
> sense of pure-height of a slur being the crappy estimate of height 
> one
> can get without triggering the line-breaking and note-spacing.  When 
> you
> say 'purer' above, you seem to be using the superlative of
> "pure(English)" rather than "pure(LilyPond)".

i'll change it to "better pure estimation"

> http://codereview.appspot.com/5364048/diff/4001/Documentation/contributor/programming-work.itexi#newcode1913
> Documentation/contributor/programming-work.itexi:1913: LilyPond is 
> smart
> enough to know if a series of chained functions are
> So if one function in the series has no pure equivalent, but you call
> the first function at a time when only pure functions may be called,
> what is LilyPond smart enough to do?

it gives up and returns a dummy value (something like #f or '()).  one 
(untested) way to see this is take one of the chained callbacks (like 
rest offset) and append a function:

(lambda (grob) (return 0))

LilyPond will ignore all of the previous work because it can't find a 
pure equivalent for the anonymous function.

> http://codereview.appspot.com/5364048/diff/4001/Documentation/contributor/programming-work.itexi#newcode1959
> Documentation/contributor/programming-work.itexi:1959: Pure Y values
> should be used in any functions that are called before
> should or must?

must

> http://codereview.appspot.com/5364048/diff/4001/Documentation/contributor/programming-work.itexi#newcode1996
> Documentation/contributor/programming-work.itexi:1996: overly-distort
> the Y-extent of an system, which could result in a very
> extra-spacing-height does not distort the Y-extent

true, will change.

> http://codereview.appspot.com/5364048/diff/4001/Documentation/contributor/programming-work.itexi#newcode1997
> Documentation/contributor/programming-work.itexi:1997: @q{loose} 
> looking
> score with lots of space between systems.  So, to
> Prove it.
>
>
> http://codereview.appspot.com/5364048/diff/4001/Documentation/contributor/programming-work.itexi#newcode2041
> Documentation/contributor/programming-work.itexi:2041: not be changed
> later in the compiling process due to other changes?
> This sounds like "pure(ComputerScience)" in the sense of a function 
> that
> depends only on its arguments, not on any other program state.  Does
> "pure(LilyPond)" include this concept.

no.  the only air tight definition of `pure' in lilypond is "a series 
of behaviors and coding conventions about which one can think 
spontaneously, extemporaneously and intuitively after having read 
through all pure-property-handling code for at least one year."

Cheers,
MS