CSS question...or disappointment

[Marc Stridgen]

21 minutes ago, Day_ said:

Right up there with the bloke that’s spent 25 years looking for Nessy on Loch Ness

Everyones gotta have a hobby I guess LOL

[Charles]

We were recently comparing our page speed to a competitor that people always go "Oh they have an amazing page speed score!"

Turns out, they just hide all their JavaScript from Google. Totally a cheat and totally against Google's rules but the almighty page speed score was high!

[SeNioR-]

1 hour ago, Charles said:

to a competitor

could you please some example?

[Askancy]

1 hour ago, SeNioR- said:

could you please some example?

I think naming names is unethical. IP.Board currently has only one licensed competitor and one open source...

However, you never understood this craze of constantly using tools like gtmetric pagespeed insight etc...

They are useful tools but in my opinion it is fair to consider them for what they are....

[sadams101]

On 9/21/2022 at 6:25 AM, Rikki said:

Hey Adlago,

I don't work at IPS (despite that cute cartoon avatar I have), so I'm able to tell you that you're doing it wrong. You should never have been editing the default CSS files, and it's unfortunate that the product allowed it all this time.

I assure you, you can do everything you did before. It'll be even better than before, actually, because it won't break on every upgrade.

I can assure you that @Adlago knows how and why there is a need to edit IPB's core site files, and that we know how to upgrade after doing so.

The reason we have to edit them is because of IPS' ongoing site speed issues, which they are always at least 2 years behind in addressing, for example this thread:

 

IPB is mistaken if it believes that CDN's will help with site speed. After going back and forth from a CDN I found they do not help. Most of the speed issues with IPB software are inherent in the IPB application, and as @Adlago has pointed out, have to do with how IPB's JS and CSS files written and served.

This is why some people need to tinker with these files, not for fun, but for speed...no CDN (if we could achieve this without editing the core files we would):

 

 

[Randy Calvert]

Again… with CSS… you don’t NEED to edit core CSS to improve SEO.  Just because you know how to change something does not mean you’re doing it in the right spot or doing it in a way that is sustainable. 

[sadams101]

You do not even know anything about what we are doing.

[Randy Calvert]

20 minutes ago, sadams101 said:

You do not even know anything about what we are doing.

How about this… what are you trying to do in the core CSS files that you can’t do in the custom.css?  Let’s see if it’s actually a true problem or if there is a reasonable alternative. 

Edited October 7, 2022 by Randy Calvert

[Adlago]

21 hours ago, Randy Calvert said:

How about this… what are you trying to do in the core CSS files that you can’t do in the custom.css?  Let’s see if it’s actually a true problem or if there is a reasonable alternative. 

And you think please. This topic does not address changes to CSS rules. What this thread is about is being able to other loading  this CSS...
Think for yourself - how much time does a server use to process php requests, and the CSS minimizer that IPS has implemented (this is a particularly good hit) for all possible  external CSS? Even the fact that all external CSS are downloaded in parallel does not change the delay in page rendering. Yes, when a site has no special claims for this - it is irrelevant.
But when a site relies on advertising revenue - every millisecond of delay is a lost cent.
The point of my topic is that the CSS that is used by a site (and this in volumetric weight is about 15% of the total volume of available external CSS) should be loaded as a priority.
This is possible only if you move used CSS rules for each site (for each site it is individual - according to the needs of the community) in one common external CSS, which you understand yourself will load quickly... All other CSS rules in files, in which they are created can be moved before the close body tag... They are not important for site rendering - and can load with low priority...
Site speed, which is strangely underestimated - costs money, and any site using ads knows this very well...

[Day_]

1 hour ago, Adlago said:

But when a site relies on advertising revenue - every millisecond of delay is a lost cent.

Go on, expand on this please as I'm struggling to work this calculation out.

If you believe that by shaving a millisecond off a page loading will earn me a cent, I would like to offer you full time employment.

[Randy Calvert]

37 minutes ago, Day_ said:

Go on, expand on this please as I'm struggling to work this calculation out.

If you believe that by shaving a millisecond off a page loading will earn me a cent, I would like to offer you full time employment.

