I want to create a guided viewing experience

[tomcrane]

This long introductory narrative uses a specific bespoke viewing experience to explore more general features of the Manifest Editor. In this case it's a format developed for the V&A, an example here.

I don’t know much about IIIF other than hearing people mention it at work. I work on web content production for a museum.

I have been given an image service URL for the big image, a Google doc full of copy and instructions for the highlights, URLs of some additional photos that go along with each piece of copy, and been told to create a guided viewing experience.¹

I open the manifest editor at the URL given to me:
https://example.org/manifest-editor?configuration=single-image-guided-viewing ²

The Manifest Editor opens with just one option available - in a list of templates for creating a new Manifest, there is only one, so I click on it.³

As soon as it opens, it prompts me for “Image or Image Service (?)”. It’s a text box.⁴

I don’t bother clicking the Help icon as I assume that this is the URL I was given, so I paste it in.⁵

It starts loading as soon as I paste, but I stay in the dialogue - it shows me a preview and the image dimensions.⁶

Satisfied that I’ve been given (and added) the right image, I click OK.

The Manifest Editor creates the first and only canvas from this image. There’s no Canvas strip, no manifest outline. Config says that annotation tools should be available.

The Image appears. Now I want to go through the Google doc I’ve been given. I need to do this 17 times, for each of the points of interest:

• Draw a box on the image to mark out the region of interest
• Provide a title for this region of interest
• Edit the text in a WYSIWYG editor.
• The text includes some images, but I’ve also got those in the Google doc too, so I can just add them using the editor, which supports basic layout. The images are already on the web.

As I add these 17 points of interest, they build up in a list. I can go back and edit previous ones - e.g., I realised I got the title wrong on one of them. I can also change the order of the ones I have already made (I realise I missed one).

After a while I hit the Preview button - I didn’t choose the dropdown options, just clicked the button.

A new tab opens showing me what it looks like for the users. Something like this.

I do this a few times as I add new regions of interest. I realise I have put the box for one of them in the wrong place so I go back and reposition the box for that one.⁷

After I’m done for today I hit the Save button.

My Google Doc says “save the Manifest in the IIIF Collection Guided Viewing”. There are two main options in this Save UI - Permalink and IIIF Collections so I assume the latter. I go in there and can see other folders related to my department’s work. I can quickly find the folder I need, and I save it in there.⁸

After saving, the Save UI shows me a link to my Manifest, and a link to my Manifest in the Preview environment. I’m not sure which of these is needed so I just send them both to the content editor who asked me to do the work.

The next day I get an email with a few update requests. I go back to the editor. A thumbnail/icon for the thing I was working on is sitting right there in the middle of the screen under the heading “Recent” - so I just click on it and jump back in. I make the requested changes, Save again, and send the link back.

Footnotes

1. This is the “already hosted” use case. Manifest editor does not need to accept new image content. ↩

2. This is a general purpose instance of the manifest editor. It has a configuration section named “single-image-guided-viewing”. This configuration:

   • adds the Ocean Liners “viewer” as the default Preview environment (although Mirador and UV still available in the list).
   • Has a single template so that File->New can only create one kind of Manifest
   • The Manifest has institution-specific boilerplate already provided, e.g., provider property, logo.
   ↩

3. It could just “click this for you” and open at the blank manifest - but that would be annoying if you came to it in that config to edit an existing manifest. The current editor is too keen on reopening what you were last working on no matter how long ago it was. The ritual of opening things helps orient the user in an application. Re-opening things should be super-easy through “Recent”. ↩

4. The config knows that this template is single-canvas, single-image-on-the-canvas, and it knows that it hasn’t got that image yet, so it can immediately ask the user for it with no ambiguity. ↩

5. The Manifest Editor doesn’t require that the user specify whether this URL is an image, or a IIIF Image service. It will deduce that by:

   • Analysing the path to test whether it’s a parameterised Image Service URL - i.e., it resolves as an image but there’s an image service sitting behind it. In which case it will deduce the info.json and ask for it
   • Requesting the URL to see what it is - if it’s an image service
   • (include adding /info.json when testing)
   ↩

6. If it’s an image service the Canvas w,h will be that of the service, unless below a threshold size, in which case the Canvas w,h will be multiplied by successive factors of 2 until they exceed the minimum.
   If it’s an image, the client can load it into an <img> tag and obtain the pixel width and height from that (doesn’t have to be visible). ↩

7. All the time this is happening, actions in the UI are updating the Manifest representation in the Vault instance that the editing controls have access to. ↩

8. The configuration established two possible targets for Save - and only these two. The one chosen is one that can POST the manifest to a IIIF Collection for persistence. ↩