Closed kampfschlaefer closed 5 years ago
One general remark: I consider switching from qt4 to qt5 fairly major, so the version should switch to 0.6.0 for this.
I thought we we're both using the same repository now, so I just made a new branch. I've deleted my jackmix repository, in fact.
So if you like the Qt5 version you can just merge the branch? I'm not saying it's perfect but it appears to work just like the master branch did.
Best wishes,
Nick/.
Sent from my HTC
----- Reply message ----- From: "Arnold Krille" notifications@github.com To: "kampfschlaefer/jackmix" jackmix@noreply.github.com Subject: [kampfschlaefer/jackmix] Qt5 (#2) Date: Fri, Dec 9, 2016 16:28
Lets have a pull request for that to discuss issues. You can view, comment on, or merge this pull request online at:
https://github.com/kampfschlaefer/jackmix/pull/2
-- Commit Summary --
-- File Changes --
M NEWS (7) D README (16) A README.md (28) M SConstruct (12) M TODO (10) M VERSION (2) R admin/qt5muc.py (17) M backend/jack_backend.cpp (2) M jackmix/SConscript (4) M jackmix/main.cpp (2) M jackmix/mainwindow.cpp (18) M jackmix/mainwindow.h (6) M libcontrol/SConscript (2) M libcontrol/controlsender_test.cpp (4) M libelements/aux_elements.cpp (6) M libelements/stereo_elements.cpp (8) M libgui/SConscript (2) M libgui/abstractslider.cpp (2) M libgui/abstractslider.h (2) M libgui/channelselector.cpp (8) M libgui/channelselector.h (2) M libgui/channelselector_test.cpp (2) M libgui/editablechannelselector.cpp (8) M libgui/graphicalguiserver.cpp (8) M libgui/midicontrolchannelassigner.cpp (8) M libgui/midicontrolchannelassigner.h (2) M libgui/midicontrolchannelassigner_test.cpp (2) M libgui/midicontrolchannelassigner_test.h (2) M libgui/slider.cpp (4) M libgui/slider_test.cpp (4) M libmatrix/mixingmatrix.cpp (10) M libmatrix/mixingmatrix.h (8) M libmatrix/mixingmatrix_privat.h (2) M libqlash/qlash.cpp (3)
-- Patch Links --
https://github.com/kampfschlaefer/jackmix/pull/2.patch https://github.com/kampfschlaefer/jackmix/pull/2.diff
-- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/kampfschlaefer/jackmix/pull/2
Agreed.
Sent from my HTC
----- Reply message ----- From: "Arnold Krille" notifications@github.com To: "kampfschlaefer/jackmix" jackmix@noreply.github.com Subject: [kampfschlaefer/jackmix] Qt5 (#2) Date: Fri, Dec 9, 2016 16:29
One general remark: I consider switching from qt4 to qt5 fairly major, so the version should switch to 0.6.0 for this.
-- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/kampfschlaefer/jackmix/pull/2#issuecomment-266057076
@nickbailey of course we are using the same repo now. but still pull requests are a good way of documenting branches and the discussions and reviews on them;-)
Our whole company works on the same repo, still we use merge requests for everything.
OK, I'll do one next time I'm at a computer.
Sent from my HTC
----- Reply message ----- From: "Arnold Krille" notifications@github.com To: "kampfschlaefer/jackmix" jackmix@noreply.github.com Cc: "nickbailey" nick@n-ism.org, "Mention" mention@noreply.github.com Subject: [kampfschlaefer/jackmix] Qt5 (#2) Date: Fri, Dec 9, 2016 20:32
@nickbailey of course we are using the same repo now. but still pull requests are a good way of documenting branches and the discussions and reviews on them;-)
Our whole company works on the same repo, still we use merge requests for everything.
-- You are receiving this because you were mentioned. Reply to this email directly or view it on GitHub: https://github.com/kampfschlaefer/jackmix/pull/2#issuecomment-266115827
That conflict was just me not properly understading markdown properly. I've resolved it now.
Here's what I think I've done:
toAscii()
method to toLatin1()
, andgetInteger()
method to getInt()
jack_port_set_name(port, new_name)
with the intention that all clients should use jack_port_rename(client, port, new_name)
instead. This is in backend/jack_backend.cpp. Their intent is to ensure that all jack clients listening to state change information be notified, and that this doesn't necessarily happen with jack_port_set_name()
(or at least that's my understanding). ::jack_client_t *client
is a private member of the JackBackend, so I just used that. Maybe I should check it's non-null? I think that would have resulted in an exit well before this point in the execution though.I believe things are working in Qt5 after this, but I've not checked it really, really thoroughly. I did also get it to crash once when adding and deleting inputs and outputs, but I can't reproduce that.
In a different universe where I had time to work on this, there are one or two improvements on my wish-list. Note this is a wish list. Not a demand list. Unfortuantely, I am very bad at building Qt widgets. You might say I would benefit from the practice :wink:
Qt5 branch now compiles with C++11 (the default these days). throw(exceptions_list) is deprecated apparently.
I've got a session tomorrow when we'll be using JackMix. I have just done the patch for it (quite simple; two ins and two outs). But I did notice there's not currently a way to split 1->2 and 2->2 channel pots if you change your mind so you can go back to the original matrix of 1->1 channel ones. I might look in to doing that. Session first, though :)
The other thing is there's no obvious way of seeing if there's any signal on the inputs and outputs. A big VU/Peak type display would be too much. I might play with the idea of changing the colour of the control knobs green at -40dB, then to orange at -18dB and red at -6dB for example. I've not looked in to that at all yet. It would just be a bit of a help if you didn't know whether the jack was routing signal in and out properly.
I reckon with the de-zipping and the fact that @dsheeler has done some testing, I should merge the qt5 branch to master and bump the minor version number.
Is there any reason not to do that, @kampfschlaefer ? qt4 is pretty much dead now, I think.
@nickbailey I did the version bump to 0.6.0. And some minor things, only compile tested.
If this still works for you, please merge to master and tag that merge-commit as 0.6.0. Then we can decide on which channels to announce this release… and if at all…
Lets have a pull request for that to discuss issues.