I recently visited Tatton Park in Knutsford where the deer were rutting. This is a beautiful park to walk and plenty to do for kids too! This was my first time photographing deer this close and they gave me plenty of opportunities. I have shared a few more photos that are available on 500px. This trip might just cause me to go an buy a new lens. A really big long new lens!
You can browse my published portfolio over at 500px. Be sure to share your feedback.
A recent trip to Granada, Spain filled me with joy after an absolutely stunning guided tour around the Alhambra. I also found a little gem of a steak restaurant called Apo restaurante on Plaza de San Lázaro 15. The trip finished with a road trip over to Gibraltar and Seville. The views of Africa from the top of the Gibraltar rock where breathtaking. The guided tour also provided an opportunity to get close to photography the Barbary macaque monkeys. Seville has plenty of sights to take in with my favourites being the Cathedral of Saint Mary and the Plaza de España. I have shared a few more photos that are available on 500px.
You can browse my published portfolio over at 500px. Be sure to share your feedback.
I have been working with a large government department where Microsoft Services helped them transform their intranet to SharePoint Online leveraging all the modern capabilities available as well as rethinking what an intranet was. This work included envisioning to fully exploit all of Office 365 and a crucial Information Architecture (IA) design which also mapped their existing, complex and poorly performing hierarchy of subsites to several new hub sites and a completely flattened site structure. I hope to write more about this work soon.
Primary intranet site navigation provider
To create a consistent user experience, the look and feel for several hub sites was replicated, creating the feel of a single intranet. The hub site navigation also needed to be replicated, and that is wherein the challenge lay. The navigation cannot be easily copied from one site to another, let alone between hub sites, and thus this solution was born.
We decided to create a dedicated hub site that only the intranet team can access. Within this hub site, we were then able to create the intended intranet navigation and get the headings, order and links just right without troubling anyone. Once we were happy with the navigation and testing successfully passed, the idea was that we would be able to replicate the hub site navigation to all our other hub sites. More importantly, the process could be easily repeated, keeping the intranet navigation updated across many hub sites. The task could also be automated but we opted to leave it as a manually initiated task.
Script to replication hub site navigation to other sites
The solution to the problem is a PowerShell script made up of two functions. One function that gets the navigation from the hub site acting as the master or primary navigation provider (supplied from a parameter). The other function replicates the navigation to other hub sites. The Get-SPOHubSiteNavigation function exports the hub site navigation from the site provided to a CSV file. The CSV file is then used by the Copy-SPOHubSiteNavigation function to replicate the navigation links to other sites. The process could be hijacked somewhat to add hub site navigation from a CSV file rather than reading it from an existing hub site. We found it was easier to get a feel for the navigation in a real SharePoint site rather than work with the information in a spreadsheet.
You can also disable the export process by setting the -Export parameter to -Export:$false. This switch makes the Get-SPOHubSiteNavigation quite useful for displaying the hub site navigation in an accessible collection. This collection can be used with pipe functions, to filter or sort the collection, for example. This collection is also much easier to work with than using the Get-PnpNavigationNode -Location TopNavigatioBar -Treecmdlet and parameters to show the full navigation structure.
As with many of my posts, I like to share any observations I have made during the work.
The cached Hub navigation appears to refresh every 30 minutes if left unchanged. You can force the cache to refresh by editing the links and then clicking cancel.
A JSON file gets saved to the /_catalogs/hubsite library which could be related to the cached navigation. It has a GUID based name e.g. 20d31c78-8a8c-499a-b953-ecc673344cef.json and appears to get modified by the System Account when changes to the hub site navigation are saved.
The cached JSON file looks like this:
This JSON file, in theory, could be read from a source site and the navigation element copied and injected into the JSON on a target site through manipulation of the JSON payload and the /_API/navigation/SaveMenuState API.
When trying to add a new navigation node using the Add-PnpNavigationNode cmdlet. If the parent node was invalid I got the following error. This wasn’t because the URL linked to a file or folder that didn’t exist as the error below suggests. Dave Garrad kindly calls out the fix for this. The fix is to add -External to the Add-PnPNavigationNode cmdlet to tell PowerShell that the link is not in the same site collection.
The script could benefit additional error handling, particularly to ensure the target site(s) provided are registered as hub sites.
Bonus tip! If you find you have lost changes to the navigation, this same JSON file is part of a library where version history is enabled. This file appears in hub sites and sites associated with a hub site. The version history might help you recover from any loss, but you won’t be able to restore it as the library appears to be read-only.
I hope you find this post useful. As always #SharingIsCaring
So I’ll start by saying how bowled over I am by the engagement I’ve had after posting a tweet sharing that I had programmatically updated the new footer. For this, thank you, this kind of engagement and praise is why I love the community and contributing to it.
I want to share some numbers with you, which was one of the main reasons for pursuing this task.
My customer had 180 sites where the footer needed configuring.
Settings, change the look, select footer (three clicks)
Enable footer, browser and select a logo and then upload it, enable displaying the footer name and then add the footer name text, apply settings and close settings (eight clicks)
Edit links (one click), create a new link, add text, add a URL, save the link, then (four clicks) and then repeat for three other links (twelve further clicks). Then save the navigation (one clicks). Eighteen clicks for footer links.