Typically a single millisecond won’t matter, but the general wisdom is that performance matters… that if a website takes more than three seconds to load, there is a percentage that is lost for each additional second.  Huge sites like Amazon, Facebook, Walmart can convert this even down to milliseconds in terms of how it impacts conversations.

https://www.kickfire.com/blog/hold-up-wait-a-millisecond-how-response-time-impacts-your-website?hs_amp=true

However most websites powered by IPB are not to be operating at that scale. In addition, most websites powered by IPB also are not primarily e-commerce applications. (Many may have products/subscriptions for purchase, but generally if e-commerce was the primary goal… a site would be using applications designed for that regard.

Performance does matter and is relevant to a point. However there reaches a point of diminishing returns. A human won’t distinguish the difference between 20ms and 30ms. They absolutely would tell the difference between 300ms and 3000ms.  

In terms of SEO, again… important, but according to Google content is KING. A site with good content would do better than a crappy site with no content but that was 30ms faster. 

Ultimately there reaches a point where you are not really “gaining” anything except bragging rights. And there’s nothing wrong with that.  However in reality in the “real world” you don’t have to have the fastest speed on the block to do well in Google. (It can’t suck, but it does not have to be the absolute fastest either.)

What’s funny for me is that I’ve typically seen third party ad networks (especially Adsense) be among the worst offenders in terms of those core web vital stats people focus so heavily on.  So while having good performance matters, it’s not the end all/be all. (Which I THINK is one of the ideas you’re saying!)  🙂

[PatrickRQ]

On 9/15/2022 at 10:05 PM, Stuart Silvester said:

You should be adding any custom CSS to the custom.css file instead of editing the CSS distributed by other apps. You'll only give yourself headaches upgrading them in future.

The customized css from before update that made .css'es hidden made it hidden and customized, without ability to revert it to default 😄

[Matt]

Core Web Vitals are the critical checks to pass, and as far as I can see there's no difference if we're talking about SEO
 

We agree that we can do better with performance and have plans to do so, but for SEO, I don't see much improvement between our version and yours.

[sadams101]

So the Google score means nothing, right?

If so, they would simply show an Pass/Fail, and not the score. 

Why would Google create this complex tool with how to fix each item that is causing slowness? 

You can continue believing that the Google score will not affect your rankings, but Google has already said years ago that site speed is a ranking factor (of course bad scores will affect you), so I will continue trying to fix IPB's inherent speed issues.

Edited October 16, 2022 by sadams101

[Randy Calvert]

3 minutes ago, sadams101 said:

Right, the Google score means nothing, right?

Rather than looking at them as there is a difference between 85 and 86… look at them more as a pass/fail. 

They help you figure out where you can improve. If you have a HORRIBLE score, yes you will be penalized. But for example in the two screenshots above, it won’t change the ranking. 

[Randy Calvert]

By the way...  here's a good read:

https://www.perficient.com/insights/research-hub/impact-of-core-web-vitals-on-ranking

It matters for SEO/ranking, but it's not as big as some people make it out to be.  

Quote

Our data confirms what Google has been telling us: CWV scores are not a large ranking factor. Content relevance and quality remain the most important ranking factors, as they should.

So...  if we say content is worth like 80% and performance is worth like 20%...  personally I would be spending my time on where the most weight is at.  🙂

[Adlago]

8 minutes ago, Randy Calvert said:

So...  if we say content is worth like 80% and performance is worth like 20%...  personally I would be spending my time on where the most weight is at.

Where do you find an estimate that the content is worth about 80%?
Content rating also depends on users - slow loading speed content is ignored by many users...

PS. Also Please note, above images from  @Mattare from different site...

 

[Randy Calvert]

1 hour ago, Adlago said:

Where do you find an estimate that the content is worth about 80%?
Content rating also depends on users - slow loading speed content is ignored by many users...

PS. Also Please note, above images from  @Mattare from different site...

 

I literally made it up.  I was using it as an example to illustrate the message that you spend time on the things that have more weighting/importance.  Google has clearly said that content matters more than speed...  but not to totally ignore speed either.

 

[Meddysong]

6 hours ago, Adlago said:

PS. Also Please note, above images from  @Mattare from different site...

Of course one of them is. @sadams101 mentioned in this thread that he has made changes to his site to address the "speed issues [...] inherent in the [Forums app]", posting a screenshot of his results. Matt reused one of those screenshots to contrast with the SEO results for this site, illustrating that the gains compared to an unaltered version of the software appear to be marginal. 

[Adlago]

 

7 hours ago, Meddysong said:

Of course one of them is. @sadams101 mentioned in this thread that he has made changes to his site to address the "speed issues [...] inherent in the [Forums app]", posting a screenshot of his results. Matt reused one of those screenshots to contrast with the SEO results for this site, illustrating that the gains compared to an unaltered version of the software appear to be marginal. 

