Re: does anybody care about LSR?

[Valentin Villenave]

On Tue, Jun 28, 2011 at 5:48 PM, Graham Percival
<address@hidden> wrote:
> As far as I'm concerned, no I don't care about LSR;

Shocker. What else is new? :-)

> the people who
> wanted it in the first place aren't maintaining it; we haven't had
> a flood of users volunteering to take care of it.  This experiment
> with "user-generated content" hasn't shown a clear net benefit to
> the project, and as more and more people use lilydev and send in
> patches, the need for something like LSR lessens.

I must disagree. People have posted tons of interesting snippets
there, ranging from the basic "how to display a whole note" to
advanced tricks used by Eluze, Gilles Thibaut, Jay Anderson&al. (there
are even quite advanced snippets by Nicolas or Reinhold).

FWIW, I use it on a daily basis whenever I have scores to publish
(which happens to be the case at the moment).

> I'm suggesting that we just dump the whole thing on Phil.  He can
> choose how picky (or not) to be about explanations, indentation,
> looking for duplicates, etc.

I wasn't aware that there were so many people wasting your time (and
other devs') repeatedly asking about explanations, indentations etc.
But yes, I am certainly confident Phil can handle it just fine. I can
help too (I know my track record with LilyPond is awfully bad these
days but in spite of every task where I've dropped the ball, I don't
think I've left a single LSR-related request unaddressed in the past
four years).

> If somebody here *does* care, then speak up.  Please note:
> 1. nobody is offering to touch the code behind it.  So don't say
> "hey, it would be great if LSR could automatically xyz" unless you
> think you can program the xyz yourself.

There certainly /are/ things "we" can do in a long-term perspective,
but probably not within Sebastiano's Java/Tomcat/ERW design. If even
Reinhold has never been able to tweak it, then there's definitely a
problem here.

However, I could see a PHP reimplementation happening (for example on
top of a MediaWiki installation). Probably can't (and won't) do it on
my own, but there are other LilyPond-related online-2.0 tasks I'm
working on right now that *could* possibly help in the future (again,
long-term thinking).

> 2. anybody with the source code can do much more efficient work by
> editing stuff in git directly.  The only point of LSR is to
> provide a quick, easy, automated repository for non-git people, so
> whenever somebody with git access touches LSR, it's a net loss for
> the project.

I think I get your point, and it does make sense e.g. for snippets
that are meant to live in the git tree (such as doc snippets,
/input/new or whatever the kids are calling it these days). But please
remember you're speaking from a git-centric point of view: many users
have very neat tricks to share (including advanced stuff that isn't
really suitable for upstream integration yet, such as Nicolas' Scheme
engravers, or my own "composite" dynamics, and other "niche" tricks
that are most useful to a handful of people (fingering charts,
thingamabobs etc.), but wouldn't make sense in the main distro.

You make a valid point in advising people who can to focus on the main
source tree rather than the LSR. And you're welcome to not be
interested in, and totally disregard it (again: what else is new). But
do keep in mind than some of us do care, and feel free to point people
towards Phil or myself.

Cheers,
Valentin.