Originally published at: https://www.world-machine.com/blog/?p=831
Over the last few posts, we’ve talked about how World Machine started, and the fact that the UI is a critical blocker. Last post, I introduced the new framework we’re moving to – QT. As promised, I’m going to try to write more frequently about small bits of the work being done to move World…
Originally published at: https://www.world-machine.com/blog/?p=831
Hmm, why not decouple the preview window and make it float as an option? That way you wouldn’t need to resize the entire sidebar (and save screenspace) if all you need is a bigger preview window. Just a thought…
Absolutely! That’s a good idea; and with QT it’s pretty straightforward to implement. I’d like all of the major UI “sections” to be breakable/drag-able so you can configure everything about the views the way you want them.
However… that can come later.
The primary focus right now is to keep things as simple as possible to make the conversion to LTE happen quickly. I haven’t always been succeeding at this But I’m trying to incorporate only the easiest UI changes/wins for now.
“every button and objects on screen required that its location was set manually and specifically” WHAT!!!
That must be bad , That can not be good…
Fortunately we have QT now…
Luve that new resize window feature!! WTG
Honestly, I still dont see any value in this. This is tremendous rework with not much added value. I had never ever problem with WM GUI. And I am user for cca 6 years. Could be better, for sure, but it does its job. On the other side, there are other areas, which are pain in the as. As artist, I need to iterate fast. If any adjustment takes 2 minutes to know result, it limits my productivity. With just 2x better speed, I could by very often twice as productive, this would be HUGE. So logical step would be GPU support (start migrate at least part of the load to GPU). There are GPU driven tools on the market, so it is definitelly possible (I still consider WM the best by far). Other major pain is texturing/coloring. Quite hard to achieve realistic look, for terrains with more climas. This could be solved by library of various presets, macros, etc… There are also some useful features missing, like dune filter. I think Geoglyph showed, what artists are really missing (pitty it does not support WM anymore, I missing it a lot) and just taking it as inspiration and go this way would make much, much more sense. I understand, that for you as developer is GUI pain in the ass, but for me as user I didnt care about GUI at all, there was no bottleneck for me. Productivity (time spent) is what allows me to earn my bread. Making some device configuration takes minutes. It takes hours of iterations to tune it.
I hear you. Although I do think the UI and UX is very important, productivity and functionality is definitely #1. LTE will certainly look nicer, but that is a minor benefit compared to the absolutely game-changing productivity improvements it will usher in! I’ll talk about this a lot more in the next blog post. It is the thing I’m absolutely most excited about.
The short version is that the combination of a profoundly new architecture for project state and QT for UI fixes/shortens a lot of the tweak/tune cycle that consumes so much time creating worlds right now. I think you’ll love it, but let me know in the next post
Well, for me it is simple math. 5 percent of time is working in GUI and 95 pecent watching CPU render other iteration. So even if you make GUI 10x more productive, you only saved 4,5 percent of time…I really believe, that you can make grat GUI, but I am not that sure, that it could cut from that 95 percent, which are real issue. Anyway I really look forward new version
For me it is 95 Percent GUI and 5 percent CPU rendering. So it is very diffrently how we use WM
In my opinion the rewrite to the new UI/UX platform is needed. While we don’t see a huge leap in features immediately, You’re missing the message: feature development has been stagnant for years because the old platform was hard to work with.
Going to QT will make it easier for Stephen to work and develop future feature additions and enhancements.
That’s worth a lot to me!
Except for super rapid GPU computing, I don’t see much that competition can beat WM, even in its current state.
They’re all good for what they do, but WM is still at the top. Robust, and with features we only can dream of in the other two. The “planet exploration” mode to look for interesting features outside of the current work area for one, and proper tiled builds, hands down best in class.
Some feedback about the UI:
- The whole idea of that left window feels outdated and takes space. Just remove it. For instance - the small preview feels like a bad habit to ever look even - it is too misleading.
And that list thing below it - again It feels outdated idea - I already read the nodes, should be hidden/alternative by default. When working I just hide that whole useless left panel; open additional window instead (drag it to the left -so the screen takes 50% left) and set it as node window and I use the main WM window to preview the 3D terrain in the rest of the 50% screen space on the right - there I can actually zoom/see something!
This set up should be the default, just an opinion.
(+ setting the preview light should have alternative control in the big 3D preview).
- Lock Preview button. Always in use - should have better control, the “F” button is slow/inconvenient; the button on the left panel is too much to always reach (and overal outdated omg) - better make it the middle mouse button! Almost all work is done with mouse already!
- If Fav tab can have double row & maybe space separators -probably I wont need to search in the other tabs ever again.
Speaking of features to get rid of or add; a “search for” type widget would be great. Press TAB and start typing any part of the name of the node you want and it’ll show up as a suggestion.
In other words; look at how Nuke is doing this. Or Unreal Engine. It’s brilliant!
TAB -> Type “er” and you shouldn’t have to type more to get the different erosion nodes in the list, and pick the one you want.
The nodes you use most should appear at the top of the search results.
A couple updates on your guys feedback!
- I love the idea of a filterable tab quick search widget. This is a high priority add, as it aligns with a few other directions I want things to go.
- I agree that view locking is a critical feature. In LTE, there is a global view lock, but individual 3d views can override that and view a specific device, useful for seeing both what you’re working on, and the final output at the same time when you have >1 3D views showing.
- I think you’ll find the leftside view possibly more helpful now as it no longer shows misleading output - it shows the same results as the main 3D view. As far as getting rid of the leftbar, I can see arguments for both ways - screen space is at a premium and you don’t always need it. I think having it be resizable and also surfacing the show/hide button to the main toolbar should go a long way towards addressing this. Eventually it will be possible to float/move/hide anything you don’t want there.
I like the idea of having view lock be an easy click. I’ve been experimenting with this, it is very helpful. I’m looking at either ctrl or alt click right now for lock the view to that device. the “F” key will remain for compatibility. This will work in both the device list and the workview.