Tiny Elf runs on a time-driven trigger. Unfortunately, the script run by one user does not have the permission to delete triggers set by another user. Thus, the original owner of the trigger needs to delete the trigger they have set up.

Original Owner - Deleting the trigger

This needs to be done by the person who has triggers currently automating Tiny Elf.

1. Open your Tiny Elf Spreadsheet

2. Go to the Tiny Elf menu and click on Delete trigger (Pauses automated notifications)

   

3. Once the trigger is successfully deleted, you should get a notification saying Trigger has been deleted.

   

Original Owner - Verifying if your trigger is deleted

1. Open your Tiny Elf Apps Script project.

   • This can be done by going to your Tiny Elf Spreadsheet.

   • Click Extensions → Apps Scripts

   • Then Tiny Elf - Rostering bot Apps Script project will open in a new window.

2. On the left, click Triggers alarm. You should see any triggers you set here.

   

3. You should not see any scripts Owned by Me. That’s SUCCESS! 💪

4. If you do see any such triggers, click on the meatball menu and hit Delete trigger

   

   New Owner - Setting up automated notifications

   1. If you have been pointed at a Tiny Elf Spreadsheet and asked to take ownership of its triggers, the first step is authorising Tiny Elf. Do this by clicking on the Tiny Elf menu → Authorise Tiny Elf.

2. Click OK on the next pop-up.

   

3. Choose an Account in the next pop-up.

   

4. Click Continue

   

5. Select All in the next screen and click Continue at the bottom of the screen.

   

6. You should see a success message. Click OK

   

7. You should now see the full Tiny Elf menu with four menu items.

   

8. Verify your trigger settings in the Settings sheet.

   

9. Click Set/reset automated trigger in the Tiny Elf menu.

   

10. You should get a message saying Success - Trigger has been set\reset. Click Ok.

    

    Congratulations! That’s it. You have now transferred ownership of the automated trigger to yourself. 🌟

It's been 21 years since I started in the Software world, albeit as an intern. Having wanted to become a Software Engineer (without knowing precisely what that meant) from a young age, I can say this was a dream come true. Digging a trench (a physical one, not a metaphorical trench) to install a smart doorbell brought a few practices in the Software industry into perspective.

Digging a trench

Trench, n
gen. A long, narrow ditch or furrow cut out of the ground. Also figurative.
Oxford English Dictionary, s.v. “trench, n., sense I.3”, July 2023. https://doi.org/10.1093/OED/4609488321

Anyone can dig a trench. Similarly, anyone can write code. But let's look at the differences between how an expert and an amateur digs a trench.

It's just a hole. It's not rocket science you might say! But you need to consider tooling, skills, experience, efficiency and quality.

Tooling

You need the right tools to dig a trench. This could be a spade, an auger, a mattock, a shovel or even a power tool. It needs to fit the space you have to work in and the size of the trench you need.

While digging my trench, I started with a spade and soon realised it did not do the job. Then it was off to Bunnings to buy a mattock and another trip to buy an auger. It was a surprise that I escaped any serious bodily harm while using the auger with a power drill.

I needed to have the right tools to dig a trench and I was unprepared.

Skills

I have dug holes for planting before without any formal training. But this trench needed to be 10 metres long, 10 centimetres wide and a metre deep. It needed to be dug right next to a concrete driveway.

But I believe this would have been done well had I learnt the necessary skills as a builder's apprentice. So it was all trial and error as I went about digging this trench. YouTube videos helped to some extent, but they did not replace the need for some form of formal training.

Experience

I had never dug a trench before. So I had no experience as to what I should be doing. The ground was rock hard (pun intended) and I had no clue as to which tools to use first.

As I went about digging the trench I came across tree roots and concrete which I didn't anticipate and had no idea how to deal with.

Efficiency

I recruited my wife and son to help on this project and it took us over a weekend to dig this trench. There were multiple trips to Bunnings, stops for YouTube instructional videos and breaks when we were totally clueless and frustrated.

Remember, it was a 10m x 1m X 10cm trench and it took 48 hours to dig from start to finish.

Quality

The quality of the trench was a huge compromise. Digging to a metre deep was too hard and it ended up being just 40 cm deep and even a little less in certain places. The width was larger and smaller in several places depending on which tool was used. Fortunately, as we were covering up the trench once the conduit was laid, it was not a big concern. But it would be a concern if anyone ever dug up the same area.