This comparison is incorrect, that's why I wrote it.
Below is an objective comparison from a test now from my friend's site @sadams101
1. IPS default theme - without ad blocks included
- there are no reported web vitals indicators in the test because it does not pass the test successfully.
Site load test

 

2. Test with a used theme - after numerous changes in the CSS loading method, without changing the content CSS rules, All advertising blocks included - more than 6-7 objects in a page.

And of course - web vitals test

Do you think it makes sense?
Each site has its own specificity in displaying its content - (how much value content has - that's another topic). That's why every site uses css rules of different importance. In order to speed up the loading of all resources, it is essential to speed up the loading of the used rules, and postpone the unused css rules...
This is impossible only using custom.css. This is also impossible using designer mode.
And that is the point of this topic of mine.
I understand the desire to improve the entire process of IPS development - but this process is slow and requires a lot of analysis and a lot of work.
Therefore, it is better to IPS restore access to all available css files in ACP.
When the IPS offers a successfully working project for the idea CSS to CDN - let this project go through online tests - who wishes to participate - and then go in for release...
I think so.

Edited October 17, 2022 by Adlago

[sadams101]

20 hours ago, Randy Calvert said:

By the way...  here's a good read:

https://www.perficient.com/insights/research-hub/impact-of-core-web-vitals-on-ranking

It matters for SEO/ranking, but it's not as big as some people make it out to be.  

So...  if we say content is worth like 80% and performance is worth like 20%...  personally I would be spending my time on where the most weight is at.  🙂

The conclusions of your article seem to support focusing on improving site speed:

 

Quote

Regardless of whether rank is significantly impacted, improving CWV will likely be beneficial for your business.

Summary

Our data confirms what Google has been telling us: CWV scores are not a large ranking factor. Content relevance and quality remain the most important ranking factors, as they should.

However, this should still matter to anyone who publishes or operates a website for reasons that go beyond the SEO benefits. Sites with faster pages get more user engagement and higher conversion rates.

From Google's perspective, the reason they care is also clear. Google wants to serve the best possible results to its users. Offering a quality service keeps usage high and drives ad revenue.

While much of its algorithm seeks to identify the site with the best content to put in the SERPs to meet the user needs related to a given query, this has little to no value if the users have a bad experience on the site containing that content.

Expect that Google's investment in CWV and Page Experience will continue to grow over time. You may see increased strength of some signals and new signals added into the mix. The concept of Page Experience and CWV scores within that will continue to be part of the landscape and will continue to grow.

 

[Randy Calvert]

I've not said you should not improve the user experience at all.  What I'm saying to be abundantly clear...

Spend the MAJORITY of your time on good content.  Spend SOME time making your site perform better.  

Again...  content is king

Too many people here focus on the pennies instead of the dollars.  

Edited October 17, 2022 by Randy Calvert

[sadams101]

I also focus on good content. Content does not fall under the purview of IPB's software engineers, which is who this thread is addressed to. This thread is focused on IPB taking away the ability to easily modify the CSS files, which in turn makes it very difficult for those who do actually care about site speed to be able to do anything to address it.

[z929669]

On 7/8/2022 at 2:45 AM, opentype said:

/snip

So far, from the 27,436 active clients you are the ONLY ONE who mentioned a clear need to make drastic changes to the core CSS files. So it is questionable to suggest that “many” will turn away, when you have no foundation whatsoever to say that. You are likely just claiming it to give your position more weight. 

I will mention it then. The IPS CSS has so many redundancies it isn't even funny. This change didn't impact me, since I do use custom.css, but in maintaining that implementation since IPB 2.x, it's been clear that the CSS is very hacky with many different people having worked on it in silos over the years. This is demonstrated by the frequent use of !important in the CSS definitions to override the otherwise ridiculously convoluted selectors that even the devs don't want to touch.

Now with talk of improving the CSS implementation, it's almost a certainty that my customizations will be broken in the future ... these past few 'upgrades' have come with many major downsides (e.g., the Topic Feed uselessness ... don't even get me started).

Such is the way of SAAS software: optimize the software for the devs and the non-technical customers whilst alienating the technical/power customers ... dumbing down the world.

[Matt]

I have mentioned multiple times that we are going to create a new CSS and JS framework that will be lighter. It is not a fast undertaking though and needs to be planned in stages.