View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0003121 | ardour | bugs | public | 2010-05-10 13:09 | 2010-07-06 18:16 |
| Reporter | oofus | Assigned To | cth103 | ||
| Priority | normal | Severity | major | Reproducibility | always |
| Status | closed | Resolution | fixed | ||
| Platform | Dell D830 core2duo T9300 2.5GHz | OS | Mandriva | OS Version | 2010 |
| Target Version | 3.0-beta1 | ||||
| Summary | 0003121: When sending MMC and relocating the playhead with transport in play, Ardour 3 sends the wrong sequence of MMC commands | ||||
| Description | When sending MMC and relocating the playhead with transport in play, Ardour 3 sends the wrong sequence of MMC commands. Ardour is sending locate then play, then locate. This second locate causes the external machine to locate and stop. The sequence just needs to be locate followed by play. Note: send MTC is off in this situation. | ||||
| Tags | No tags attached. | ||||
|
|
This seems to have changed slightly recently. Every now and then a sequence that causes the correct behaviour is sent, even though this sequence looks ultimately incorrect as well, in as much as it sends too many messages. This sequence seems to be locate, locate, play. One too many locates ! The predominance is still for locate, play, locate to be sent though. |
|
|
I can see why this happens, but I'm rather afraid of breaking other things by tinkering. I think we need a list of the different situations in which transport locate / play occur so I can check. |
|
|
Is there not just one code path that deals with the transport transitioning into play regardless of what happened prior to that ? I'll think through where locate/ play happens and add it as a further comment. |
|
|
there should only be Session::start_transport() and Session::stop_transport. however, some of the work of stop transport is done asynchronously in a different thread, so stop_transport() is not quite as clean and simple as it may first appear. likewise, locate is done partly in a realtime thread, then partly in the butler thread and then is finished in the realtime thread again. |
|
|
This should be better, or fixed, in SVN. |
|
|
Fixed. |
|
|
fixed |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2010-05-10 13:09 | oofus | New Issue | |
| 2010-06-14 15:21 | oofus | Note Added: 0008239 | |
| 2010-06-14 20:47 | cth103 | cost | => 0.00 |
| 2010-06-14 20:47 | cth103 | Target Version | => 3.0-beta1 |
| 2010-06-25 21:31 | cth103 | Status | new => confirmed |
| 2010-06-25 21:57 | cth103 | Note Added: 0008328 | |
| 2010-06-26 11:03 | oofus | Note Added: 0008330 | |
| 2010-06-27 15:52 | paul | Note Added: 0008336 | |
| 2010-06-28 17:26 | cth103 | Note Added: 0008348 | |
| 2010-06-28 17:26 | cth103 | Status | confirmed => feedback |
| 2010-06-28 20:50 | cth103 | Note Added: 0008352 | |
| 2010-06-28 20:50 | cth103 | Status | feedback => resolved |
| 2010-06-28 20:50 | cth103 | Resolution | open => fixed |
| 2010-06-28 20:50 | cth103 | Assigned To | => cth103 |
| 2010-07-06 18:16 | oofus | Status | resolved => feedback |
| 2010-07-06 18:16 | oofus | Resolution | fixed => reopened |
| 2010-07-06 18:16 | oofus | Note Added: 0008402 | |
| 2010-07-06 18:16 | oofus | Status | feedback => closed |
| 2010-07-06 18:16 | oofus | Resolution | reopened => fixed |