When it was too hard to cut through the roots, we tunnelled through it. That meant we had to use a flexible conduit in that location rather than the hard conduit used elsewhere.

If my Builder friend ever saw this trench, he's going to faint. 😮

Trench ≠ Software Engineering

So what has digging a trench got to do with Software Engineering? I would take those same principles of tooling, skills, experience, efficiency and quality, and apply them to Software Engineering practices. These are by no means the only considerations when building software, but are the parallel lessons I took away while digging that trench.

Writing Software

Software Engineering, n.
The practice or process of designing, developing, and maintaining software; the field of study concerned with this.
Oxford English Dictionary, s.v. “software engineering, n.”, July 2023. https://doi.org/10.1093/OED/2334731322

Anyone can write code. Even a toddler writes code these days with tools such as Scratch. But would it turn out to be like my trench? Let's draw some parallels to it.

Tooling

You can write code in a simple text editor such as Notepad or an Integrated Development Environment (IDE) such as Visual Studio or VS Code. While Notepad would allow you to write code that would work, it would not help with syntax checking, IntelliSense, debugging and other advanced features provided by an IDE. Having the right tool aids productivity.

Also, tools such as ReSharper, Snyk and GitHub Co-pilot give the developer an edge. Similarly for other disciplines in Software Engineering, tools such as Azure DevOps, JIRA, Figma and Package Managers make all the difference.

Skills

Before I started to program professionally I used physical paperbacks such as "GW Basic - User's Guide and Reference", "Microsoft Visual Basic 5 - Step by Step" and other bulky books to learn programming in addition to what I learnt in high school.

As I went into a professional setting in 2002, MSDN was a popular resource and Microsoft Certified Professional study courses led the way in learning new languages and skills. I was doing my degree in Computer Science while working and these resources helped me gain skills quickly.

Currently, I would recommend Pluralsight, Microsoft Learn, and specific Platform certifications for Developers to skill up in their areas of interest.

While a degree in Computer Science or a related field can really set you up to win in Software Engineering, I have seen several smart and enthusiastic people transition from other careers by following a good boot camp. So don't fret if you don't have a degree in CS, there are other ways to skill up.

I'm a firm believer in certifications. But certifications do not trump experience. Certifications are a way to keep up to date on technology and receive the minimum knowledge on that particular version of a platform or tool. Getting a certificate does not mean that you are an "expert".

Experience

An experienced developer with a few years under their belt works differently from someone who's new to the workforce. They know certain things inherently based on the years they have dedicated to a technology or a platform. For example, while I can read and understand PHP, I don't call myself an expert on the LAMP stack as I would do with ASP.NET on the Microsoft stack.

But the years of experience do not count if a developer does not upskill and keep up with the latest changes in the technology they specialise in. So experience and skills go hand in hand as you see in the graph below.

The above graph shows how a developer who is learning continuously skills up compared to a developer who has stopped that continuous learning or only learns on the job (pertaining to tasks they are assigned).

Efficiency

Long-term efficiency is the name of the game. Remember you are running a marathon and not a sprint. So be efficient, but pace yourself.

One of my mentors (Hi Andy 👋) says, "Efficiency is how fast would you build a website/software for your girlfriend, wife or partner". It's a labour of love but done in the shortest possible time. You deliver a quality output but you are super efficient as you go about it. In the world of Agile, you would be super agile in this instance.

If you are efficient, that gives you the time and space to learn new things on the job as 70% of a developer's learning is on the job.

Quality

Quality is not only ticking some checkboxes to say that your output passes the listed acceptance criteria. The end product should be something you are proud of. It should be something you can show off to the world. It should do what the label promised and do it well.

Your end product should be functional, reliable, performant, secure, and maintainable. This would involve rigorous testing, continuous improvement, and adherence to industry best practices to keep that quality.

Bringing it all together

Having the best tools in the trade is not going to make you better at Software Engineering. Nor are textbook skills. You need a combination of tools, skills and experience to produce quality output efficiently.

Sensible Decisions

Do I have more tools in my garden shed? I do! But that doesn't make me a better trench digger. Similarly, having the best Software tools doesn't make me a better Software Engineer.

