Page Speed Improvement

[Wise]

Description
After the last commits, the previews of the master and staging branches now as fast as possible. The page size has decreased to 1/10. First Contentful Paint and Largest Contentful Paint time values were decrease 5 times. Regarding the current test details from Frankfurt / Germany;

Tests Results

• According to the test of the https://pagespeed.web.dev/ site, there are no additional suggestions for performance. The site get 100/100 for both mobile and desktop.

• According to the test of https://tools.pingdom.com/ the site get 99/100 score. (Only difference of test result due to the first access time.)

• According to the test of https://gtmetrix.com/ the site get 97-99/100 score. But I can't test efficiently because it doesn't allow me to test from Frankfurt/Germany. Also, there are offers a number of recommendations that can be applied.

• According to the test of https://www.webpagetest.org/ we get A/A+ results in all titles without giving numbers. But if we want to go further, he makes a suggestion that I will mention later.

First Byte	FCP	Time	TBT	Requests	Total Bytes
.112S	.244S	.340S	≥ .000S	11	87 KB

Based on the average of 4 webpage tests, we get the above results. And that's means we are faster than 99% of existing web pages. After this short brief, I will list the improvements we can implement.

Upcoming Features and checklist

1. Image files will use native lazy-loading

• Lazy-loading of single image
• Lazy-loading of image sets/src-sets
• Multiple image format support
• [Canceled] Serving multiple images as a single image in tables

2. Pages will get more responsive

• Support for traditional mobile phones
• 2x-4x Retina-ready pages for desktops (Also related to the first topic)
• Sub-menu support (Changed)

3. Favicon file will load simultaneously with the images

• Does it cause an issue to render favicon file with photos?
• Does change the order of favicon file cause a performance issue?

[Thilo Molitor]

Does this need any interaction from me?

[Wise]

@tmolitor-stud-tu , the first part is actually complete. I'm undecided whether to open the pull request before the other parts are completed. By the way, the maintainers don't want the dropdown menu feature to be used. He stated this both in the issue answers on his own page and while designing the code (I think), he acted in this direction. Shall we cancel this feature?