Indeed, I’m still using Git for Cantabile, and it really works nicely. I used to just use Dropbox, but periodically ran into problems with conflicted files, and I was never 100% sure I hadn’t lost/overwritten stuff. With Git, everything’s explicit, you can scan what has changed, and revert file changes easier than in Dropbox. Also particularly useful if you have multiple Cantabile machines sharing the same song/setlist/rack files.
Neil
Using GIT to manage Cantabile is certainly not for everyone, but it is truly brilliant.
Hi Roland,
I found the issue here. The old bindings system didn’t support matching on a device id in the MMC sysex message. The new system does and it supports a wildcard mode that will match on any device id (to match old binding behaviour). The problem was on upgrading the old binding it was mapping the device id to 0 instead of Any - and I’m guessing the device you’re using has a non-zero device id.
I’ve made a fix for the next build, but in the meantime, you should be able to just change the device id in the source side of the binding to “Any” by deleting the 0 and it should work:
This was an oversight on my part related to reading back CC values from MIDI targets. Normally this isn’t possible (once you send a CC value it’s sent and you can’t easily read it back). In the previous binding system I had a mechanism in place to track last sent CC switch values so they could be read back.
Anyway, I’ve re-implemented this for the new bindings. Now it works for CC, fine CC and switch CCs. I’m still testing but hope to include in the next build.
Brad
Thanks. I’ll give this a try later this evening. (Eastern US time)
Thanks. I realize this a somewhat non-standard usage, but it allows me to use a regular foot switch like a sustain pedal as a “latching” controller.
Neat, I think I’ll take a thorough look at my utility-rack, I’ve made a way more complicated solution for this. Thanks!
Build 4116 up now, mostly bug fixes:
Thank you for the quick response! The “tooggle” binding works now, just like in the pre-4100 builds.
And the MMC binding worked after I followed your suggestion of setting the Device ID to Any.
Awesome!
I’m thinking, “Surely, more people have taken the plunge,” but does the “12” at the end of the link mean that only 12 people have downloaded the C4116? I must admit, I’ve downloaded C4116 and haven’t installed it as all my stuff was working at C4113. I’m waiting for Brad to post a stable before I convert again.
Just curious… Does anyone know how many total Cantabile users are on the planet? This forum statistics states there are 1.4k users all time and 153 active users in the past 30 days. That would lead me to the conclusion that <10% have tried C4100.
Just posting out loud.
Still haven’t had time to try yet, and as I am still building songs for my big gig in early Aug, I may hold off now until after that.
If it aint broke and you don’t want to break it before the gig …
Yes, the testing on 41xx is very limited at the moment… I’ve deliberately not pushed it into the official release stream yet because I wanted to get some confidence testing with the community here first. I’m ready to push it out to the official release stream as experimental but I’m interstate this week and will wait till I’m back next week before doing so.
Current state should still be considered experimental, but will no known issues.
I’ve been running it since release. A couple of versions hicupped but the solution usually followed quickly.
“404 Not Found”
I’m not sure what’s happening, but copy the link from C4116 to the browser header and change the 6 to a 7, then ENTER, and 4117 will download.
Oops… fixed.
After upgrading I seem to have an issue with a ReaJS script I’ve made to split the program changes from LivePrompter into a few midi messages. It is in the background rack. The first time the module is called it works as expected, but then it suddenly stops sending the midi messages, and I have no clue why. There are no messages to see in the MidiMonitor, but if I insert debug statements it is clear that the statements in the module are executed. So, somehow they are not send. What really puzzles me is that it works the first time. Anyone else that have noticed something like this on the new version? Still works perfectly on the spare laptop running on 4068.