Have I gained skills in digging a trench? Yes, YouTube videos do count. But am I skilled in it to dig another trench? I would say No! Similarly, learning new skills as a Software Engineer doesn't make you an expert. But it does make you a better beginner.

Have I gained some experience in digging a trench? Definitely. But would I dig another trench? I don't think so. I will leave it to the professionals. Similarly, I would be more comfortable working with a tech stack I'm experienced in rather than another tech stack I have passing knowledge of.

You pay more for the right tooling, skills and experience. You go to the experts because you need a quality output that is produced efficiently. If you find the right experts or that brighter agency💡, you get it done right the first time.

So next time when you need to dig a trench or get some Software developed think of Tooling, Skills and Experience that deliver Quality Efficiently. You don't want to end up with an expensive auger and a half dug hole.

P.S. - I couldn't resist adding this recent video from my smart doorbell (which is the reason I dug a trench that became the inspiration for this article)

Happy 24th Birthday! I'm glad to be celebrating with you. 🥂

It's with great joy and excitement that I pen this letter to you. As I reflect on this 24th birthday milestone, I can't help but feel a deep sense of gratitude for the incredible journey we've shared together.

Luminary is not just a great digital agency, it's a collection of awesome human beings doing great things together. You have been more than just a workplace to me; you have been a source of inspiration, growth, and fulfillment.

I joined Luminary eight years ago, with an eagerness to contribute my skills and make a meaningful impact in the software industry. Little did I know that this would mark the beginning of an incredible chapter in my career of 21 years. Over the past eight years, or approximately one-third of Luminary's lifetime, I have witnessed firsthand the remarkable growth and success of this organization.

I started as a Senior developer, eager to learn and make a difference. With each passing year, Luminary provided me with opportunities to step up and take on greater responsibilities. This allowed me to be promoted to the role of Technical Lead, and eventually, I reached the position of Technical Director. Luminary has not only recognized my potential but has also nurtured and supported my passions, enabling me to evolve both professionally and personally.

I am grateful for the owners of Luminary, Marty Drill (CEO and Founder), Adam Griffith (Managing Director), and Andy Thompson (CTO), who have created an environment that fosters innovation, collaboration, and growth. Their leadership and vision have been instrumental in shaping Luminary into the vibrant and dynamic organization it is today. I would also like to extend my heartfelt thanks to the leadership team, all our team members, both past and present, who have contributed their talents and expertise to make Luminary a powerhouse in the digital industry. Each individual brings a unique brilliance to our collective effort, and I am fortunate to work with such bright minds on a daily basis.

One of our proudest achievement this year was receiving the prestigious overall site of the year award in the Australian Web Awards for the website we built for UNICEF Australia. This recognition is a testament to the tireless dedication and passion that each Luminary team member puts into their work. We strive to make a difference in the digital landscape and create experiences that leave a positive impact on people's (and especially children's) lives.

In the words of Confucius 孔子, "Choose a job you love, and you'll never have to work a day in your life." I can honestly say that I am yet to work a day at Luminary. The passion, enthusiasm, and commitment that permeate our workspace make it feel less like a job and more like a fulfilling journey. The challenging projects, the creative problem-solving, and the opportunity to shape the digital world all fuel my love for what I do.

However, as our MD Adam said, it's important to acknowledge that it isn't roses every day. We are all human, and we experience our fair share of ups and downs. Yet, it is during these moments of challenge that Luminary truly shines. We come together as a team, supporting and encouraging one another, finding solutions, and emerging stronger than before. This camaraderie and resilience make Luminary not just a workplace but a family.

Talking about Family, my wife Suzie and son Jordan are also a part of that Luminary family. They are known and have made connections over the years. As my colleague Sarah said, when the kids (who are teenagers now) meet up once or twice a year, it's like meeting your cousins who live far away. It's the small things which make Luminary great.

It's also that outward look which you foster as a BCorp - People, Planet and Profit makes you unique among your peers, Luminary! As Liam, our Engagement Director, said, with Luminary you don't experience the stress that hinders some other agencies.

As we celebrate your 24th birthday, Luminary, I am filled with gratitude for the incredible journey we have shared. You have provided me with a platform to learn, grow, and make a meaningful impact in the digital world. Here's to many more years of success, innovation, and brightening the digital landscape.

