Epix-Incorporated
Re-add panic mode
I originally removed panic mode in #527 However with fruther thought Adonis panic mode is actually very useful in cases where Adonis fails to load properly. And instead of just failing it still continues the loading process but enables panic mode. This makes Adonis more resilient when bad PRs or something else break Adonis. And more future proof.
I would like for the formattedchangelog to be made more efficient memory-wise because the client is given a changelog and a formattedchangelog for basically no reason
I would like for the formattedchangelog to be made more efficient memory-wise because the client is given a changelog and a formattedchangelog for basically no reason
Well that shouldn't be a part of this PR
And if we did that Changelog would have to be stored somewhere else than the shared folder because the shared folder is replicated to the client
or you could get rid of the formattedchangelog
Well then the changelog wouldn't be formatted
Adonis should never really fail to load in the first place. It would be a integral flaw if Adonis fails to load, which would be fixed separately. I will let others comment if they have other opinions.
Adonis should never really fail to load in the first place. It would be a integral flaw if Adonis fails to load, which would be fixed separately. I will let others comment if they have other opinions.
True. But when Adonis fails to load there is no visual indicator of this happening. It becomes hard to debug to see exactly at which point the error occurs, and you have to dig in the Roblox console to attempt to find the source, if you have access to it in the first place. A loading error in Adonis can occur from many things: Loader script errors, settings file errors, user plugins errors, cloud module download errors, server core module errors, server dependancies errors, server command module errors, server plugin errors, ACLI injection errors, ACLI bootup errors, client module errors, client plugin errors, client theme errors, client remote communication errors, Adonis netcode errors etc. For errors occuring in the server side of the main module this would provide a huge boost in debuggability, ie. immediate feedback instead of waiting a few seconds to see if Adonis just took a bit of time to load (which it sometimes does), with the added bonus of recovering from errors in some cases.