> purposes. I would like to be able to see, study, and manipulate the > performance of a score, not just have it played to me, I would like to > be able to hear other people's interpretations while studying the same > score, I would like to be able to use linear track-style and other > block or structural editing operations to edit a score structure, and I > would like to be able to derive likely scores from performances and > experiments.
Mmmm, well, all of this sounds to me more as belonging to the AI realm than any other thing, if I am getting it right. RG would then become more of a sort of technical infrastructure providing MIDI and notation capabilities to performance and analysis algorithms, which would allow perhaps for some kind of manual (even realtime) interaction. As you well know, but let it be stated once more for the benefit of the casual reader, the greatest challenge here is how to translate between fuzzy, highly symbolic information (the score) and the crude reality (a live performance) and then keep them synchronized when you change any of them. A tall order, indeed. > notation software, but it isn't a knock-off of any single program. Hmmm, it is the intention what counts, isn't it then? > But to set out deliberately to produce and distribute an exact > replacement for an existing proprietary program, unless there is a > really strong necessity, is not a righteous thing to do. To replace Well, there is always that discussion about how an idea can be the sole and rightful property of one single person or corporation. Nevertheless, I agree that just cloning is not ethical. But you forget mentioning one important open source point. The freedom to improve or tailor the software to your needs. As good as it is and addresses my score writing needs, many times I have thought "if I could add/change this in Sibelius..." I have been tinkering with Sibelius' script language and plugin writing, but there are some base flaws in the internal data structure (how the score is organized in memory) that prevents things like different time signatures in two different but simultaneous parts, for instance. Besides, I'd rather contribute to improve an open source project than a commercial one. AFAIK, mscore interface is inspired in Sibelius' because it works very well and is thus a very good starting point, but I heard from Werner that he doesn't have any interest in producing a faithful clone. For starters, he is not copying Sibelius' internal data structure (bless him for it) and I think he gets many (most? all?) engraving ideas from Lilypond. Cheers, Luis
