Re: Lilypond error behaviour

[David Wright]

On Tue 19 Apr 2016 at 13:22:50 (+0200), Stephan Neuhaus wrote:
> On 2016-04-18 20:29, David Wright wrote:
> >That begs the question. How do you define "clearly malformed input."
> >If it is malformed, it can't be clear.
> >
> >Of course, if you mean "clearly-malformed", then I contest this
> >vehemently.
> 
> I think I was precise enough.

Yes, the fault was with me. I failed to guess whether you meant
"clearly-malformed input" or "clearly malformed-input" and so
I tried to cover both possibilities in turn. It's just a case of
"man eating tiger".

> But I must say that I don't feel comfortable with the tone of this
> discussion, so I will, with respect, bow out of it.

I've felt comfortable using LP for nine years and, until this thread,
hadn't given much thought to how errors are treated. I was more
concerned about "mistakes" in specifying what I wanted and got.
LP is great because you can work on mistakes even before all the
errors have been eliminated. That really helps when you're working
to tight deadlines.

Now it appears that there are people who want to change that
because they want to add some sort of post-processing by machine
(embedding LP in Makefiles, or into a server), or even just the
imposition of some tidier design principle that has nothing
to do with printing music.

I think I'll risk making you uncomfortable in order to argue
against arbitrary non-generation (or deletion) of PDF output.
Sorry.

Cheers,
David.