Re: problematic commit 02fe038744e634b42f1a3377c4f0dc3d25e80344

[Valentin Villenave]

On Sun, Oct 24, 2010 at 3:43 AM, Graham Percival
<address@hidden> wrote:
> I would have liked to see this in codereview before it showed up in git 
> master.

Feel free to revert it.

> changes.tely:
> - why is \include  "still recommended"?

Because \include works with "arabic.ly".

> pitches.itely:
> - if \include is still recommended, then why is \language discussed first?

Because it's way simpler.

> - I see a @noindent, which is heavily discouraged by the doc policy.
> - I see @exmaples, which are also discouraged by the doc policy.

Have you had a look at the current doc?
http://lilypond.org/doc/v2.13/Documentation/notation/writing-pitches#note-names-in-other-languages

I see "talking through the code", which is heavily discouraged by the
doc policy and IMO much worse than having a clean @example block.

> - does \language support scoping?  If so, it would be nice to see that
> discussed.

Nope. That's why it isn't discussed.

(As Neil pointed out, scoping wouldn't be much of an interesting
feature in this case.)

> - perhaps we should not consider arabic.ly as a "language", but rather
> as an "instrument (or music) style", like bagpipes.ly or gregorian.ly.
>  If so, then this is the appropriate time to discuss it.

So I suggested. But since I wouldn't make this choice on my own, in
the meantime I updated the docs in a way that preserved the logic we
used so far: not only is arabic.ly still regarded as a language, but
the languages are still referred to with a .ly extension in the table
(which wouldn't make any sense if we emphasized and recommended the
\language command).

I believe that deprecating the "language.ly" approach deserves more
thoughts and discussions, and before we do that, this commit fills the
gap.

Cheers.