Improving Performance for WordPress Sites with Translations
Translations play a crucial role in making websites accessible to users from different regions and languages. However, recent performance analysis has revealed that translations can significantly impact server response times for WordPress sites. In light of this finding, WordPress contributors are proposing several technical solutions to enhance performance for the vast number of WordPress sites that utilize translations.
The Impact of Translations on Loading Time
According to a comprehensive performance analysis conducted earlier this year, localized WordPress sites experience loading times that can be up to 50% slower than non-localized sites. This discrepancy in loading times is influenced by various factors, including the themes and plugins used on the site. Pascal Birchler, a Google-sponsored WordPress core committer, highlighted these findings and emphasized the need for improvements.
Identifying Potential Solutions
Following extensive discussions on GitHub, the WordPress Performance team has identified six potential solutions to address the performance issues associated with translations. Each solution offers distinct advantages and disadvantages. Let’s take a closer look at these proposed solutions:
1. Solution A: Use different file format
This solution suggests storing translations in PHP files instead of the current .po and .mo files. The .mo file would serve as a fallback option. Birchler argues that loading and executing another PHP file would be a faster approach. A proof of concept for this solution is available on GitHub.
2. Solution B: Native gettext extension
This solution involves leveraging the native gettext extension, which is a popular library for handling translations. By utilizing this extension, WordPress can optimize the translation process and potentially improve performance.
3. Solution C: Cache translations
Caching translations can significantly enhance performance by reducing the need to fetch translations repeatedly. This solution proposes implementing a caching mechanism specifically designed for translations.
4. Solution D: Lazily evaluated translation calls
By lazily evaluating translation calls, WordPress can avoid unnecessary translation operations during page load. This approach aims to optimize the translation process and minimize the impact on performance.
5. Solution E: Optimize/Rewrite existing MO parser
This solution focuses on revamping the translations parser to bring significant improvements to all localized WordPress sites. Birchler suggests that combining this solution with a new PHP translation file format (Solution A) can minimize the i18n overhead. However, introducing a new format comes with its own set of risks.
6. Solution F: Splitting up translation files
This solution proposes splitting translation files into smaller, more manageable parts. By doing so, WordPress can optimize the loading process and potentially improve performance.
Testing and Gathering Feedback
The Performance team plans to further test these proposed solutions on a larger scale through the Performance Lab feature project. Before proceeding, they are seeking feedback from the broader WordPress community. The deadline for providing feedback on the proposal, which includes benchmarks and additional analysis details, is August 6, 2023.
Conclusion
The performance analysis conducted on WordPress sites with translations has shed light on the impact of translations on server response times. To address this issue, WordPress contributors have proposed several technical solutions aimed at improving performance for localized sites. These solutions range from using different file formats to revamping the translations parser. By gathering feedback from the community and conducting extensive testing, WordPress aims to enhance the performance of translated sites and ensure a seamless user experience for all.
