Any sort of roadmap for Xenforo 2.3 ?

[Chris D]

The current editor is a massive problem, but I think that it highlights the problems that the XF team would have if they were to take this in-house. The more features you add to the editor the more problematic it becomes making it work reliably with different browsers, different releases of browsers and different operating systems. I can see why they would want to use a 3rd party product for this so they can concentrate on the core.

Froala has caused problems for many of our users and been unpopular with them since XF started using it though and needs replaced even if it is with something simpler and less fully featured.

Yeah absolutely not going to write our own. But we’re being a lot more careful in choosing our next editor. I have been working on this while the rest of 2.3 is finished, amongst other things.

Not likely to be in 2.3 but we acknowledge it’s an important project.

 

[Frode789]

Some of the earliest work on 2.3 was towards improving Lighthouse metrics, though we may need to revisit this again with INP in mind.

Bearing in mind this is only "lab data" from relatively barebones local installations, here is a mobile performance score comparison of the thread view between 2.2 and 2.3 on identical server configurations:

Spoiler

View attachment 288723 View attachment 288724

Great work! Love to see a better focus on performance Improvements.

Feel free to share more details/examples about performance improvements if you have

 

[Alpha1]

Bearing in mind this is only "lab data" from relatively barebones local installations

Please do tests with some of the most popular addons installed. Especially addons with major content types. With and without CloudFlare optimizations.

Is the XF3.0 style going to focus on further optimization as well as SEO? I ask because the way XF currently works is many nested DIVS, heavy CSS, JS and queries all loading things not needed for the current page. It seems to me that a new approach to styling is required to greatly improve that.
For WordPress Optimization we use a plugin to unload unused styles and scripts:

Asset CleanUp Pro: Performance WordPress Plugin - Gabe Livan

There are often times when you are using a theme and a number of plugins which are enabled and run on the same page. However, you don’t need to use all of them and to improve the speed of your website and make the HTML source code cleaner, it’s better to prevent those styles and scripts from...

www.gabelivan.com

While WordPress is a different beast, preventing unused things from loading on a page would be really cool for XF performance. Especially when many addons are used on XF.

 

[The Dark Wizard]

Please do tests with some of the most popular addons installed. Especially addons with major content types. With and without CloudFlare optimizations.

Is the XF3.0 style going to focus on further optimization as well as SEO? I ask because the way XF currently works is many nested DIVS, heavy CSS, JS and queries all loading things not needed for the current page. It seems to me that a new approach to styling is required to greatly improve that.
For WordPress Optimization we use a plugin to unload unused styles and scripts:

Asset CleanUp Pro: Performance WordPress Plugin - Gabe Livan

There are often times when you are using a theme and a number of plugins which are enabled and run on the same page. However, you don’t need to use all of them and to improve the speed of your website and make the HTML source code cleaner, it’s better to prevent those styles and scripts from...

www.gabelivan.com

While WordPress is a different beast, preventing unused things from loading on a page would be really cool for XF performance. Especially when many addons are used on XF.

Theres no way xF Devs can realistically test that many addons/combination of addons right? They would have to first request access to said addons (many that aren't free and sold on other sites) and it would require dev work to create an automated performance testing suite.

 

[JoyFreak]

Theres no way xF Devs can realistically test that many addons/combination of addons right? They would have to first request access to said addons (many that aren't free and sold on other sites) and it would require dev work to create an automated performance testing suite.

Also that the dev team are more concerned about the core performance rather than third party add-ons. It’s up to the add-on developer to optimise their own add-ons for better performance. The dev team can’t always monitor these things, it’s too much work.

 

[Jeremy P]

Most of the improvements were global (ie. they impacted every page) so add-on pages should benefit all the same. If an add-on does something in particular that doesn't perform well, there's only so much we can do to mitigate that.

We're optimizing everything we are reasonably able to, but frankly XF Lighthouse scores are already in pretty good shape given the dynamic nature of the platform. A lot of our CSS bloat is due to FontAwesome (which, as has been noted elsewhere, will be remedied in 2.3). Otherwise the "unused CSS" and "unused JS" statistics only paint part of the picture by virtue of being an isolated data point. The unused bits shrink quite a bit when navigating between a handful of pages, and splitting them too aggressively can hurt other metrics. Still, we're looking to squeeze out improvements whereever we can.

 

[tajhay]

Appreciate the drastic improvement in communications.

 

[Starbucks]

Appreciate the drastic improvement in communications.

I agree.

 

[Mr Lucky]

Please do tests with some of the most popular addons installed. Especially addons with major content types. With and without CloudFlare optimizations.

I don't think it's practical, as others have said it should be the other way round. Just get the core solid.