Happy Birthday, Luminary - H5YR! (For the uninitiated, it means High Five You Rock!)

With warmest regards,

Emmanuel

As a member of the Umbraco Headless Community Team, I was aware of this and we as a team had made several suggestions on the direction of this feature.

Documentation

The official Content Delivery API documentation is pretty comprehensive and is a good place to start.

Installing 12RC1 on your local machine

If you want to try it out on your local machine, the Package Script Writer by Paul Seal is an excellent resource. Within a few minutes, you will have Umbraco 12 RC 1 installed and ready to test out on your local machine.

Installing 12RC1 on an Azure Web App

But of course, I wanted to try this out on Azure and share it with the devs at Luminary. I've got this nifty script that I run on an Azure Cloud Shell. In around 20 minutes I get a fully functional Umbraco 12 RC 1 with the Content Delivery API enabled.

wget https://www.speedycmsinstall.com/scripts/azure/v12rc1/create-umbraco-v12rc1-demo.sh
chmod +x create-umbraco-v12rc1-demo.sh
./create-umbraco-v12rc1-demo.sh

You can fork the bash script from GitHub and tinker with it for your own needs. If you need more details on this script, you can read all about it in Scripting an Azure deployment of Umbraco 9.

Starter Kit installed

My bash script also installed the "Starter Kit" which gave me some content to play with while testing out the new Content Delivery API.

Trying out the Content Delivery API

I tried the most obvious choices for testing out the content delivery API. Note that the following URLs may not work in the future as this test environment is ephemeral.

Retrieving all the content

Retrieving all the content was easy enough.

curl --location "https://umbracov12rc1-demo-19175.azurewebsites.net/umbraco/delivery/api/v1/content"

But it seemed I was missing something. I know that my name was included in this Starter Kit. But the Delivery API did not return that item to me.

{
    "total": 30,
    "items": [
        {
            "name": "Home",
            "route": {
                "path": "/",
                "startItem": {
                    "id": "ca4249ed-2b23-4337-b522-63cabe5587d1",
                    "path": "home"
                }
            },
            "id": "ca4249ed-2b23-4337-b522-63cabe5587d1",
            "contentType": "home",
            "properties": {
                "heroHeader": "Umbraco Demo",
                // removed for brevity
            },
            "cultures": {}
        },
        {
            "name": "Products",
            // removed for brevity
        },
        // removed for brevity
        {
            // item #10
            "name": "Biker Jacket",
            // removed for brevity
        }
    ]
}

Since it said that there were 30 items in total, I tried the take query parameter as follows:

curl --location "https://umbracov12rc1-demo-19175.azurewebsites.net/umbraco/delivery/api/v1/content?take=30"

Voila! I can now see 30 items and myself in the JSON results.

{
    "total": 30,
    "items": [
        {
            "name": "Home",
            // removed for brevity
        },
        // removed for brevity
        {
            "name": "People",
            // removed for brevity
        },
        // removed for brevity
        {
            "name": "Emmanuel Tissera",
            "route": {
                "path": "/people/emmanuel-tissera/",
                // removed for brevity
            },
            "id": "6374bda9-034a-4109-a837-72ed6a325f11",
            "contentType": "person",
            "properties": {
                "seoMetaDescription": "",
                "keywords": [],
                "umbracoNavihide": false,
                "photo": [
                    {
                        "id": "77db56ab-2f79-4eb2-82fd-eb7266e6b366",
                        "name": "Emmanuel Tissera",
                        "mediaType": "Image",
                        "url": "/media/vvlnw31z/emmanuel_tissera.jpg",
                        "extension": "jpg",
                        "width": 1600,
                        "height": 1200,
                        "bytes": 772875,
                        "properties": {},
                        "focalPoint": null,
                        "crops": []
                    }
                ],
                "department": [
                    "Australia",
                    "mvp"
                ],
                "email": "",
                "twitterUsername": "",
                "facebookUsername": "",
                "linkedInUsername": "",
                "instagramUsername": ""
            },
            "cultures": {}
        },
        // removed for brevity
        {
            // Item #30
            "name": "Contact",
            // removed for brevity
        }
    ]
}

⚠️ I believe the Content Delivery API is missing a pagination block to let consumers know how many records have been returned and how they should reach for the next set.

