I feel that UE has taken the wrong path in script development. there are essentially just two purposes for using scripts in game development: hot updates and reducing programming costs. It seems that there is nothing more advanced than Blueprints when it comes to cost reduction, it even expanded into the TD position. Thus, in my opinion, UE should focus on how to better support Blueprint hot-reloading, instead of creating a new scripting language.
1 Like
My take on this is that we are just seeing the intermediate state Verse is currently in. From my understanding of the GDC Verse Talk it is intended to be more than just another scripting language. It wants to establish a new standard (e.g. everything running in a transaction, getting Meta-Verse-ready). But being a standard that may be adopted by others also means it cannot make Unreal specific assumptions, it needs to be general purpose.
Being a scripting language that’s even built on top of the Blueprint VM as far as I know is just the easiest/first step for Epic to get it integrated into UEFN/Unreal Engine and expose it to a large user base. So the standard can be iterated and improved upon as well. Regarding UEFN/Unreal Engine I guess Blueprint could be converted to Verse similar to how Blueprint-nativization used to work in UE4. Or the Blueprint Editor may even become the graph based code editor for Verse itself. In my opinion Blueprint and Verse don’t exclude each other, they are more of a complement similar to how Blueprint and C++ code complement each other.
Not to mention back in UE3/UDK days we had Unreal Script as the main scripting language but there already was a visual graph based scripting language called Kismet as well. Fun fact: there are still references to Kismet in the code.
Verse is the correct path. C++ is all over the place and can confuse even vets in game dev. Verse is meant to streamline not necessarily make it easy to learn and that is not the issue programmers faces but overly complexity of c++ integration even for simple task that could be done in one line takes a few as you have to call back ol features and then connect with new version of that feature so verse hopefully remove this part.