Re: Lilypond structure / implicit - explicit / with statement

[David Kastrup]

Bernard <address@hidden> writes:

> On 04-04-16 18:35, David Kastrup wrote:
>> Bernard <address@hidden> writes:
>>
>>>
>>> I think I am getting quite close now. Thanks,
>> I have no idea what you think you are doing but I have severe doubts
>> that you are getting close to anything useful.  It looks to me like you
>> are just randomly putting syntactic entities into a bag and shake.
>>
>> Have you read through the Learning Manual?  It should tell you how
>> LilyPond organizes its input.
>
> Thanks David and others, but I give up.
> For your info I did read the manual, much more then I wanted to.
> My ultimate question boils down to how to assign a value to a
> property. That should not even be a question.

With \set / \override, and you actually had working code doing just
that.  For some properties, it might make sense to set them right at the
creation of a context rather than afterwards (like instrument names or
number of staff lines).  For those you use context modifications spelled
with \with which may contain assignments or \override inside.

If you just had followed the code examples given for setting properties
in the manual, all would have worked fine.

> Lilypond advertise itself being easy, and it is.
> {c'} does create output.
> Even a bit more complex examples are made.
> This because code is generated implicit. But at the same time it hides
> what is behind, creating confusing.

> You have implicit knowledge and explicit knowledge. Using implicit
> knowledge creates confusing.
> As being new to Lilypond I might see inconsistencies, which are, for
> those with many experience not noticeable any more.
> I might learn those inconsistencies but first I have to comprehend them.
>
> For me, when someone come new to work for a company, the first two
> weeks are most valuable. He can see things  nobody else can see, who
> do work their many years. Use those info or it be lost.
>
> Sorry but just assigning a value to a property, when then property is
> known and the value is known, should not be such a hassle.

It isn't.  Nobody except yourself knows why you chose _not_ to stick
with the working code for that but rather walk through arbitrary
permutations of syntactic elements.

> I thought I was close. But you are right, I am not. And I am not
> spending more time to just assign a value to a property.

I have no idea why you did not just stick with the code that worked.
Instead you invented and listed a number of variations that didn't, for
a number of reasons that are not actual inconsistencies.

The only "inconsistency" I am able to see is that context modifications
(like context definitions) omit a context name for their sets/overrides,
and that they use assignment syntax without \set unlike to how context
properties are set inside of actual music.  The first is rather obvious
(how could a context modification apply to anything but the respective
context itself?).  The second is kind of a wart due to a mixture of
historical and technical reasons.

But that's pretty much it.

-- 
David Kastrup