Besides, if I understand this correctly:

Once 2.3 is out, we'll be getting in touch with major XenForo style producers to help them get ready for the 3.0 style,

then devs may either be getting a preview (a good idea IMO) or else some useful info in order to be ready.

 

[Alpha1]

To clarify: I'm not asking the XF team to optimize 3rd party addons. I'm asking the XF team to optimize their platform better for the use of addons. Including official addons.
XF and XFCloud are intended as a platform to install addons on. Therefore it makes sense to evaluate what performance issues you encounter when you install a bunch of well coded addons on the platform. And which of those performance issues can be improved upon by how the platform works.

 

[Jeremy P]

We already minify CSS globally, and provide tools for add-ons to minify and bundle JS. If you install a bunch of add-ons which require a bunch of CSS or JS to be loaded in order to function and don't split them up where sensible, we can't do much about that. There's some room for improvement in the page markup itself, but we're limited in how much we can change there in 2.x since we want to avoid breaking backwards compatibility as much as possible.

 

[Ivancas]

but rest assured we are working hard behind the scenes

Kier, you definitely need to improve the lack of communication in the future.

 

[lucien_lies]

Appreciate the drastic improvement in communications.

totally agree

 

[Alpha1]

If you install a bunch of add-ons which require a bunch of CSS or JS to be loaded in order to function and don't split them up where sensible, we can't do much about that.

I would love to be able to toggle off the loading of the CSS of a specific addon on a specific page. And the same for queries.
When I load a page in debug mode and read through the queries then I see many addon queries that should not be loaded on that page. 99% of admin will never review this and are therefore unaware. Let alone report any issues to addon developers. But it has a significant impact on page performance and therefore on Google ranking. Especially if there are slow queries or many unnecessary queries.

Something that would be great to help with addon performance in general is to add a simple analysis tool that exposes such issues, so admins are made aware of slow or high number of addon queries above a certain threshold. So that the analysis shows that:
Addon X adds 400ms and 90 queries to Thread pages.
And a link to the addon support page next to it.
Then admins can report it to the developer. Developers can improve upon it. Admins can also evaluate if the performance cost of an addon is worth the benefit.

 

[FTL]

@Chris D will you be releasing more official add-ons in the future?

 

[mattrogowski]

When I load a page in debug mode and read through the queries then I see many addon queries that should not be loaded on that page. 99% of admin will never review this and are therefore unaware. Let alone report any issues to addon developers. But it has a significant impact on page performance and therefore on Google ranking. Especially if there are slow queries or many unnecessary queries.

That's doesn't really make sense - if they're being run then it's because the addon is running them, they wouldn't be there unnecessarily. I'm not sure how you'd come to the conclusion that they "should not be loaded on that page"

If an addon is coded very badly and is running queries on every page instead of specific pages or actions then it would be specifically coded to do that, it won't be practical for XF to know whether or not a query "should" be there or not.

 

[Suzanne O]

Personally i wouldn't go with so many add ons because when major releases come out you have to disable them all the time.
where as if it were all widgets it would be fine.

 

[MySiteGuy]

I would love to be able to toggle off the loading of the CSS of a specific addon on a specific page. And the same for queries.
When I load a page in debug mode and read through the queries then I see many addon queries that should not be loaded on that page. 99% of admin will never review this and are therefore unaware. Let alone report any issues to addon developers. But it has a significant impact on page performance and therefore on Google ranking. Especially if there are slow queries or many unnecessary queries.

Something that would be great to help with addon performance in general is to add a simple analysis tool that exposes such issues, so admins are made aware of slow or high number of addon queries above a certain threshold. So that the analysis shows that:
Addon X adds 400ms and 90 queries to Thread pages.
And a link to the addon support page next to it.
Then admins can report it to the developer. Developers can improve upon it. Admins can also evaluate if the performance cost of an addon is worth the benefit.

Debug mode will already tell you this. Toggle the addon on/off and you can see the query count change, as well as load speed.

 

[FTL]

Personally i wouldn't go with so many add ons because when major releases come out you have to disable them all the time.
where as if it were all widgets it would be fine.

That's indeed the problem. I've had that a lot with Firefox, for example. Most of my add-ons for XF however, are from @AndyB whom I'm confident will keep them up to date.

 

[Suzanne O]

That's indeed the problem. I've had that a lot with Firefox, for example. Most of my add-ons for XF however, are from @AndyB whom I'm confident will keep them up to date.

I am talking about the xf third party addons.
if i get the tapatalk addon it seems to be ok.
If i used one from @Ozzy47 i'd have to disable it.
if it was the ones i bought from xf then they'd be fine.