> Mozilla invested a lot into R&D in XR in the late 2010s, and in late 2018 they released an experimental browser called Firefox Reality. It was a great entry into the XR field, helping establish what a browser in these devices really looks like, and figure out the unique challenges. Today we’re excited to take up this experiment and continue this work as a complete project.
Well, I hope they'll contribute to Gecko to make it faster at rendering 3D with WebGL, or even help adding WebGPU support.
Because for now, perfs aint good compared to chromium based browsers... Firefox is often way slower when using heavy 3D apps.
You have to shuffle heaps of data between Javascript and the gpu, I’ve found Firefox’s js engine to be slower at doing that. Firefox also seems to call the garbage collector much more often, and do it’s business a bit slower.
Normally you don’t notice, but trying to keep things at 60fps it shows up rather quickly.
If you have a nice reproducible case of the garbage collection problem, please file a bug https://bugzilla.mozilla.org/enter_bug.cgi?product=Core&comp... and I'll take a look. I've seen some similar issues, but I don't have anything good to work against right now. (The first thing I'll do is grab a profile with https://profiler.firefox.com and that'll immediately tell basically what's going on. You could grab one too, and that'll be useful even if there's not an easy way to give me a reproduction.)
There's also a very recently landed change that might help, if the problem happens to be excessively long minor GCs (which is what I've seen with similar sites): https://bugzilla.mozilla.org/show_bug.cgi?id=1751162
Shuffling data to the GPU could be an interesting case too, but that's not handled by the JS engine itself so I'm not as familiar with how that works. It sounds like it might be worth a bug if there isn't one already (and feel free to file; we'll triage and determine if it's a duplicate.)
Ok so what's slower is the JS engine which is a more reasonable claim. My experience is it depends on the types of work your doing but in general, indeed, chrome seems faster. But Firefox has improved a lot with quantum.
no GPU drivers do not support webGL natively.. Although gecko use Angle from google, 3D rendering is a multifactorial problems that can be optimized from many aspects of a web browser (display lists vs immediate rendering, various caching systems, direct compositing, a lot of low level optimizations, reducing memcpys, etc)
Last time I checked (which was admittedly 2-3 years ago), Chrome had considerable lag with VR whereas Firefox managed much better. Might depend on the GPU, of course.
This is really interesting. Last night I was looking at FireFox Reality's repo and was pleasantly surprised to see [some newer commits][1]. Now today I see that Wolvic is [a fork of FireFox Reality][2] and that the commits were contributed from that project. I'm happy to see a private company fund new work on this front.
> There is, of course, Firefox. Brave has a lion as a mascot, there are Puma and Dolphin browsers, and many smaller and historical browsers and projects within the browser space are named after animals
But the wolf name is already taken and on a browser based on Firefox too https://librewolf.net/
Note that I am still mad that librewolf was not named Windwolf or Waterwolf instead.
I think Waterwolf would get confused with Waterfox. Only a matter of time before wind and earth also get claimed by Firefox forks, though. New ones seem to pop up every year or so.
Very excited to see this, my ongoing prediction is that the metaverse will be WebXR based. No 30% cut from walled gardens, and sites can evolve from static, 2D flat experiences to immersive ones. Of course, apps will work day one on any headset in a platform-agnostic manner. And for the spark, just wait until the Apple headset ships with WebXR and WebGPU supported by default....
Also, my team are currently working on bringing Unreal Engine to WebXR.
A VR browser is however very useful eitherfor browsing 3d content or having access to content that is useful for other VR apps. Especially for standalone headsets where you can't just switch to your pc desktop.
I've used passthru on the Quest 2 with a browser window hanging in my living room. There absolutely is an allure to having a tv as big as my wall playing a youtube video while I'm folding a mountain of laundry. Working, serious work it's a bit of a stretch at this moment yeah
I haven't used immersed yet, but the whole "I can see my environment and my hands and type on a keyboard I can see with N floating windows" thing is neat and will only get nicer as the resolution improves
Yep, it's grainy / pixelated but it's not like I need to read fine print to do that sort of task. I did it as a novelty, but idk I could see myself doing that sort of thing again.
I build a WebXR app at work. It's a virtual environment for taking foreign language classes. You travel around the world and talk about what you see in-language.
