OK, so within my sort of small section here, I'm going to talk about some of the smaller demonstrators or experiments that were developed, one of which on the back of Simplesite and one of which has become after that, to answer two of the questions that were raised at the start of the project. So.

Basically discovering IIIF content. I mean IIIF as most of you will probably know the international Image interoperability framework can help provide Open Access to a wealth of images from institutions across the world all using the same free shared tools and services. The IIIF Consortium itself provides a very nice guide which brings together a great list of these institutions and showing how you can find published live resources. Many of these institutions actually provide well-documented APIs to allow more complex searching. To find the paintings or images or objects in the collection that you're actually required for your research or your activity.

Now, the thing about making use of these open APIs is they're great. There's lots of information in there, but they can be quite complex, so this is just a quick snapshot of the results from the National Gallery Washington API, which I've used quite a lot, which is really good, but it does take a little while to learn the documentation because it is just quite a complex system, so.

The question was, is using the current data that's being provided by the National Gallery Washington or the other in other institutions? Can it be made easier just to get a preview of the trip if contents in there? So here you actually need a long URL. There is a web page to do this as well, but if you're wanting to pull the data directly to make use of and to do work with, you need to construct this big URL and then you actually get all of this data back. You need to go through it to find the IIIF stuff which is actually named. There. You got IIIF manifest URL. And the canvas ideas. You've got some IIIF information in there and to get them out to do something with it.

So the idea was to develop a simple if discovery system and it was carried out to demonstrate what could be done. So for individual institutions or across multiple collections at the same time. This work was supported by three collaborations between three different projects. The practical IIIF project, the EU Horizon 2020 Shock project and the EU horizon 2020 I period HS project now. That is the link to the live site. It does work, it's up and running and you can explore that in your own time afterwards.

Now this is how it basically works, so you have a website where a user can have a keyword search and this goes to a simple discovery endpoint. Now that endpoint then reformats that query into the complex query required for a given institutional API. You get the results. These custom results come back to the end point, which then reformats them into a simple list of resources and pushes them back to a standard viewer. So what the user sees is a keyword search and the IIIF results. Even though you're making use of all of the powerful API options from the different institutions, so this is the current website I'm not going to the lot of into the technical details of how it works. There is a lot of documentation on the web page explaining that. And we just want to sort of show how it works today, but the user interface is a simple search box with a limit and page number option and the viewer default option either Openc Dragon or Mirador, and in this case, you can search across seven different collections.

At the moment the Art Institute of Chicago National Gallery, National Gallery of Art, Washington. The Smithsonian Stats Museum for Kunst in Denmark, the Vienna and the welcome trust. And if you do a search for C, it returns you 187,554 objects. A lot of these are coming from the Smithsonian and you can explore. I mean, it's just open C dragon. You can zoom in and zoom out and explore the images as you require, but openc regard on its own doesn't tend to give you any context of what the images are and where they're from.

So by clicking on the last icon on the top right, you can actually then scroll down the list and get direct connections to the individual images as required. Now Openc Dragon, as I said, provides great for exploring large sets of zooming images, but it's not so good for metadata and context now, Mirador. Another IIIF compliant viewer provides a lot of options for metadata and context for the images, so depending on the needs of a user, the simple discovery system allows you to toggle back and forward between the two options, so you can just push on the mirror door button and you'll get the mirror viewer or push back to get the open sea dragon, you can jump back and forth between the two, and mirrored are obviously gives the option to see all of the information about the individual objects and the images that are inside the.

Search result now the simple IIIF discovery system with a little while ago it was created on the back of the simple site system and mentioned all of the code is up on GitHub and there is an archive on zenodo for people to use. And all of these links will be available in the material we upload and save on the public website at the end of the webinar.

Now the next thing I was wanting to talk about was one of the issues within the National Gallery, which is where I work is. We look to try and support exhibitions and publications with high resolution images and provide more content for guest visitors and users. So over the last ten years or so, the National Gallery has been publicly presenting high resolution images to support some of these activities.

