After updating to beta 8, there is a decrease in performance. What I mean is:
- on December 5, PSI made a useful adjustment to its test platform, which eliminated the large TBT times.

Dec 5, 2024
The CPU throttling factor for PageSpeed Insights has been adjusted to account for the low CPU performance benchmarks typical in PageSpeed Insights production environments. In general, this should result in increased lab data TBT scores for sites on mobile. Field data and desktop scores should not be affected. 

On December 7th, after updating to beta 7, PSI tests showed perfect performance, which I wrote about:

 

Now, after updating to beta 8 - mobile performance is already severely reduced, (65-75 points) without me having made any changes to my page.

What does this decrease mean:
- First Contentful Paint - delayed by over 3 seconds
- Largest Contentful Paint - delayed by over 5 seconds
This happened with beta 8.

I've reviewed the commits and there's nothing that would change it from our end. There are no structural CSS changes, or changes on how JS/resources are loaded.

What does the report say the delay is caused by?

1 hour ago, Matt said:

What does the report say the delay is caused by?

The only difference I see in the test is what I wrote above - too much delay of FCP and LCP. And that is provided that the first byte is under 200 milliseconds and TBT is 20 ms. It is inexplicable to me because I have not changed anything regarding images in blocks and anything else between beta 7 and beta 8.

See test now

Edited December 16, 2024Dec 16 by Adlago

If you scroll down, you can check what is causing the delay to FCP/LCP.

1 hour ago, Matt said:

If you scroll down, you can check what is causing the delay to FCP/LCP.

What website do you use to do these tests?
Could you tell me?

2 minutes ago, Marco Junior said:

What website do you use to do these tests?
Could you tell me?

This

I'm getting a better mobile scoring on my vanilla installation of v5 Beta 8, with higher scoring on desktop. I have no plugins or add ons.

Edited December 16, 2024Dec 16 by Gary

@Gary Your LCP is a text block that actually has a large render delay

Which doesn't really bother me or my members. It has never taken that long for any part of my site to load, whether it be on a mobile or a desktop. Especially considering my target audience are those in Australia and New Zealand.

I much prefer focusing on content rather than squeezing every millisecond for a score.

Edited December 17, 2024Dec 17 by Gary

For me beta 8 gives a very good performance

Yeah, I'm not sure why it's flagging an image load as blocking. I'm not sure what we can do about that.

The one thing I’ve learned over the years is that Google is not always right and many times breaks or does not even follow its own guidance. 

18 minutes ago, Randy Calvert said:

The one thing I’ve learned over the years is that Google is not always right and many times breaks or does not even follow its own guidance. 

This is actually a very good point. Just one random example

 

Why do I think beta 8 has a degradation? Here are two tests - 4.7.19 and 5.0. beta 8 - on the same server, same admin (ha ha), same test platform - with a difference of a few minutes...

Why more images visible in second screenshot?

Maybe this has something to do with lazy image loading in the first test results versus the second. 

Maybe hitting a case of their browser not wanting to download as many images in parallel for the v5 case. 

Just my initial thoughts on what might be going on…

Or is there only one image for both? 

Edited December 17, 2024Dec 17 by KT Walrus

4 minutes ago, KT Walrus said:

Why more images visible in second screenshot?

Maybe this has something to do with lazy image loading in the first test results versus the second. 

Maybe hitting a case of their browser not wanting to download as many images in parallel for the v5 case. 

Just my initial thoughts on what might be going on…

In beta 7, the images were the same, and as can be seen from the test linked in my first post on this topic, it is comparable in performance to that of 4.7.19 - unlike the lowered performance for beta 8.

36 minutes ago, Adlago said:

In beta 7, the images were the same, and as can be seen from the test linked in my first post on this topic, it is comparable in performance to that of 4.7.19 - unlike the lowered performance for beta 8.

Have you looked in your browser’s dev tools to see the timeline of where the delay is coming from?

Seems to me it would be related to the number of HTTPS requests and whether the browser is doing most requests in parallel. 

I always use Cloudflare in front of my servers to optimize page loading. 

Make sure you are using HTTP3 for your site. 
 

https://http3check.net/?host=preview.invisionalpha5.com

2 hours ago, KT Walrus said:

Make sure you are using HTTP3 for your site. 

I don't like Cloudflare - and in general any configuration related to cache prioritization. In my hosting plan I have the option (and I used it for two years) to configure an Apache main server and an Nginx cache server. Due to the fact that cache management from Nginx and cache managed by the IPS platform cannot be synchronized to be maximally effective, and I have seen all sorts of antilogical states in the thousands of tests I do (this is my hobby - not a commitment) - I gave up using the Nginx server and now I work only with Apache - and this completely satisfies me.

1 hour ago, Adlago said:

I don't like Cloudflare - and in general any configuration related to cache prioritization. In my hosting plan I have the option (and I used it for two years) to configure an Apache main server and an Nginx cache server. Due to the fact that cache management from Nginx and cache managed by the IPS platform cannot be synchronized to be maximally effective, and I have seen all sorts of antilogical states in the thousands of tests I do (this is my hobby - not a commitment) - I gave up using the Nginx server and now I work only with Apache - and this completely satisfies me.

I was thinking about Cloudflare not as CDN but as an HTTPS server that really knows how to serve websites most efficiently. HTTP3 connections is only one example. Cloudflare is free to put in front of your site and doesn’t take much setup out of the box. Just flip a DNS switch and Cloudflare just works. If it doesn’t help, just turn it off and use Cloudflare’s DNS. 
 

But you can also optimize your site by using some of the many services Cloudflare includes in your account. 

If you're going to compare versions, you are better off having multiple installations.  Remember...  Google frequenly changes how it measures stuff.  Sometimes even daily.  Saying "nothing changed" ignores the fact that Google can and does change.  

This issue is related to cache blocks. I reduced the cache time to 10 minutes and after a few tests:

Then again bad tests, but after 10 minutes again very good.
Now I left it at 5 minutes - and again often very good test results.
@Matt   Please review the "Cache sidebar, header and footer blocks" management.

2 hours ago, Adlago said:

This issue is related to cache blocks. I reduced the cache time to 10 minutes

Is there an option for that in ACP or is it from the server?

8 hours ago, EliasM said:

Is there an option for that in ACP or is it from the server?