2016 is coming to an end in about 1 month. I decided to write something about VocaDB – where we are we now and where we’re going.
The good news
VocaDB is self-sustaining for the most part. We have a healthy number of active editors making sure that the database is kept up to date. The server is fully funded through our Patreon program. I feel that VocaDB has the feature set it needs to be a useful discography database. Finally, we just reached a milestone of 100 000 song entries – a proof that VocaDB scales as much as it needs to!
The bad news
The main problem is still that the development and maintenance of VocaDB is solely my responsibility. 2016 has been stressful for me overall, and I haven’t had the time or motivation to spend on VocaDB as I once did. Fortunately this doesn’t affect the day to day operations here, just that the website isn’t evolving as much anymore, but maybe it doesn’t need to. Additionally, the tech behind VocaDB frontend is getting old, meaning large parts of the code would need to be rewritten if I’m going to keep improving the website. Needless to say there’s no resources for such massive operation.
VocaDB as a service
The most important part of VocaDB is the database backend: the amount of information we have, the way it’s structured and how it’s accessible through our web APIs.
My greatest wish is that other developers would start building more services around VocaDB’s database: decentralized, specialized user interfaces making use of our APIs and massive amount of information. I’ve written a few such tools on our github page as an example: the rankings generator, UTAU Lyrics Wiki Importer, Vocaloid hierarchy page.
I’m going to focus my development efforts in the backend: the database structure and web APIs, and try to leave the user interface level things to others.
Documentation/editing guidelines (wiki)
Next year I hope I’ll finally rework the documentation. There’s an unbelievable amount of guidelines and best practices for editing on VocaDB. All new documentation will be written in the wiki, because maintaining the documentation there is simply so much easier than adding documentation to VocaDB itself. The downside of having a separate website for the documentation is that it can be difficult to find. To fix this, I’ll look into bringing the documentation content from the wiki directly to VocaDB. The wiki will need to be updated first – it’s based on Creole syntax that no one uses anymore. This means merging changes from the Roadkill Wiki 3.0, which is a large amount of work by itself.
More accurate title translations
I was never completely happy with the content translation system on VocaDB, but the fact is that designing a perfect multilingual website is very difficult. Some time ago, I reworked how translations are handled for lyrics, and I’ve been quite satisfied with that system. I’m going bring the same system to editing title translations (songs, albums, artists) on VocaDB. This means that, instead of the silly “non-English” option, you’d have an actual language option (Japanese, Chinese, Korean etc.) for the original language, similar to lyrics.