These sets of images have been put together to provide additional content for exhibitions, publications, and research projects. Now this current or old. System was based on the spoke IIP Image Viewer which I built a number of years ago now and as part of the practical IIIF project we wanted to upgrade that to a new system that made use of the layer presentation version three API and the new specifications. So this is the new system.

This is how it looks. It's styled a lot of based on Bootstrap, which is a JavaScript library CSS. Now the structure of the old system was based on a hierarchical. Group of images and type titles and descriptions of those images, and these are all stored in a MySQL database. Now this was easily replicated using groups of Traer manifests, which is a group of images with a title and a description and supporting metadata. These groups of manifests can then be packaged together into leaf collections, so a IIIF collection is a selection of subcollections or sub manifests.

So the simple system was created through our users to explore a complex of collection and then display the the results in Mirador. So that's the link to the live site. I'll leave it as a title in the next few slides so people can have a look if you'd like, but effectively what it does is it's one page of PHP that looks at the current time players collection or manifest, pulls out the details of titles and any metadata that's in there and displays them in a nice, user-friendly, simple screen. On any child resources, be they collections or manifests are listed below.

Quite similar to some of the functionality that's available in Mirador, but in this case, each of the steps have their own distinct URL, so you can reference or stop or start at any particular point in a complex IIIF collection.

And you can look at them either as an interactive list shown here and that happens when there's at least one collection in there. You get an interactive list or. My list of manifests can be dropped straight into mirrored or as you can see here, so you can then explore the images as required and as you want to and explore them so this particular set of images was put together in relation to an exhibition on brochu's lesson in drawing. So there's quite a lot of different drawings in there in relation to one particular painting.

Now the addition is. This is one of the things we've been doing is supporting the publications, so the National Gallery Technical is one of the publications we produced every year looking at the technical examination of paintings and a number of the research that comes out of the National Gallery.

So the old system provided you a link to where you could find a PDF of the article. So this new system actually puts it in line. So if additional metadata is added to the IIIF manifest to indicate a linked PDF, the system can then spot that and will pull in and load in the PDF directly and present it just next to the images. So in this case you can zoom in to this thumbnail as you were reading the particular article there on Reynolds painting technique, and that's worked really smoothly now.

The system is currently optimized to exploit additional where the data in the IIIF document to define the bread crumbs so you can see at the top there's a breadcrumb trail taking you from the home page, right the way down to the given manner set of manifests you're in, and all of the collections nested collections back up to the top, and this comes from the metadata that's stored in the system. You can make the system work from any IIIF collection. It doesn't work at the moment because the development work I was doing, but some additional work is going to be required to make that a bit more robust to allow you to just drop any collection in there and explore it as you as you want.

Now the last thing I wanted to say before I hand on is as part of this work and the result of an interesting discussion on Slack. So I put the question is if one has a big complex IIIF collection. Is there a way to quickly preview all of the images in it without needing to go up and down all those nested collections of manifests, so the actual screen shot on the front page is from an example of how you can do that.

So we built a small collection previewer. Yes, so this is the link of the previewer, and this does work with. We've tested it with a number of IIIF collections in version two and three, and the ones I've used work so far, it seems pretty good, and what it does is it will load you all of the images referenced in that collection or in those manifests into open sea Dragon.

So here you can see an example. All of the images that are actually referenced in the IIIF collection Explorer that I've just presented there. It's about just shy of 1500 now. This does work quite well. It can take a little while for them to load in. If you're loading in that sort of number of images. It can be a little slow. It's currently hard, limited to 2000 because things started to get a little problematic at that point, but it is possible to put larger ones in there, so future work was going to look at potentially combining this with the two player discovery that I mentioned at the start the GUI so that people might be able to have limits and pages and offsets so that you can actually work your way through larger collections of images.

And I will stop there. So that's the IIIF collection preview IIIF Collection Explorer and simple IIIF discovery. 

Thank you.