That is a total of 29 clicks per site — a total of 4,860 clicks for all these sites.
Almost five thousand clicks is a lot of mouse work and this number doesn’t even factor in browsing to each site nor typing each of the values and URLs, which leaves a lot of room for errors too!
To emphasise just why I wanted to achieve this programmatically. Colleagues started to perform this manually while I did some R&D. During this time a collection of sites were updated. However, even with signed-off designs, change request and final confirmation, the customer still changed their mind hence why I like to code and work lean. This kind of change is not uncommon. However, for any organisation that centrally manages sites, has a flattened modern information architecture (IA) or a Hubified intranet with many sites where a standard and consistent footer might be wanted, not being able to manage the footer through PowerShell or other means is a problem. Even managing variations of a standard footer is likely to be an issue for if there is no way to handle it programmatically.
The intranet team within a large government department I have been working closely with over the last two years are responsible for close to 200 sites. Performing tasks against these sites and making changes to them efficiently using PowerShell or other means is essential. Finally, we should not forget that trends, standards and regulations change over time. One such example is that regulations for internet and intranet sites within the government might require that certain information must be readily available from the footer, which would lead to a need for wide-spread changes. Even with their carefully designed flat site structure as part of their information architecture (IA) which leverages Hub sites this change is still site-by-site as the footer configuration is not inherited from a Hub site and therefore set independently.
Exploring how the Footer works
So with all this in mind, I started to explore how this could be achieved using PowerShell. The options and tools are limited and restricted given the environment. I felt that if this could be through the UI, then there must have been an API I could play with to achieve the same result. Off I went in my lab using Fiddler, SharePoint Online Client Browser, Edge Insider developer tools, Postman and VS Code. Troubleshooting and reverse engineering things is a childhood pastime. I found many other exciting easter eggs during this exercise that I hope to share too, but it was like there was a bounty on resolving the footer configuration.
Colleagues in my team were also looking for a solution to similar problems for their customers. A colleague quickly followed up about some undocumented verbs he found in the site design schema (see Site design JSON schema on Microsoft Docs). This news also confirms what I learned from some PG colleagues. Sadly, the footer only has site design support right now.
What I discovered through Edge tools that each change to the footer through the editing footer links or from the change the look pane invoked an API post to “/_API/SaveMenuState”. This post had a JSON payload with the configuration of the footer and also included any footer links. Win!
Initial approach and exploration
I took off with examples of this JSON payload that I obtained from Developer Tools after applying several different footer configurations. I started exploring how the API worked using Postman (this is an excellent tool by the way). To use Postman, I had to create an SP app to get an auth token from AAD so I could use Postman with SPO APIs. I gave the app full tenant access so I could explore things better. Getting started with Postman was a great lesson and one I want to share in a separate post, but Postman is now a vital tool for me when working with SharePoint Online. Once I had played with the API, I set out to achieve the same behaviour in PowerShell.
I set out with the intent to share my work as I know this was going to be of interest to lots of others. I decided to create multiple functions, so each part of the footer can be configured separately. I also built this with a view of contributing and sharing it as part of the SharePoint PnP module. The code is functional and not perfect. Let me walk you through it and share the entire thing at the end too.
Note: the script uses the debug log the output information ($DebugPreference = "Continue").
Enable/Disable SPO Footer – These two functions were created to quickly enabled and disable the footer with a simple function call.
Get/Set SPO Footer (Logo and Text) – It is useful to be able to get the configuration of both the footer text and the footer logo. These functions can be helpful to validate the configuration, for example. These both exist in the JSON as nodes with consistent GUID for their “title” and similar they both have a “key” that identifies them. To simplify the functions I created an overarching function to run them all.
Enable/Disable-Footer – Enables and disables the footer
Get/Set-SPOFooter – Get the configuration of the footer
Get/Set-SPOFooterText – Set the footer text
Get/Set-SPOFooterLogo – Set the footer logo
Set-SPOFooterLinks – Create footer links
Here is the script with all the functions and some examples. I’m going to reach out to Vesa and the PnP team to see about getting this included somehow. But, for now, enjoy. If you have any issues or comments, please let me know!
Quirks to be aware of
During my testing, I noticed a number of quirks I had to work around or that I want to share and make you aware of.
When using Postman between sessions, you must refresh your auth token to continue accessing SPO.
No value is available when reviewing $site.FooterEnabled using the Get-PnPSite cmdlet so you cannot check to see if the footer is enabled or not.
When getting the footer, I noticed there seem to be several different states for how the footer configuration which might impact the GET functions however I have tried to mitigate this risk.
Enabled with no configuration
Enabled, but the JSON appears similar to when it is it disabled
Enabled with configuration and or links.
I noticed some sites had the footer enabled after the footer feature was released, while others it was disabled, this was why I built the function to check how the configuration of the footer.
I hope you find this post helpful. Your engagement and feedback are what drive me to write posts like this, so please keep it up!