Retrieve a particular item via ID

Request:

curl --location "https://umbracov12rc1-demo-19175.azurewebsites.net/umbraco/delivery/api/v1/content/item/6374bda9-034a-4109-a837-72ed6a325f11"

Response:

{
    "name": "Emmanuel Tissera",
    "route": {
        "path": "/people/emmanuel-tissera/",
        "startItem": {
            "id": "ca4249ed-2b23-4337-b522-63cabe5587d1",
            "path": "home"
        }
    },
    "id": "6374bda9-034a-4109-a837-72ed6a325f11",
    "contentType": "person",
    "properties": {
        "seoMetaDescription": "",
        "keywords": [],
        "umbracoNavihide": false,
        "photo": [
            {
                "id": "77db56ab-2f79-4eb2-82fd-eb7266e6b366",
                "name": "Emmanuel Tissera",
                "mediaType": "Image",
                "url": "/media/vvlnw31z/emmanuel_tissera.jpg",
                "extension": "jpg",
                "width": 1600,
                "height": 1200,
                "bytes": 772875,
                "properties": {},
                "focalPoint": null,
                "crops": []
            }
        ],
        "department": [
            "Australia",
            "mvp"
        ],
        "email": "",
        "twitterUsername": "",
        "facebookUsername": "",
        "linkedInUsername": "",
        "instagramUsername": ""
    },
    "cultures": {}
}

🌟Worked as expected.

Retrieve a particular item via Path

Request:

curl --location "https://umbracov12rc1-demo-19175.azurewebsites.net/umbraco/delivery/api/v1/content/item/people/emmanuel-tissera"

Response:

{
    "name": "Emmanuel Tissera",
    "route": {
        "path": "/people/emmanuel-tissera/",
        "startItem": {
            "id": "ca4249ed-2b23-4337-b522-63cabe5587d1",
            "path": "home"
        }
    },
    // removed for brevity - same as above
}

💪Worked as expected and I just needed to only know the path of the content item I was hoping to retrieve.

Retrieve draft/preview content

There are times when you need to preview draft content and the Content Delivery API does a great job with this.

curl --location "https://umbracov12rc1-demo-19175.azurewebsites.net//umbraco/delivery/api/v1/content/item/6374bda9-034a-4109-a837-72ed6a325f11" --header "Preview: true" --header "Api-Key: my-secret"

Response:

{
    "name": "Emm Tissera",
    "route": {
        "path": "/people/emmanuel-tissera/",
        "startItem": {
            "id": "ca4249ed-2b23-4337-b522-63cabe5587d1",
            "path": "home"
        }
    },
    // removed for brevity
    "properties": {
        // removed for brevity
        "email": "",
        "twitterUsername": "dAmazingNut",
        "facebookUsername": "",
        "linkedInUsername": "",
        "instagramUsername": ""
    },
    "cultures": {}
}

👀As you can see there is preview content where I have changed the name of the content item and added a twitterUsername. These changes are not visible on the public API.

Adding a new Document Type

To just test how easy this was, I added a new Document Type called Pet. I populated the details of my two dogs and made this request.

curl --location "https://umbracov12rc1-demo-19175.azurewebsites.net//umbraco/delivery/api/v1/content/?filter=contentType:pet"

Without writing a single line of code in the CMS back office, I get this JSON response.

{
    "total": 2,
    "items": [
        {
            "name": "Shadow",
            "route": {
                "path": "/people/shadow/",
                "startItem": {
                    "id": "ca4249ed-2b23-4337-b522-63cabe5587d1",
                    "path": "home"
                }
            },
            "id": "2b9116f6-a45f-404d-8ef2-64900148eca9",
            "contentType": "pet",
            "properties": {
                "petName": "Shadow",
                "breed": "Staffie",
                "isCuddley": true,
                "dateOfBirth": "2016-10-11T00:00:00Z"
            },
            "cultures": {}
        },
        {
            "name": "Athena",
            "route": {
                "path": "/people/athena/",
                "startItem": {
                    "id": "ca4249ed-2b23-4337-b522-63cabe5587d1",
                    "path": "home"
                }
            },
            "id": "64bf848f-06f3-4856-b730-84d4dbbeab08",
            "contentType": "pet",
            "properties": {
                "petName": "Athena",
                "breed": "Staffie",
                "isCuddley": true,
                "dateOfBirth": "2019-02-11T00:00:00Z"
            },
            "cultures": {}
        }
    ]
}

The Verdict

That was a quick whirlwind tour of the new Umbraco Content Delivery API which is a part of the core offering and I'm loving it. I need to explore the localisation features and see if it passes my GILT or GUILT – Localisation crimes test.

