Dmxkeypad

[kripton]

trafficstars

Adds a DMX KeyPad to the simple console allowing ranges of channels to be manipulated at once and it's easier to use on a touchscreen than faders. I don't consider this feature-complete but it should be production-ready. All commands that can lead to strange behaviour (crashes or 100% CPU load) should be impossible due to buttons beeing disabled.

TODO:

• documentation
• add it to the webserver somehow (together with an "clear universe" button as we have on the simple console

Discussion thread: https://sourceforge.net/p/qlcplus/discussion/development/thread/6039d20d/

EDIT (janosvitok): discussion thread in the new forum: http://www.qlcplus.org/forum/viewtopic.php?f=12&t=3091

[janosvitok]

I rebased the commits to current master (to get rid of the merge commits): https://github.com/janosvitok/qlcplus/commits/pr/280

I did some quick tests. It looks nice and useful.

Bug: Pressing FULL several times after "1 THRU 10 AT 0 THRU" each time adds AT

[matthijskooijman]

@janosvitok if you force push your rebased work into the branch used to create this PR, this PR will be updated automatically.

[edogawa23]

I wonder if there still is interest in getting this completed and included/merged. I realize that it's unfinished, but still very useful even in this state, in my opinion.

I have looked at kripton's branch and manually merged it to a new branch in my own fork, to get a rebased version. I'm sure there is a more elegant method to do that, but I still have a hard time getting along with git and github and could not find one. So I apologize beforehand if this makes things more complicated. Here is the link.

The patch is small after all, two new files, a few changes that had trivial conflicts that needed to be resolved(because kripton's branch hasn't been rebased in a long time) for simpledesk.cpp/h, and the qmake src.pro file.

So today in the morning I had a go on adding keyboard input fo the DMX keypad and got it working after ~ 3 hours without much hassle. Here's my mapping:

Channel/value digits: main or numpad digit keys +: main or numpad "+" -: main or numpad "-" AT: "a" CLR: "Escape" key THRU: "t" FULL: "f" BY: "b" Enter: "Return" or "Enter"

Also as a rather unexperienced coder I don't fully understand the code yet, especially that state machine part of it. I see that when a command is considered complete (after pressing ENTER or FULL), the current channel selection is cleared, but I cannot understand why (yet). It would be great to be able to keep the selection and further modify the value until CLR (or a new button for that purpose) is being pressed.

I have used a few hardware consoles and thus can think of several more improvements, but for now I want to find out if there's a chance to "revive" this PR (and interest of kripton, janosvitok or whoever might be able to contribute...)

[mcallegari]

I kept this on hold basically because I planned this feature for QLC+ 5 and didn't want to invest effort on v4. Seeing how it is going (my bad...) I can dedicate some time to this and bring it to a usable state. Dunno when though :(

[yestalgia]

With the keypad being a more promenent mode of input in QLC+ 5, is it worth enabling this to get users used to the concept?

[edogawa23]

I have the extended patch from above ready to apply on top of current QLC+ v4 GIT, but as Massimo seemed to oppose to adding it to v4, I didn't even try and pull request it. For me it's really useful in theater during setting up/testing/focussing and even while writing cues for a new lighting design, I have used it for a while and haven't noticed any bad side effects.

See here, I've just pushed this to github again, on top of current master.

[kripton]

Thanks for bringing this up again. Since QLC+5 has a keypad by now (see https://github.com/mcallegari/qlcplus/blob/master/qmlui/qml/KeyPad.qml), I would propose to sync the two implementations we have now. Either we adapt this PR to match the code of what's inside QLC+5 or we change QLC+5's implementation to what is found here. I would assume the first approach has the higher probability of getting merged.

[mcallegari]

Technically, the only issue here is on macOS, where Simple Desk has serious horizontal scaling issues. Roadmap-wise instead, I confirm that I would prefer to have this as a new feature in v5. Integrating this PR would give one more excuse to v4 people to not even try v5. Right now I received useful feedbacks on v5 from one single QLC+ user - repeat - one single user. To me this means: nobody uses it. It's really depressing.

In any case, there is a KeyParser in the engine now: https://github.com/mcallegari/qlcplus/blob/master/engine/src/keypadparser.cpp and test unit: https://github.com/mcallegari/qlcplus/blob/master/engine/test/keypadparser/keypadparser_test.cpp

So yes, to guarantee a consistent supported syntax this PR should be migrated to using the engine parser first.

[yestalgia]

This seems like a sensible way forward.

I'm going to be dedicating more time to helping out on this project in the coming months including switching to +5 in my professional capacity. Money is one thing but I know it will help more for me to dive in and get my hands dirty.

Don't feel discouraged Massimo! It's a big change but it will be worth it for sure. I'm seeing the same issue happening in GrandMA 2 vs 3 but people are coming around.

Thanks for the discussion.