We have books that we host which we import a copy of and host on the Pressbooks Network. There are images missing from the imported version hosted on the PressBooks Newtwork.
I’ve found that image src URL is different in the book on Pressbooks Network versus the book we host. Problematic images have #fixme appended to the end of the src URL.
Where did the #fixme come from and how can I fix these images?
Thank you!
Hi Heather – this usually happens during a clone or import routine where we’re unable to locate the source image. In this case, it may be because the images in the source book aren’t hosted in the book, but are being provided by external CDN. Not sure if that’s what’s happening here, but we’re happy to look into it for your hosted network and hopefully suggest method for fixing quickly. Can you email URLs of the two books and any details about how you cloned/imported them from one location to the other to us at premium support?
In this case, I went to the original book. It looks like the path for this image is currently: https://content-cdn.one.lumenlearning.com/wp-content/uploads/sites/14/2023/03/01150545/1024px-Margaret_Floy_Washburn.jpeg. In the clone, it looks like it is https://content.one.lumenlearning.com/introductiontopsychology/wp-content/uploads/sites/14/2023/03/1024px-Margaret_Floy_Washburn.jpeg#fixme. My suspicion is that whenever this clone/import was made, the image was pointing to that latter URL, which no longer exists, as the media has been moved to a CDN server. Does that sound possible?
HI Steel,
Thank you for responding! The books on the Pressbooks Network have been recently imported. Both image paths have the same segments of 14/2023/03 which looks like it could be the date the image was uploaded. It’s also interesting that it doesn’t affect all images. Images hosted on AWS have the correct path in both versions of the book.
@SteelWagstaff Missing images have been discovered in another book. The issue is very similar. The book that is missing images is https://printed.lumenlearning.com/introbusiness/.
The following pages are missing images:
There could be more missing images we haven’t discovered yet.
It appears that the missing images have a URL that begins with content.one.lumenlearning.com and ends with #fixme.
Any insights you can share would be greatly appreciated! Thank you!!
Heather
@SteelWagstaff Checking back in to see if you saw this message. Thanks!!
Hi @Heather_Taylor I did not. Looking now. For future reference, these kinds of questions are best suited for our premium support email address – they’ll always be seen and responded do within 1-2 business days.
Heather – went to look at the first example provided. The Pressbooks book points to this URL: https://content.one.lumenlearning.com/introductiontobusiness/wp-content/uploads/sites/15/2023/02/L1.1.4.png
The image that you’re trying to reference appears to be hosted at https://content-cdn.one.lumenlearning.com/wp-content/uploads/sites/15/2023/02/19132358/L1.1.4.png instead.
To fix, you ought to be able to simply replace the broken URL with the correct one (or set up redirect rules so that content.one.lumenlearning.com → content-cdn.one.lumenlearning.com. It appears that you may have originally hosted these images at one subdomain before moving them to their current location?
Hi Steel,
Thank you for the response. 
Can you share with me how to set up redirect rules? Would each image require a redirect rule or would it be possible to set up a redirect rule with a wildcard that looks for any URL that has #fixme appended to the end?
I’m looking for the most efficient way for us to implement a solution.
Thanks!!
I think those are URLs that you folks control. I have no idea what’s happening on those servers or how the CDN was set up. There are probably good ways to set up permanent rewrite/redirect rules, but they might depend on whether you’re using NGINX or Apache or if you have other server level configurations that we’re just not privy to. My colleague @Christopher_Murtagh is our ops expert and will definitely know more about this domain if you want to connect someone from the devops side at Lumen with someone here?
I’ll touch base with leadership at Lumen to see which direction they would like to pursue, either updating the image URLs or creating redirects. If we choose the later, I’ll touch base with Christopher.
Thanks so much! I appreciate it!
Thanks Heather. It looks like the original books (like Introduction to Business – Simple Book Publishing ) have correct URLs. I suspect that whenever you all made the change to how images are hosted, the URLs got updated in these books. The cloned/imported copies we have on our PB hosted instance don’t point to the same URLs anymore. This happens during our cloning routine – when we find broken images we append a little #fixme to the end of the URL: pressbooks/inc/cloner/class-downloads.php at a4ab87ad08ef8ae2c1eddf985c79e8dce5bcaaa0 · pressbooks/pressbooks · GitHub. This means that the images were broken at the time of the import/clone into the Pressbooks-hosted instance. If you were able to reclone or reimport the books, that might also take care of the issue quickly, since I think the URLs are only broken in the Pressbooks-hosted instance. I tried cloning the Lumen Learning book but was blocked from doing so, presumably due to a setting you have in place for the PB network you all are running for the ‘canonical’ version of these texts. Does that make sense? Happy to take a quick call with you and any colleagues if that’s helpful. My calendar is up to date here: https://meetings.hubspot.com/steel-wagstaff
That makes sense! I was wondering if reimporting the books might take care of the issue. Thank you for the additional context!
