Hi, gentlemen,

I checked the performance for several examples in firefox. In the main.js we can see many performance.now(), which can be found in b4w in the end. It can cost some time.
So to reduce client's side overhead, I would consider detecting the client on the sever and then send the tailored b4w. Osg is class based, I honestly don't believe it can be faster. Maybe just I am aged

Reply to post of user aldine_jade17
where should I put the code? I don't get it Am I going to open a notepad and place the code there?

Cool, I believe you can do that, also can try with chrome firefox app designer, or try intel xdk. You can debug there the awsd problem and have more fun

linkhttps://www.blend4web.com/en/article/119/

~/blend4web_sdk_free/deploy/tutorials/examples/
make a game example gave the details about how to make a game.
I believe Golem is a good example. When someone login, there will be a spawn of "Golem".

The usage I could thinking about:

1. Make beautiful Web, mobile or PC games.
2. Use it in APP design. Put a shiny 3D model on the APP will be cherry on the cake.
3. Charming website with interactive 3D, It can be immersive with stereoscopic images.
4. Static 3D model with free view points selection.
5. Industry design, Microsoft Hololen holography application …
6. Real time 3D communication, Intel 3D real sense image sensor brings real world in 3D. B4W can be used as a channel for true 3D communication.

Anyway just for fun.

Dear great developers and all beloved B4W users,

Level design and AI flow chart are two intuitive parts of fun for making a game. With sculpt mode in Blender, we could easily make caves and tunnels, which might not be easily done by other great tools. However AI part will be mostly rely on hard coding.

I believe most of AI flow charts nowadays could be influenced by BGE logic or vice versa. Graphic node view reduces the work load by replacing hard code with node connections. NLA is the only graphic UI logic tool we have now. It could be natural to try to use NLA to approach Golem first. NLA is cool, I can easily get hundreds of lines. But debugging it is painful, at least for me. The page param in NLA sheds some light to AI automation actually. Well it is inside JSON, if I am not wrong. So it could be hard to know what is exactly going on.

May I have your opinions regarding AI implementation in the future?

Thanks in advance.

Ответ на сообщение пользователя dallas3d
3. Any other ways to go from blend4web to immersive virtual reality with a head mounted display?

Cool!

Hard truth from Unity manual:

Each VR device has a different plugin.
….

So I believe WebVR is a good start.
Some slides
Change 3 to b4w on the slide above.
Get the threeVR from git. Change it to b4w friendly js.
Debug & fun.

Ответ на сообщение пользователя veliefendi
Thanks for your consideration. If i can succeed to learn and use, i like make something like this:
http://3d.cl3ver.com/uWfsD?tryitlocation=3

If i can succeed, i would like to share my knowledge whit Blend4Web comminty.

Webgl 1.0 gets 65k vertices limit. cl3ver seems to use some tricks. Looking at their shadow closely, we could find it is not a straight line.
[url=https://youtu.be/R1ROxTnD3E0 could be a start for how to build a city on blender.

Besides, since B4W is event triggering, you will have huge advantage for interior design. (just enter another level)

Hello,

Thanks for your article.

May I know whether you would limit yourself by the level of detail, LOD, requirement when you start the conceptual work?


Thanks.