Yes you are correct!Sharlybg wrote: Tue Dec 11, 2018 7:21 pm It only give me the feeling that it is there to avoid newbie to do things wrong.

Yes you are correct!Sharlybg wrote: Tue Dec 11, 2018 7:21 pm It only give me the feeling that it is there to avoid newbie to do things wrong.
Multiply looks like a very bad way to control texture color to me.B.Y.O.B. wrote: Tue Dec 25, 2018 10:48 am I have also thought about this.
In the old LuxBlend there was an "M" button for each texturable property, and if you pressed it, the original color would be multiplied with the texture ("M" as short for "Multiply").
I have done a mockup/hack how this could look like:
Video: https://youtu.be/08W_nVxosuw
Yes, makes sense.lacilaci wrote: Tue Dec 25, 2018 1:29 pm Your mockup is close, but if instead "M" there was value input in % to signify how much effect does the texture has over color (basic mix not multiply)
Yes I even forgot about node wrangler. Super useful when working with 2.8 and cycles.B.Y.O.B. wrote: Tue Dec 25, 2018 1:31 pmYes, makes sense.lacilaci wrote: Tue Dec 25, 2018 1:29 pm Your mockup is close, but if instead "M" there was value input in % to signify how much effect does the texture has over color (basic mix not multiply)
Instead of changing the UI we could also make a shortcut that inserts a mix node or HSV node.
We could try to mimic some of the top features of node wrangler.
Do not listen to legacy requests.. it made vray a complete fuckfest of settings that don't need to be there...B.Y.O.B. wrote: Fri Mar 01, 2019 10:45 am I have created a github issue to keep track of all the different threads about simplification/streamlining of settings: https://github.com/LuxCoreRender/BlendL ... issues/243
I will read them again after the port to 2.8 and I will re-evaluate them. A lot of the ideas in these threads unfortunately require to break old scenes (meaning you would have to downgrade BlendLuxCore to render a scene you saved with an old version). Usually I want to avoid this at all costs for obvious reasons (imagine if your asset library has to be re-done every few months).
Also, I will have to find a solution to the following problem:
When I rewrote the BlendLuxCore addon after the experiences I made with the old LuxBlend, my goal was to cut away as much unnecessary stuff as possible. For example, I felt that light strategy and metropolis parameters were settings that did not need to be changed in 99% of scenes, so I left them out, with the thought that if a power user needed them in his 1% scene, he could export to scn/cfg and manually set them there. However, over time these settings that I deemed unnecessary crept back into the UI, which was caused by user requests.
This is a problem I need to solve sooner or later if I want to prevent the UI to become like the one in the old LuxBlend: 100 settings but in almost all your projects you only tune 10 of them.
It's a difficult task: shielding the user from himself, without taking away his freedom and without holding back power users who really know what they are doing.
Look, I believe you make correct choices, so far there isn't much to complain about.B.Y.O.B. wrote: Fri Mar 01, 2019 12:39 pm I don't think you can write off wrong choices of settings as the user being an idiot. Most of the time, the UI is not communicating enough information (or the right information) to enable the user to make the correct choice for his situation.
For example, I just looked at the path depth tooltip and found that there is none! There should be a tooltip that helps the user decide how high to set the path depth. And if he still chooses a very high value, there should be a warning and if he runs out of VRAM, the error message should contain a list of possible causes, among them the unusually high path depth.
These kind of things get overlooked because a) we developers are too experienced with LuxCore, we don't even hover over the path depth long enough to see that there's no tooltip and we don't choose 500 path depth, and b) we don't do organized testing with real users in a lab (because it's too much effort and cost to handle). Our users could help us by reporting such things as missing tooltips.
When designing software, in my opinion you should always imagine an intelligent but inexperienced user.