Re: MacOS X development and support

[David Kastrup]

James <address@hidden> writes:

> Hello,
>
> On 5 March 2012 07:45, Phil Holmes <address@hidden> wrote:
>> ----- Original Message ----- From: "Graham Percival"
>> <address@hidden>
>> To: "Colin Hall" <address@hidden>
>>
>>>> Let's drop Lilypad. It's getting in the way of regular releases.
>>>
>>>
>>> Disagree.
>>>
>>
>> Disagree.  I'm with Colin.  I've already said that I can't see the point of
>> it on Windows: it's a cut-down version of notepad.  If it's the same on Mac,
>> I don't see the value.
>
> \PointAndClickOn
>
> That was a great feature for a 'mac user' like me. While I can run CLI
> in terminal.app, getting another editor to work with pointAndClick was
> fruitless and frustrating (at least for 2.12 and 2.13.x when I had my
> mac).
>
> So I'm with Graham on this one.

As far as I understand, in spite of the initial activities of core
LilyPond developers, LilyPad is basically an external application that
we just wrap and don't actively codevelop (meaning that its overlap with
actual LilyPond development nowadays is rather minimal).

The principal goal is setting up the user with an environment where he
can start working right away.

We could pick a different target for that purpose.  Personally, I want
to get to a state where Emacs will make a compellingly useful part of
the music creation toolchain.  But it would still not be something you
would want to install as part of an installer for LilyPond.  It's more
like it would be nice to have some functionality looking for an
installed version of Emacs and integrating with it.

I think Frescobaldo might be a nice fit by now (as far as I understand,
it does no longer require KDE, merely Qt).  It does quite more than
LilyPad, integrates nice with a GUI, and still does not obscure working
with LilyPond, like the applications NtEd and similar do.

If you take a look at <URL:http://frescobaldi.org/download>, you'll find
that it is almost but not quite in a situation where we could use it on
all supported platforms when precompiled.  It would be more challenging
to let GUB actually compile it (and dependencies).

A perspective for delivering it as an integrated part of installers for
those platforms where we have to offer a complete bundle due to a lack
of sane packaging systems.  Oh, by the way: any of the OSX users have an
idea what is up with Fink?  I see

<URL:http://pdb.finkproject.org/pdb/browse.php?summary=lilypond>

    lilypond            2.12.2-1        GNU Music Typesetter
    lilypond-devel      2.13.5-1        GNU Music Typesetter

Which is not exactly impressive.  But I am digressing.

The point I wanted to make is that if one can cut out a coherent set of
requirements, it might be possible to get people from the Frescobaldi
community involved, and challenged into making MacOSX work if that's a
precondition for getting Frescobaldi into our installers.

The problem is that we can't even cut out a coherent set of requirements
for keeping LilyPad supported, let alone start on something new.  So it
would be clear that whoever was going to get interested in this project
would have to start with a _lot_ of fumbling in the dark with uncertain
outcome.

-- 
David Kastrup