Ted Felix

Apart from that, the new logic is looking fine along the paths I've looked at so far. I will continue using it and see if anything else turns up. Thanks for testing this. It does appear as if the remaining XRuns are mostly harmless, but I'm sure we can achieve an XRun-free implementation. Hopefully Philip has time to look at it. Otherwise I'll see if I can figure out the remaining XRuns before the 26.12 release.

Just pushed [570968]. Please test latest git. The "new logic" checkbox now enables/disables based on whether "Use JACK Tranport" is checked.

Preferences: Enable/disable new logic checkbox

It seems like the choice to use or not use the new logic should only be available if "Use JACK Transport" is checked. Definitely. This was supposed to be a hopefully quick hack that would just go away shortly, so I didn't put enough thought into it. We were also struggling with finding a good name for the "New JACK Transport Logic" checkbox.

The "Use JACK Transport" checkbox enables JACK transport. Rosegarden will monitor the JACK transport and respond. It will also update the JACK transport based on its own transport. E.g. pressing play in rg will also start JACK transport playback. Doing any work with the JACK transport (e.g. syncing with Ardour) requires that "Use JACK transport" be checked. The "New JACK Transport Logic" checkbox controls whether you are running with the original code which has limitations and problems, but works...

The "Use JACK Transport" checkbox enables JACK transport. Rosegarden will monitor the JACK transport and respond. It will also update the JACK transport based on its own transport. E.g. pressing play in rg will also start JACK transport playback. Doing any work with the JACK transport (e.g. syncing with Ardour) requires that "Use JACK transport" be checked. The "New JACK Transport Logic" checkbox controls whether you are running with the original code which has limitations and problems, but works...

I've updated the FAQ to include the HR timer. Please review: https://www.rosegardenmusic.com/wiki/frequently_asked_questions#what_does_system_timer_resolution_is_too_low_mean

If I'm reading image 2 correctly, that's a 9.6ms error. That would definitely be audible. Unless it's happening gradually and not snapping back at some point. It's more error than I would expect from a 250Hz (4ms) clock. But the HR timer ones look great. So I think the takeaway here is that we should recommend the HR timer for those who can't upgrade their 250Hz kernel. I'll update the documentation and my Linux MIDI Guide to cautiously recommend the HR Timer as an option when only a 250Hz kernel...