Board Thread:Technical Updates/@comment-26339491-20191203204132/@comment-4970747-20200217115813

But then if they do want to get that full-sized version, they don't get the original image - they get a crappy compressed webp file that they can't use anywhere else instead, because it thinks it's a jpg/png instead.

As others have mentioned, there are plenty of ways to effect this change purely for web crawlers without disrupting the anonymous experience. At the wiki I'm most active on, editors have spent a considerable amount of time building up a large repertoire of photos. Now getting to these photos is a nightmare. We used file pages to help boost the SEO of these images on their own - now all that work has just been thrown down the drain. Not to mention image categories, which are absolutely impossible to access now without links in the main namespace.

In anything, what should have been done is just improving the file pages themselves. Make them more appealing to crawlers. Encourage editors to complete file descriptions. I'm all for SEO and trying to improve it, but is making things better by just throwing out the bad bits really the right move? Or should those weak spots be improved or strengthened?