Andrew Bernard <[email protected]> writes: > Hi Phil, > > Many benefits. I just rewrote my recent flatten-ly tool in Python, > more as an exercise than anything else. (Thinking vaguely that some > people may find it more maintainable than Scheme. A foolish > notion!). I used Python 2.7, oblivious of the fact that lilypond > requires 2.4.5. So many things that one takes for granted in 2.7 are > missing in 2.4.5 that I abandoned any notion of publishing a Python > version, as I am not prepared to go back to a much less functional > Python and rewrite half the program. > > 2.4.5 is very old now, and as with any system, there have been > enormous strides forward in terms of bug fixes and > improvements. Python 3 is also out, of course, and this is yet better, > although the 2 series still has a strong following. > > While accepting the dictum that if it ain’t broke don’t fix it, this > also has to be balanced with the idea that it eventually becomes > necessary to trade in your car for a new one before it falls to > bits. Since lilypond is under active development, apart from the issue > of lack of human resources, it seems to make sense to me to have its > componentry on an upgrade path as well. > > Python 2.4.5 was released in 2008. That is now a considerably long > time ago. I don’t think users would be happy to continue with a 2008 > release of lilypond, by way of analogy. > > Moving to Python 3 would be good.
Don't really see the cost/benefit now. Moving to 2.7 should be easy in comparison: LilyPond already runs with that on several platforms. It's "just" a matter of updating GUB. No idea how complex that is for Windows etc. Python 3 is not backward-compatible. -- David Kastrup _______________________________________________ lilypond-devel mailing list [email protected] https://lists.gnu.org/mailman/listinfo/lilypond-devel
