Traktor needs an API

NI Hopeful

I'm not very privy to the latest and greatest these days across all the different DJ suites, but I have been DJ for decades and am a software developer and do not understand why Traktor does not have even a simple API to read currently playing tracks.

Years ago I built some software for myself to organize/tag my music in the way I see fit so I could log and rate mixes, get mix suggestions, create playlist etc. I've so badly wanted access to a Traktor API so I could just automatically give myself suggestions instead of having to switch over to my app and manually type in the search. Imagine being able to get my own suggestions in an overlay and then click on it to load it in a deck, not even having to touch Traktor, just using my own plug-in.

I just searched rekordbox and serato and it seems they too lack an API. This seems like such an easy way to differentiate from them. Get some traction with hot third party plugins and it will give people a reason to switch to your software. With the rapid advancement in machine learning it would only be a matter of time before someone built a superior cataloging/song suggestion plug in. It would also be trivial to tie videos directly into tracks as well.

Am I missing something here? I realize the software must be a mess given it's age and probably not easy to extend but what gives? Are all these companies really going for the walled garden approach?

rmn

Yeah exactly my thoughts. You can get this to work by modifying the Traktor QML files. There is a little JS engine in there that allows you to do http calls and read certain properties. It is a bit of a hassle to setup but if you know your way around some debugging tools you should be fine.

Sadly, it is a walled garden approach, yes. Currently there are 10 options: either you buy a NI Device or you get really good at reverse engineering MachO stuff.

SanderSantema

That would be the dream. I would really appreciate just a specification of the collection.nml xml file as well, there's weird stuff (i.e. tags with unknown function or potentially unused etc. etc.) in there which makes scripting of it quite hard.

rmn

Are you referring to id3 tagging? Because that ****** is the wild west.

https://community.native-instruments.com/discussion/comment/38338#Comment_38338

There are definitely ways to figure out .nml, .tsi and all other the custom and proprietary stuff NI is doing. But Frankly, after reading into all developments and browsing the forum, I am no longer convinced that investing more time in Traktor is wise. Then again, in my opinion, there is no viable alternative. Traktor is really good at its core functionality. It sounds great. It is stable. It is fast. But everything other than core functionality is pretty much broken. Over the years I pulled a lot functionality out of Traktor. FX presets, custom deck states, dynamic mappings, macro controls, scaling, and playlist management (currently in process). Basically, I use Traktor as an engine.

Brett Whittaker

They should chose a business model of open model publishing of their standards and interfaces instead of being so proprietary.

nskcharles

As a software developer there are a lot of things in Traktor that are tantalizing and frustrating. A full object model and scripting language would do the trick... :-D