Issues Related to Performance seen during lighthouse audit

[tabshaikh]

Lighthouse performance audit for different pages: Home Page: https://openlibrary.org/ Author Page: https://openlibrary.org/authors/OL18319A/Mark_Twain Edition Page: https://openlibrary.org/books/OL27459710M/Holding_Up_the_Universe

	Desktop	Mobile
Home page	58	62
Author Page	80	87
Edition Page	52	59

Homepage Desktop Performance Summary:

Homepage Mobile Performance Summary:

Author Page Desktop Performance Summary:

Author Page Mobile Performance Summary:

Edition Page Desktop Performance Summary:

Edition Page Mobile Performance Summary:

Issues Common to all pages:

The Key issues needed to improve Openlibary performance are below(in order of priority) ::

• [ ] Properly size images #3223
• [ ] Serve static assets with an efficient cache policy
• [ ] Twitter preview blocked for IA hosted images #1841
• [ ] Preload key requests #3221
• [ ] Serve images in next-gen formats #3222
• [ ] Preconnect to required origins #3224
• [ ] Avoid flash of invisible text #3226
• [ ] Ensure text remains visible during webfont load
• [ ] Minimize main-thread work
• [ ] Avoid an excessive DOM size
• [ ] Reduce JavaScript execution time
• [ ] Avoid chaining critical requests
• [ ] Keep request counts low and transfer sizes small

Audit Report can be found here -> https://archive.org/details/ol-audit desktop-author.pdf author-mobile.pdf edition-mobile.pdf edition-desktop.pdf

Stakeholders

@tabshaikh

[tfmorris]

I'm getting very different times, perhaps because of network speed differences. We need to find a way to test this reproducibly so that we're not chasing our tails.

We should use HTTPS for all URLs to avoid the initial redirect. The home page is critical, but other page types are important too, so we should test Author, Work, & Edition.

Here are mobile and desktop scores for various URLs at my home office: 48 73 Work https://openlibrary.org/works/OL85892W/Dracula 95 99 Author https://openlibrary.org/authors/OL31727A/Bram_Stoker 65 98 Homepage https://openlibrary.org

As one would expect with the differing scores, the order and magnitude of the performance improvement suggestions change as well. Below are the ones for the home page:

Mobile - Score 61

Opportunité - Estimation des économies Différez le chargement des images hors écran - 3 s Préchargez les demandes clés - 1,17 s Diffusez des images aux formats nouvelle génération - 0,15 s

Desktop - Score 91

Opportunité - Estimation des économies Différez le chargement des images hors écran - 0,36 s Dimensionnez correctement les images - 0,24 s Préchargez les demandes clés - 0,13 s

[tabshaikh]

@tfmorris Yes maybe due to different locations and network speeds the audit would show different results, these issues I created for keeping homepage in mind but I would audit the author and editions page too and update all the issues for author and edition's page too :) Thanks for your feedback

[github-actions[bot]]

Assignees removed automatically after 14 days.

[mekarpeles]

https://pagespeed.web.dev/analysis/https-openlibrary-org/415u7hl5ah?form_factor=mobile