[DiscordArchive] What about 1 file changed +36,497 -352?
[DiscordArchive] What about 1 file changed +36,497 -352?
Archived author: Bench • Posted: 2023-11-21T18:35:53.571000+00:00
Original source
Was mostly referring to the other parts of the code tidbit since no plans to cherrypick a system for the general core at large
Archived author: Kitzunu • Posted: 2023-11-21T18:36:26.612000+00:00
Original source
there is a basically complete, but abandoned PR for RBAC closed
Archived author: Nyeriah • Posted: 2023-11-21T18:38:10.494000+00:00
Original source
It had a crash I think that wasn’t addressed, rest was ready, back then. It probably has conflicts so we’d need to start over though
Archived author: Kitzunu • Posted: 2023-11-21T18:38:22.136000+00:00
Original source
https://github.com/azerothcore/azerothco.../pull/8418
Archived author: Natrist • Posted: 2023-11-21T18:56:14.141000+00:00
Original source
RBAC is fairly simple to implement. Can probably be done over a weekend if someone is willing to land a hand with the SQL stuff as that's not my area of expertise.
Archived author: Natrist • Posted: 2023-11-21T18:57:32.446000+00:00
Original source
Anyway I don't think it's a big priority right now. I think we should clean up the code as much as possible. I think we should start with the message handlers.
Archived author: Natrist • Posted: 2023-11-21T18:58:39.495000+00:00
Original source
Most message handlers were initially implemented on guesses and many variables are unnamed. If we start there we can clean them up and it will lead straight to refactoring a lot of the code they call.
Archived author: Natrist • Posted: 2023-11-21T19:05:19.371000+00:00
Original source
Also we will probably need to write a command manager class at some point for commands to keep them tidy.
Archived author: Abillister • Posted: 2023-11-21T21:31:21.464000+00:00
Original source
RBAC can be as simple or overtly complicated just based upon however it is designed. I always preferred a database driven process in which the 'logic' is stored in tables vs being in the code. I like 'relationship' data and being able to properly see all of the relationships within the code. This method I find tends to be considered a bit more 'complicated' depending upon how normalized you like your DB to be.
My personal experience with application development was that it was also faster/easier to make changes to 'data' vs code. So by building the logic into the tables and having the code just dynamically execute that there were fewer stages between development and production.
BUT, we also had some parts of our application that loaded SLOWLY because the code was being built dynamically when the page was loaded. BUUUT - that was also in 'classic ASP', which was not exactly the greatest backend to begin with (and it was mostly written by junior developers, myself being amongst that list as I came from a VB scripting side of things as a domain server admin - which used VBS which easily translated into classic asp and simple java).
Archived author: Natrist • Posted: 2023-11-21T21:33:27.990000+00:00
Original source
RBAC is cool but permissions in code are more efficient.