Those... aren't the only two options. I use Three.js.
I never really saw what the hype was about React. It seems like most people who got really into it never really had much experience building componentized code before, so what they think of as being the "great" thing about React doesn't really need React at all.
Virtual DOM is overhead, overhead I can't really afford. I have to hit 120FPS on the Quest 2 with an immediate-mode renderer, when React's V-DOM was designed for a 60FPS retained-mode environment. It's just... too much of a leaky abstraction to deal with when I have about 8 milliseconds to do absolutely everything. And yes, I hit 120FPS.
Honestly, managing 3D objects is one of the smallest parts of my job. Some days, it can be all consuming, but other times, I can go weeks without touching the graphics. The biggest concern I have is with data management and efficient transmission and decoding of assets.
I continue to be astonished (and dismayed) that people use a browser on the desktop to do serious work. Really, the browser provides a horrible user experience, no matter where you use it.
I miss native apps and the engineering that went into them, as compared to the slapdashery of web apps.
Serious work can be as little as searching for something and reading content. No reason why a web browser can not do this - in fact it was built for this purpose.
The name is a little confusing, and makes it clear they are not native English speakers. It's not a bad name, but if someone says to me, "Hey, did you try Wolvic yet?" I'll spell it "Wolfic" or "Wolfik" or "Wolfick", definitely not "Wolvic". Confusing.
Whereas to my UK eyes it's instantly reminiscent of Volvic[0] mineral water, which is definitely a V sound and doesn't read or sound strange at all (see also: pelvic, civic, Slavic). It wouldn't occur to me that there might be an F sound.
Interesting observation. Fwiw, I am definitely very much a native english speaker (in fact, I don't speak any other (non-programming) language very well), and I proposed the name. As the post says at the end - it doesn't hurt that the domains are available if you don't use a "real" world. I guess we'll see if this happens.
Apologies if it's not the type of feedback you are looking for, but I strongly agree with the OP. I find the name so confusing that I have trouble even thinking about it. Quite possibly the problem is me, but if you are trying to reach an American audience, consider doing further audience testing before investing too heavily in the name.
I would pronounce the 'v' in Wolvic like in "vision" and the in your words 'f' like in "elfic". Which would make it clear Wolvic is written with a v. Would I be wrong?
(not a native English speaker)
(we have a town called Volvic here, known for its water. Pronounced like in "vision". I would pronounce Wolvic the same way, except the first letter would be pronounced like in "well")
Also not a native speaker, I'd use the same pronunciation as you ("v" in "wolvic" is the same as in "elven", "f" in "wolfic"/"wolfish" is the same as in "elfic"/"elfish"). But the difference is quite subtle, I see it can be hard to tell apart in casual conversation when people pay less attention to enunciation.
Wolvic sounds more natural to me than your other suggestions. If someone verbally told me about the project, my first guess would be to spell it as Wolvic.
I can see possibly someone non native spelling it as Wulvic or Woolvic (if they heard the name in a conversation). I'm sure some may misspell it with an f also but if the project grows, it won't matter.
I use FFR on my Oculus Quest 2. IME, text quality is more constrained by the res of the device, and I personally find it painful to read on my Quest - whether in FFR or Immersed - for extended periods. Others don't seem to have the issue, and people on HN have posted about working full-time in VR, so YMMV.
The question is about font rendering. Do they rasterize text into a texture, then draw the texture on a 3D plane in VR introducing artifacts at the second step, or do they rasterize text directly in 3D space using the view matrices?
It poisons (and dumbs down) the ecosystem to post generic dismissals like this, so please don't. Of course, most projects end up failing - but that's a bad reason to direct this sort of barb at specific ones.
If you have a substantive point, make it thoughtfully; if not, please don't comment until you do.
> Mozilla invested a lot into R&D in XR in the late 2010s, and in late 2018 they released an experimental browser called Firefox Reality. It was a great entry into the XR field, helping establish what a browser in these devices really looks like, and figure out the unique challenges. Today we’re excited to take up this experiment and continue this work as a complete project.
> [links to https://blog.mozilla.org/mozilla/update-on-firefox-reality/]
A Gecko-based browser would be refreshing!
Igalia really seems to do interesting stuff.