Re: Evolutionary User Strategery

[Trevor Bača]

On 7/8/06, Erik Sandberg <address@hidden> wrote:
> On Friday 07 July 2006 16:46, Fairchild wrote:
> > LilyPonders -
> >
> > The only reasonable solution is to maintain upward compatibility in the
> > LilyPond processor.  New features should be added without changing existing
> > syntax.  If it is deemed absolutely necessary to change semantics or define
> > conflicting syntax, provide for optional interpretations based on the
> > version specified.  Older ly files should generate consistent results as
> > LilyPond migrates.  More exhaustive regression tests are necessary.
>
> This one is not possible either: Let's say that you typeset a score in v2.4,
> and that lily, due to a bug, makes some text collide with some notes. Let's
> say that you work around this by tweaking the extra-offset of the text. If
> now the bug is fixed in v3.0, your text will look ugly when compiled in 3.0,
> and you will have to remove the tweak manually, regardless of how clever
> convert-ly is.
>
> Conclusion: Full backward compatibility is not desirable.

Agreed.

The upgrade-tradeoff isn't just a consideration with lily -- it's a
consideration with *any* software. Spent years programming in C and
wanna try python now? Get ready to spend time learning and rewriting
things, but look forward to about 1/10th of the complexity when you're
done. Wanna upgrade your OS? You'll have backwards compatability, most
likely, but with far fewer features than we would expect with a new
lily version. And there's still gonna be stuff that's changed or moved
around.

One of the lily's -- ie, the project's -- strengths is the
introduction of increasingly better syntax.


--
Trevor Bača
address@hidden