Something that limits my actions and not lets me go above very low-poly style
(but this concept sounds impossible)
|
|
Old:
I hate Blender API, but, every time I try to do something like fingers manually it is just a never ending nightmare.
For me it is easier to finally read API documentation and make bunch of solutions on all situations, than to model by self.
Should I try it or it is the worst idea ever, as very often there are impossible Blender API limitations?
P.S.
I’m that type of person who will drag around just 177 vertices until they are not a perfect match, maybe I should just ditch the unnecessary optimization stuff and make the high-poly version first and then decimate it and make manual tweaks to low-poly on need? (Though it still will be about 20%+ geometry, than the low-poly from start thing)
Ah, and maybe I should give it a go in Unity C# (then finish it in Blender), as it is less limited + much more AI chat friendly.
Op, forgot maybe i should try Blockbench or similar, though don’t think I can make stuff like fingers easily in there, though it should rather fit the current stylistic needs.
(Someone used it, is it bunch of cubes on export or a solid optimized mesh, can it make a skinned one for character?)
‘Stylizing’ (for lack of a better term) things just moves the difficulty onto something else. It will still take plenty of practice to make that style look good, alongside being able to portray certain things in a style that would otherwise limit that.
OP, you’re still just looking for easy ways out. Modelling hands is as hard as drawing hands, yes. But if you never tackle the real challenges then you never improve.
Coding this stuff isn’t going to be any easier than modelling it either.
but i did solve the walls and roofs quick in Blender for my ugly needs though:
Potentially it will even compensate the time took to create it, as I still made only one level and a quarter ready of second was corrupted thanks to old Win 10 and local AI issue (i.g?) (still not restored previous progress from backup, btw)
Even the finger case could be turned into a generator, as basically the core is just 5 sausages.
But I guess the easiest and right way in current case is to use Blockbench, as the voxel building is pretty robotic, more close to a programmer way. (and that will solve the ultimate vertices madness)
You wrote that with AI though. You no doubt have no real idea how it works under the hood. Its effectively not even your own work. You’re skill as a programmer and an artist hasn’t gone anywhere.
Oh, ho-ho about that you’re wrong, as the AI IS VERY BAD WITH BLENDER API (Though i definetely needed to actually read the docs to know about limitations on second time, holy crap it can’t even do a multi-line for text in a sidebar)
P.S.
I actually even read the code of main addon, and considering a million of bugfixes, the AI actually made me suffer more than to write it manually after properly reading the doc (which i still hasn’t)
Upd.
Lol with AI the same thing in Unity C# would be more robust and easier to make, though for a main addon it would take 1.3x+ more lines of code
I don’t say that i don’t use the AI, i say that i had to fix like every tenth (in some cases even worse?) line of code of bugs.
I.g. missed the edited message again.
The Unity C# with Claude is no effort and easy (plus i at least have some month in it and finished junior pathway).
But Blender - no no no, no at all, if ur trying to catch me by the hand here
if i was a bad person in this one github, i would’nt even bothered fixing the critical bug that was noticed later, if you delete the default cube (all modes deleting), lol. (but i admit even here i bashed the claude again, it seems that i bashed it so much with Blender that i got blocked from it, lol)