Is there any progress on integrating a Games DB?
#16
Bumping this question if there has been any progress on integrating a Games Database to enable a native Game Library view in Kodi?
Reply
#17
It doesn't look like much has happened, Wintermute0110's document looks like a reasonable design to get started with. It could probably be pared down to only a few fields to simplify the initial PR so you have a minimal database that can be scraped to and holds name and platform I guess.

I think its more that the work to do this looks pretty thankless from a glance at some of the database code. Lots of copy and pasting with a bit of renaming so its not particularly exciting to do, hence no one rushing to implement it assuming you work from duplicating the code for one of the existing databases.
Reply
#18
(2021-03-28, 20:08)zachmorris Wrote:
(2021-03-24, 17:42)RockerC Wrote:
(2021-03-19, 14:04)docwra Wrote: I personally don't see whats wrong with another simple Sqlite db, similar to how TvShows works, but the developers were talking about refactoring the entire system at one point at which I lost interest Big Grin
(2021-03-21, 07:56)Wintermute0110 Wrote: My idea is to take the music database and start from there. The database backend will be SQlite.

Looking historically at all those other linked threads that mostly discuss huge databases refactor attempts for Kodi, so please be aware that it could be a risk that Kodi's core developers will not accept large changes this time either if a PR is asking for too many changes to be done to the database schema, especially if any significant changes affect the other databases/libraries in Kodi.

Still, I will be crossing my fingers and wishing for the best possible outcome, as at this point, I think that any native games database making it into Kodi would be better than no native games database in Kodi.

Personally, I had hoped before that someone would first just create a very simple and basic native games database that would quickly get approved for merging directly into Kodi 'as-is' and then later once that native games database is in there that could later be extended and improved. This as it is usually harder to say no to continuous smaller changes with minor enhancements and imporovements.

I'd think this is the obvious way forward as a starting point - copy/paste from video or music db code and replace it with game, likewise change the metadata properties to be in-line with whatever the game metadata property names are, and then shoehorn it into kodi as an extension from video, music, images etc to video, music, games, images, etc. Anything more complicated than that will likely never get merged, anything more simplistic wouldn't be taking advantage of the codebase already present. I'm sure this is a monumental task regardless of how it's handled.

At this point, I'm personally a fan of this approach. We're coming up on a decade since the inception of retroplayer. It's puzzling to me why "Games" within Kodi hasn't had more traction with regular Kodi contributors or other people capable of contributing to Kodi. I was so excited when I first heard about retroplayer, I thought it was going to be an all hands on deck type of situation. Video Games are in the DNA of XBMC/KODI after all. I scratch my head every time I see a Kodi update released. --PVR gets x,y,z improvements. Cool I guess....

-take this with a grain of salt. I'm a user that has never contributed to a single line of code to Kodi.
Reply
#19
I think its more that retro gaming is more niche and perhaps doesn't appeal to many of the core contributors. There are also other couch solutions that are more mature which is probably why it doesn't have as much traction in the user base who might otherwise have been interested in the featre. Personally I'm a big fan of how this makes kodi an all in one solution for all couch entertainment and might even be willing to do some of the grunt work of duplicating one of the existing data base designs, but I'd want some steer from someone more familiar with the code base as to what needs implementing for it and how it needs plumbing into the wider code base.
Reply
#20
(2022-11-09, 02:30)dickalan Wrote: At this point, I'm personally a fan of this approach. We're coming up on a decade since the inception of retroplayer. It's puzzling to me why "Games" within Kodi hasn't had more traction with regular Kodi contributors or other people capable of contributing to Kodi. I was so excited when I first heard about retroplayer, I thought it was going to be an all hands on deck type of situation. Video Games are in the DNA of XBMC/KODI after all. I scratch my head every time I see a Kodi update released. --PVR gets x,y,z improvements. Cool I guess....
Don't be fooled, that large number of changes equals large number of contributors, you might be surprised how small the current developer team is and most of the PVR related changes boil down to the contributions of < 3 active devs (numbers might be totally inaccurate so, but that's how it feels) - well, the same goes for retroplayer and with garbear being busy, the (by far) most active retroplayer dev, can't contribute huge changesets at the moment.
In the end it's volunteers work that means you rely on devs spending their freetime (or a fraction of their freetime) to do stuff - which as you can see, is not sufficient to implement all the features that our userbase is wishing for.
That being said, I don't think anyone is going to block contributions towards having a game database, I pretty much gave up on refactoring the whole thing, cause I simply lack the time to do so - and we've never blocked any db changes due to a "potential" refactoring at some point, for example there's been rather huge work on musicdb in the recent years.
Reply
#21
(2022-11-18, 17:51)OmniBlade Wrote: I'd want some steer from someone more familiar with the code base as to what needs implementing for it and how it needs plumbing into the wider code base.

That's kinda hard to give upfront.
You'll most likely need:
* A new window for listing games: See GUIWindowVideoNav
* An implementation of the vfs: See CVideoDatabaseDirectory
* The actual database implementation
* Scraper
* JSON RPC library similar to what's there for Video and or Music
* Filter implementation, sorting etc.

Workig on it, you'll find out, I probably forgot a trillion things in my listing.
Reply
#22
(2022-11-18, 17:51)OmniBlade Wrote: Personally I'm a big fan of how this makes kodi an all in one solution for all couch entertainment and might even be willing to do some of the grunt work of duplicating one of the existing data base designs

@OmniBlade Have you given this any more thought? Just to plant an initial seed that others can then help to grow. It does not sound like you would step on anyone's toes.
Reply
#23
As @a1rwulf has pointed out there are a lot of moving parts that need to be replicated. While the mind is willing the flesh is weak and I'm not sure I can dedicate enough time to it to actually drive it forwards myself. I would dearly love to see retro gaming get even better support in Kodi though as the integrated experience is just so much better than the dark days where you had to launch separate processes.
Reply
#24
FYI, Kodi has applied to Google Summer of Code for 2023 and on its idea page they list a bunch of gaming-specific ideas, including Game Library with a Games Database:

https://kodi.wiki/view/Google_Summer_of_Code/2023

https://kodi.wiki/view/Google_Summer_of_...ific_ideas

So let us hope that both Kodi gets picked for GSoC 2023 and that there are one or more qualified students that are interested in Kodi and the gaming interface part of it.
Reply

Logout Mark Read Team Forum Stats Members Help
Is there any progress on integrating a Games DB?0