I couldn't see the swagger documentation on Azure and that's something I will explore further.

The team working on the Content Delivery API have done a great job in the first release candidate and I'm hoping they will fix the current limitations they have listed to make this product shine.

Provide your own feedback

If you find issues that are not already reported, please report them on the GitHub tracker by following the link or selecting “🐛 Bug Report” when creating a new issue. That reminds me that I should file a suggestion for that pagination issue I encountered.

Public release

I'm looking forward to the public release of Umbraco 12, which is targeted for June 29, 2023.

The Possibilities

With the Content Delivery API in place, a developer could create new Document types and have Content Authors editing adding and updating content even before the front-end has been built.

This allows for a composable architecture where the content is just another API. With the Content Delivery API in place, Umbraco has placed itself in a strong position in that hybrid headless space.

Setup in five steps

Your Tiny Elf bot can be set up in five easy steps.

1. Make your copy of the Tiny Elf spreadsheet
2. Change the team data
3. Change the settings
4. Authorise the script
5. Set the trigger

You are done 🎉

Now, wait for the next team member to be rostered 🔔

Features

Tiny Elf comes with a few nifty features.

• The sheet owner can change settings directly from the settings sheet. No need to touch the Google Apps Script files.
• The bot can check if the event exists before sending out notifications. This helps with skipping notifications for meetings that are not held on a daily/weekly basis.
• When rostering team members, the bot checks if they are available; have accepted the event invitation; they are available and have accepted the invite. It can also skip the check.
• You can set and reset your automated trigger specifying when you want it to run.
• Allows you to send a notification up to 7 days ahead or on the date of the meeting.
• Allows you to specify custom messages to be sent via Slack with dynamic replacements for certain values.

Open-source

The Tiny Elf source code is available on GitHub with an MIT license. It can be extended and customised to suit your unique requirements. Bug fixes or enhancements that benefit others are welcome. Please submit your pull requests. 😊

The nitty gritty in the spreadsheet

Tiny Elf runs off a Spreadsheet which has team data and all the settings. The Tiny Elf spreadsheet is shared publicly and a copy can be made for your use.

The Team sheet

The Team sheet contains data that is spread across the following six columns.

• Name
• Email address
• Slack member ID
• Enabled
• Roster for days
• Last host date

The Settings sheet

The Settings sheet contains the following settings and each setting is explained in detail.

Calendar event details

• Event title
• Event start hour
• Event start minute
• Event end hour
• Event end minute
• Check event exists
• Roster team members when

Automated daily reminder trigger settings

• Trigger hour
• Trigger minute
• Trigger days before
• Trigger on days

Slack message settings

• Slack Webhook URL
• Team member notification summary
• Team member notification body
• Team member notification footer
• Busy notification summary
• Busy notification

The 'Instructions' and 'License and distribution' sheets

These two sheets contain enough details for an end-user to use Tiny Elf without looking at the source code.

Some real-life scenarios

Tiny Elf is already in use across three teams and here are some ways they use it. Note that each team runs their unique copy of the Tiny Elf Spreadsheet but uses the same code. The data and settings are unique to each scenario.

The settings for each scenario are mentioned here as examples/inspiration. Note that only significant changes are mentioned here for brevity.

Weekday meetings with a Friday special

Here is how the first team who uses Tiny Elf has configured it. The majority of their team attends meetings daily and some of them only on Fridays.

The setup

Roster for days - Mon, Tue, Wed, Thu, Fri is set for those who attend daily and Fri for those who attend only on Fridays.

Trigger days before is set as 0 since this notification goes out each weekday.

Trigger on days is set as Mon, Tue, Wed, Thu, Fri so that it runs every weekday.

Team member notification body is set as

Hi $[slackHandle] :wave:
You are rostered to host *Hello World* today. Here 
are <link to our wiki|instructions> for you to host it!
Give us a shout here if you can't make it.

