QML mapping requirements#6
Conversation
| 1. User Requirements | ||
|
|
||
| 1.1 Good performance, responsive scratching (Real-Time) | ||
| 1.2 Easy reassign buttons and sliders | ||
| 1.3 Easy map a new controller from the scratch | ||
|
|
||
| 2. Mapping Developer | ||
|
|
||
| 2.1 Everything at one place, one file, folder, pack | ||
| 2.2 Easy to share | ||
| 2.3 Reuse common programming skills | ||
| 2.4 High abstraction level | ||
| 2.5 Complete API documentation | ||
| 2.6 Good IDE support .. code completion. | ||
| 2.7 On the fly edits, no Mixxx restart. | ||
| 2.8 Stable: Compatible to future Mixxx versions | ||
|
|
There was a problem hiding this comment.
- User Requirements
1.1 Good performance, responsive scratching (Real-Time)
I believe this can be achieved, however scratching is IMO orthogonal to QML. The missing piece here is the ability to take message timestamps into account, which has nothing to do with QML perse and is mostly an API extension in general.
1.2 Easy reassign buttons and sliders
I think this depends on the definition of easy. If the mapping is written in a Components-oriented fashion, it won't be just simple point-and-click, but we can still retain the C++ dispatcher code that is point-and-click friendly.
1.3 Easy map a new controller from the scratch
This also hinges on the definition of easy. Knowing QML and MIDI fairly well, I'd say its easy. For someone that has never opened a text editor with code, not so much.
2.1 Everything at one place, one file, folder, pack
2.2 Easy to share
That should be fairly achievable for QML, however it does require a little bit of automated work to make the mapping "export-ready". So the current plain copying files wouldn't work and instead we'd need an export button that prepares the mapping accordingly. That prepwork would primarily consist of ensuring the built-in shared modules are bundled correctly.
2.3 Reuse common programming skills
2.4 High abstraction level
2.5 Complete API documentation
2.6 Good IDE support .. code completion.
2.7 On the fly edits, no Mixxx restart.
2.8 Stable: Compatible to future Mixxx versions
Should all be possible, especially when we can take advantage of QMLs built-in API versioning mechanics. IDE support hinges a little bit on the maturity of the qml-language-server.
This is a first draft.