Issue 1503, fix continuous behavior

[jamesmisson]

This still needs work and lots of testing so I'll leave it in draft.

This addresses #1503, where some Presentation v2 'continuous' manifests, and all Presentation v3 continuous manifests, weren't displaying as continuous scrolls (i.e. all canvases stitched together).

The problem with the v2 manifests supplied in the issue was that UV was only checking for continuous viewingHints on the sequence level, not manifest level.

I've updated the 'isContinuous' method in Manifold to check both viewingHint (v2) and behavior (v3), and both range/sequence level, and manifest level: IIIF-Commons/manifold#64. So this PR can just use isContinuous to check and display all the canvases at once (rather than viewingHint as before).

This resulted in some glitches in the viewport navigation window (OSD's little map of the whole canvas in the bottom right). This was caused by some code in OpenSeadragonCenterPanel.ts that deliberately resized the viewport navigation for continuous manifests, presumably to make the long scroll format fit more neatly into a long thin viewport. I've rewritten this resizing stuff (currently the minimum and maximum widths/heights are hard-coded).

I've also moved the viewport navigation to the left hand side for vertical continuous manifests. There was a bit of code to hide the thumbnail panel for continuous manifests (also now using isContinuous rather than viewingHint), presumably because with this kind of material the long viewport navigation does the job thumbnail navigation was doing for non-continuous content, so it makes sense for it to be where the thumbnails normally are (and not clash with the right panel too). But the current positioning with the title and the zoom buttons looks bad; it should really be to the left of the title and the buttons (like the left panel usually is), so I'll come back to that.

Some manifests for testing:

• Cookbook (v3, continuous behavior, horizontal): https://iiif.io/api/cookbook/recipe/0011-book-3-behavior/manifest-continuous.json
• The original UCLA scroll from issue 1503 (v2, continuous viewingHint, vertical): https://iiif.library.ucla.edu/ark%3A%2F21198%2Fzz001nwthk/manifest
• Mahabharata mega scroll (v2, continuous viewingHint, vertical): https://librarylabs.ed.ac.uk/iiif/manifest/mahabharataFinal.json
• IDP scroll (v2, continuous viewingHint, horizontal): https://data.idp.bl.uk/iiif/3/manifest/967B14BF8CFF43A3B12837F2829B7BAE

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Preview	Comments	Updated (UTC)
universalviewer	Ready	Preview	Comment	Oct 20, 2025 5:01pm

💡 Enable Vercel Agent with $100 free credit for automated AI reviews

[kirschbombe]

These fixes look great, @jamesmisson - no notes other than I agree about the positioning of the navigation :) Thank you for your work on this.

[LanieOkorodudu]

Thanks @jamesmisson , for addressing issue #1503 . I know this is still a work in progress, so totally understand there’s more to refine. Since you mentioned the left panel being hidden for the thumbnail, I also noticed that the Next and Previous buttons are currently not showing in the image display. Thought I’d mention it in case it hasn’t been caught yet.

[demiankatz]

Thanks, @jamesmisson, this is a big improvement over being completely broken. :-)

What do you think needs to happen to get this out of draft mode?

From my perspective, there are a couple of code cleanup things (see below) but I'd otherwise be happy to accept this as-is in the spirit of incremental progress. If you want to further improve the UI before merging, though, that's fine too.

Please let me know if I can do more to help things along!

[vercel]

@jamesmisson is attempting to deploy a commit to the mnemoscene Team on Vercel.

A member of the Team first needs to authorize it.

[demiankatz]

These are some suggested comment clean-up tasks. I'll commit them all myself so as not to waste more time. After the resulting code rebuilds, I'll do a final round of testing.

[demiankatz]

Unfortunately, the navigator seems to have completely disappeared now, so something is not quite right.

[jamesmisson]

Hi @demiankatz , I made a mess of this, sorry. The commits you made three days ago were on top of in-progress work I had accidentally committed (I've abandoned that work now as it got too complicated to finish before the release). I've now reset my branch to the working commit, then made sure any relevant suggestions (about commented out code) were still applied, then pushed again.

[demiankatz]

Looks like some of the old comments didn't go away as expected, but I'll just re-remove them now.

[demiankatz]

Looking good to me now. Thanks again, @jamesmisson!