Re: compilation error

[Jan Nieuwenhuizen]

Mats Bengtsson writes:

> I have to agree with others on this list. Jan and Han-Wen do a
> terrific job with LilyPond. One of the few objections I have is
> that they sometimes are too eager to adopt new technology.

Yes, possibly.  The point is that I believe in open source, and broken
software needs to go or get fixed as quickly as possible, especially
when it takes time to work around.

> Of course it would be nice to be able to view pictures with
> examples also in the info version of the document, but I see
> absolutely no reason for LilyPond to be a pioneer for this
> feature.

I can think of a reason :-) LilyPond's info docs are the ones that
gain most by adding images.  Much more so than say, gcc's.

> When the technology has stabilized and is available
> on most installations, we can take the step.

Adopting it early is a way to stabilize it.  I wrote the technology a
year(!) ago, let's get it being used already.

> I'm afraid you underestimate the number of people who actually
> compile LilyPond themselves and have to struggle with development
> versions of other tools just to be able to build LilyPond.

Possibly.  We try our best to discourage people to do so.  I see no
reason whatsoever for a user to compile LilyPond.  There are fairly
recent development snapshots for Red Hat, Debian and even Cygwin
nowadays.

> At the moment (at least before 2.1.28), the situations is fairly
> good, I only needed to upgrade Guile and install mftrace with friends

(note that you don't need any custom software to build LilyPond on
Debian/unstable)

> to be able to compile LilyPond on a standard RedHat 9.0 machine.
> Please, don't make this situation worse.

Ok.

Jan.

-- 
Jan Nieuwenhuizen <address@hidden> | GNU LilyPond - The music typesetter
http://www.xs4all.nl/~jantien       | http://www.lilypond.org