Someone just donated me 100 Gold Reward coins on Reddit just for pointing out the TMFX cache setting. This is worth real money, because he was so happy to finally shave off load times on his otherwise high-performance rig. This seems like a good indication that this TMFX default means trouble to users out there who do not know about the setting.
On a 5900X CPU TMFX' caching roughly doubles my load times from less than 5 seconds (100+ modules active) to nearly 10 seconds. This also applies to every F5/reload or relogin using Firefox, whereas it only seems to apply on first login using Chrome.
Usually only a few different effects are used at any time and that load/compilation times of single effects are often hardly noticeable (unless it's Mega Ice Shell). On the other hand many users don't know about TMFX' cache setting and just assume that their load times are affected by the total number of modules.
Users may not benefit from caching, but suffer from it being enabled by default.
Users don't seem to know the setting exist, even less so that it affects their load times.
Users may assume that their total number of modules is what slows load times considerably when in fact it is TMFX' caching default.
Hello.
Someone just donated me 100 Gold Reward coins on Reddit just for pointing out the TMFX cache setting. This is worth real money, because he was so happy to finally shave off load times on his otherwise high-performance rig. This seems like a good indication that this TMFX default means trouble to users out there who do not know about the setting.
On a 5900X CPU TMFX' caching roughly doubles my load times from less than 5 seconds (100+ modules active) to nearly 10 seconds. This also applies to every F5/reload or relogin using Firefox, whereas it only seems to apply on first login using Chrome.
Usually only a few different effects are used at any time and that load/compilation times of single effects are often hardly noticeable (unless it's Mega Ice Shell). On the other hand many users don't know about TMFX' cache setting and just assume that their load times are affected by the total number of modules.