It's not because the technology is there or quasi there that sampling will automatically be a threat to real musicians. A good violinist is good because he IS GOOD not because he has a Stradivarius, so a good programmer, bearing in mind he has the right tools, will be as rare as a good violinist. What makes a violinist play well? Aside from his finger coordination it's his lifetime experience and having developed fine ear for maximum expression, if the sampling programmer is not trained in a similar way or at least show a highly sensitive ear, I doubt very much he will do miracles. I noticed that people hear about sample libraries such as VSL and expect to do the same as the demos when they have very little music training if any. I think in the future people will not be stunned anymore when they find out a work was done on computer but rather say, "good programming", and whether the job was done by a real violinist or a computer programmer, people won't care much anymore. The machine will be developed to the point where it will simply be ANOTHER tool to make "good" music. Let's remember that we are still in the renaissance of sampling and herb is Palestrina, in 5-10 years more and more great sampler/musicians will emerge and become just as important as Itzhak Perlman as contributor of music. So that's how I see, as another tool to reach the same goal depending on the musical competence in whatever ways of the programer and of course using good sampling. Jay shows a nice balance between the two and is a good example of what one can accomplish with the right knowledge and tool and therefore why not be a candidate to rival live performer, true, a piece that would require longer sustain notes and more legato will still be to the advantage of a real violinist, as William said the variance in expression is too much.... but that's only a question of time before the string lib are refined even more.