I have a web site with an image slider. I keep the some of the image tags empty as the images load on when slide comes into view for faster page load. The image tags defined
Set the image src
attribute to #
:
<img data-src="img/portfolio-desktop1-small.png" src="#" alt="Thumbnail">
The HTML passes the W3C validator just fine, and modern browsers know not to try to load the non-existent image.*
For contrast, using a src
value that references a non-existent file results in an unnecessary HTTP request and an error:
<img data-src="img/portfolio-desktop1-small.png" src="bogus.png" alt="Thumbnail">
Failed to load resource: The requested URL was not found on this server.
*Note: I've read conflicting information on how browsers handle the #
. If anyone has definitive information, please add a comment.
Also see related answer by sideshowbarker about the action
attribute:
https://stackoverflow.com/a/32491636
It seems the src="#"
trick used to be a decent workaround but not anymore. What I do now is create a Gulp build task to post-process src="#"
to use an inline data URL of a tiny invisible one pixel SVG.
The essential gulpfile.js bits look like:
const onePixelSvg = '<svg xmlns="http://www.w3.org/2000/svg" width="1" height="1"></svg>';
const placeholderSvg = `"data:image/svg+xml;base64,${Buffer.from(onePixelSvg).toString('base64')}"`;
const task = {
buildHtml() {
return gulp.src('src/html/**/*.html')
.pipe(replace('src="#"', 'src=' + placeholderSvg))
.pipe(gulp.dest('build'));
}
};
gulp.task('build-html', task.buildHtml);
The advantage of using a build task is that 1) the source remains uncluttered, 2) the HTML always validates, and 3) the inline data URL prevents the browser from making an unnecessary and invalid network request.
What happens if you just remove the src attribute then add it on the fly when you need it. The src attribute isn't required. And in my opinion I wouldn't worry about what the w3c validation tool says anyway. As long as you test it in the necessary browsers and it works.