RE: Evolutionary User Strategery

[Fairchild]

Alexander -

Best proposal I've understood.  If/when implemented I see only two
continuing costs.  The user must maintain selected multiple versions - seems
reasonable.  Developers must make stable versions available in perpetuity,
but continuing support is not necessary - also seems reasonable.

Anybody willing and able to implement?

                            - Bruce

-----Original Message-----
From: Alexander Brock [mailto:address@hidden 
Sent: Monday, July 17, 2006 3:42 PM
To: Fairchild
Cc: address@hidden
Subject: Re: Evolutionary User Strategery


Hello,

> However, modifying 'finished' scores to be acceptable by the 
> latest version is not reasonable.  Upgrade modifications require 
> significant effort.  The convert-ly program helps, but misses a lot. 

One solution could be to write a wrapper which reads the \version and 
decides which installation of lilypond must be used. If there is no 
exact match just use the next higher version. That would increase 
execution time by something like 1/20sec on normal PCs.

Alexander