Change Details

The mobile skin performs various transformations on content that would be more efficiently done inside the parser itself rather than retroactively on the parsed and cached content. To support our work in lazy loading of images and references to improve the experience for users on 2G connections we would ideally like to omit the HTML for the image tag and references list inside the parser itself. Obviously we do not want to impact the HTML version of the desktop site until we can make a strong case for performing these changes for all users. Questions to answer: [] Is this even an option, will @timstarling scream at us and chase us with a pitchfork? [] What are the trade-offs of maintaining a separate parser cache? [] How would we go about doing this? [] Is it possible to vary parts of the parser output e.g. at the references level and images level? Duration: 4hrs

The mobile skin performs various transformations on content that would be more efficiently done inside the parser itself rather than retroactively on the parsed and cached content. To support our work in lazy loading of images and references to improve the experience for users on 2G connections we would ideally like to omit the HTML for the image tag and references list inside the parser itself. Obviously we do not want to impact the HTML version of the desktop site until we can make a strong case for performing these changes for all users. Questions to answer: [] Is this even an option, will @tstarling scream at us and chase us with a pitchfork? [] What are the trade-offs of maintaining a separate parser cache? [] How would we go about doing this? [] Is it possible to vary parts of the parser output e.g. at the references level and images level? Duration: 4hrs

The mobile skin performs various transformations on content that would be more efficiently done inside the parser itself rather than retroactively on the parsed and cached content. To support our work in lazy loading of images and references to improve the experience for users on 2G connections we would ideally like to omit the HTML for the image tag and references list inside the parser itself. Obviously we do not want to impact the HTML version of the desktop site until we can make a strong case for performing these changes for all users. Questions to answer: [] Is this even an option, will @timstarling scream at us and chase us with a pitchfork? [] What are the trade-offs of maintaining a separate parser cache? [] How would we go about doing this? [] Is it possible to vary parts of the parser output e.g. at the references level and images level? Duration: 4hrs