Loading time performance regression?

Regarding the repro on Mac + Chrome, I think @Anupam_Das posted a bug on Chrome bug list

@Anupam_Das: Do you mind sharing the bug link? We can probably all vote it up

@Leon: We isolated it to a webgl2 bug on Chrome for macOS

you can workaround it for now by disabling webgl2 on this specific user agent

1267322 - [WebGL 2] Seeing stalls in getUniformBlockIndex() despite waiting for program readiness with COMPLETION_STATUS_KHR - chromium

Please upvote!


@Leon we have a workaround in place that should fix EVERYTHING (and you can keep using webgl2):
Babylon.js Playground (babylonjs.com)

Please tell me what are your numbers now


I was able to test on my Mac too and got about the same results as your’s @Deltakosh, about a second with Safari and about 4 seconds with Chrome after testing several runs. And with your PR it’s down to about a second or faster on Chrome. :slight_smile:

I also tested on FireFox where it’s taking about 3 seconds. After copying your same exception but for Mac and FireFox it’s down to about a second too thou…

1 Like

Anyone’s ever noticed the weird behavior of Chrome on Mac? Sometimes clearing the cache or even restarting Chrome makes the loading about 20% faster (I’m not speaking about just bjs or webGL) this is even true with the bloody Apple website :wink: May be yours noticed that the same apple website actually loads faster with edge on Windows :wink: I don’t know what these Guys are doing to sell us a rig that is twice the price of its windows counterpart and leaving such parts (as WebGL) unattended for years, sometimes for decades. I have been a mac user for (well, since mac :wink: and I still like Mac and work on Mac (but then, I play on Windows :wink: I simply cannot endorse the way these Guys are doing what is supposed to be their job :unamused:

Hell Yeah! :champagne:
Now it feels fast again :slight_smile:

Thanks for the awesome work!

iMac Pro
MacOS Monterey
Chrome 101

Babylon.js v5.5.5 - WebGL2 - Parallel shader compilation
time1: 0.5
time2: 1200.8000000715256
time3: 1369.3000000715256
Babylon.js v5.5.5 - WebGL2 - Parallel shader compilation
time1: 0.5
time2: 1177.6000000238419
time3: 1375.1000000238419
Babylon.js v5.5.5 - WebGL2 - Parallel shader compilation
time1: 0.5
time2: 1276.6999999284744
time3: 1459.8999999761581

Is there going to be a patch release for this today?

cc @RaananW

i’ll release 5.5.6 in an hour or two.


It was a hell of a ride :slight_smile: Thanks for your patience :smiley:


I’m sorry to have to say this, but the problem does not seem to be fixed in 5.5.6

Here it works

Here i simply removed the snapshot param which goes to version 5.5.6

and again I have load times of ~8sec

Was the fix properly applied?

These are the times I am getting -

But the PR you referenced (i.e. the PR the snapshot you have used ) is not yet merged, so it’s not applied in 5.5.6:

Still test disable meshes because gltf loaders them them on later by deltakosh · Pull Request #12487 · BabylonJS/Babylon.js (github.com)

I’ll merge this soon. Is that the fix for the issue?

Yes this will fix it.
I thought this was going out in version 5.5.6 :sweat_smile:

1 Like

I’ll see when we repeat a new patch. Probably will do that later today

My bad I thought I merged it but I did not (I’m getting old apparently)

1 Like

@RaananW I really need to get a patch release out with the fix,
do you think you are going to be able to get one out today?

1 Like

Sure, 5.6.0 will be released in an hour (and a bit :blush:)

1 Like

@Deltakosh @RaananW sorry Im wayyyy late to the party. I was out all last week and wasnt able to follow up. We will test on 5.6.0, the PG example works so I think we are good. Thank you all.

1 Like