mulling over some tools (i’m porting to the SDK) and the byproduct is some of the new stuff adds features enabling the users to create some assets that would have been generated automatically at export time themselves. The question arises do I shift responsibility to the artists (who usually can’t be trusted so I’ll probably have to add export checks) or implement to auto stuff which is quite involved with the sdk.
You are obviously talking about technical and not artistic aspects, so if there is a subjective, solid, reliable and unique solution to the problem, I would definitely go with the automatic solution.
Computers like to do that sort of things for us.
I would also consider how it could be scaled in the future compared with the other one.
This might sound, to someone, like a Microsoft policy “don’t let the user do what we can do for them”, which I am sure many of us disagree, but I dont think this is the case.
I’m erring on the side of automatic, but there is something to be said for “complete scenes” as it were where all files use the same export process. The assets can be visually checked to be correct and in the scene for example.
When you say “visually”, how much of work does it represent for an artist, how many times a day doe they need to do this “tedious” process and how accurate can you ensure they will be?
Most people don’t like routines and artist are not precisely the exception. If at some point it becomes a tedious process, then it is probably that mistakes will appear, along frustration and bad mood.
If it can be automatized, then let the artist use their artistic skills to do artistic work.
For example, if you have 100 boxes in a scene and you have to manually check if none of them has a flipped face, what process would you pick?
If at some point any artist realizes that the “tedious” process could had been automatized, you better get used to hear
we know what, not precisely words of love about the programmers behind.
i usually try to support both. and which is the main one depends on situation.
i don’t like to delegate full responsibility to an artist, but i also don’t like when an artist has no idea how automatic way works. so i give an option – do it manually or automatically. in 90% cases an artist picks an automatic way. but always has a chance make it ‘better’ manually.
that’s another thing i usually try to do – force an artist to check what the automatic way makes. in this case an artist becomes responsible for what will be delivered in to the process (production).
both these solutions are help me to minimize supporting of the tool in the future. half a year latter i don’t remember how this tool works. and very often i am very surprised when see that an artist uses it (successfully and with his pleasure!) quite different than i designed it.
If the task can be done better by a human than a machine, then there is more than one way of doing it. In that case you would have to evaluate in deep the cost-benefit of both.
This is how I structure things too. I don’t lock anything off, but I present an automated way that saves time. All processes can be re-created manually. I see it as a reward system – if you use them, you’ll save time. But it’s right that the artists understand the process.