Monday to Thursday meetings

The second team who uses Tiny Elf has configured it to run only from Monday to Thursday.

The setup

Roster for days is set as Mon, Tue, Wed, Thu for all team members.

Trigger days before is set as 0 since this notification goes out on the day the script is triggered.

Trigger on days is set as Mon, Tue, Wed, Thu so that it runs every weekday excluding Friday.

Team member notification body is set as

Hi $[slackHandle] :wave:
You are rostered to host *FED standup* today. You can lead with:
• What we did yesterday
• What we’re doing today
• Does anyone want to bring anything up
Here are some <link to our wiki|tips> for you to host it! 
Give us a shout here if you can't make it.

Note that you can use the Unicode character • (hex 0x2022) instead of markdown as the Slack documentation suggests.

Meet every other Friday - Notify on Wednesday

The third team has a meeting every other Friday and wanted the notification for the host to go out two days in advance.

The setup

Roster for days is set as Fri for all Team members.

Enabled is set as No for one team member as they were on extended leave. This prevented Tiny Elf from picking them as the host.

Check event exists is set as Yes. This checks to see if the event exists on the calendar before the script is run.

Trigger days before is set as 2 since this notification goes out on Wednesday.

Trigger on days is set as Wed so that it runs every Wednesday only. But as it checks for the calendar event, notifications are sent out only every other week.

Slack message with instructions on your intranet

The Slack notification can contain links to your intranet or company wiki. Use the Slack format of <https://intranet-address.com/page|text description of link> to include links in your notification.

Notify on a private channel

The notifications can be sent to a private Slack channel. The only requirement is that the person setting up the incoming webhook has access to that private channel. Once the webhook URL is configured, Tiny Elf works the same as it would on a public channel.

History and how Tiny Elf came to be

We have got a team of 15+ developers who meet every Monday to Thursday and this increases to 20+ developers every Friday. We wanted to give each team member a chance to host the meeting to take away a central authority and give each individual an opportunity to grow by chairing and running a meeting.

But this was easier said than done since our virtual meeting software (Zoom and Google Meet) didn't natively support rostering. So it became a manual task to keep track of who was rostered and who was next. Going alphabetically by name didn't help since some team members could not to attend all meetings due to other commitments or meetings at the same time.

Hmmm... What if I could automate this process and give the assigned host a heads-up?

Automation - version 1

As I was already maintaining the list of hosts in a Google spreadsheet, I wrote some JavaScript on the Google Script Platform to perform some automation. The code essentially checked each team member's Google calendar for any overlapping meetings and if there were none, it sent out a notification on a shared Slack channel asking them to be the host. It also recorded the last date they hosted the meeting so that it could rotate them after all the others have had a go at hosting the meeting. It also incorporated some smarts to assign certain team members only on a Friday.

Automation - version 1.1

This automation was working so well that a second team requested this bot to help them assign hosts from Monday to Thursday. That was easy enough, but I had to make changes to the source code as almost everything was hard coded. This included the meeting time, title and message templates.

Automation - version 1.2?

Tiny Elf got enough notice that a third team now requested the bot for their meeting. But that meeting was held every other Friday. Also, they wanted notifications to go out on a Wednesday.

Another copy? Another go at changing the source code?

TypeScript to the rescue

Luckily, I had started learning TypeScript and this was a project I was converting from vanilla JavaScript to TypeScript.

Next steps

I have already opened an issue for myself to automate integration testing as it does take time when I have to do it manually.

When reviewing the draft of this post, my wife asked if messages could be sent as an SMS or a WhatsApp message. Tiny Elf wouldn't be a Slack bot then - but it's something worth thinking about.

Resources

As I was learning TypeScript and grappling with Google Scripts, these resources were of immense help and I have borrowed some code/direction from them.

• How to send Slack alerts from Google Sheets / Google Apps Script by Rowan Barnes
• Google Apps & TypeScript Project Template by Ian Sanders
• TypeScript tips from Borislav Hadzhiev

My requirements

I needed two extension methods to do the following:

• A method to convert "Yes"/"No" to it's equivalent boolean value
• A method to convert a date string to a Date object (or the minimum Date if it's a blank or malformed date string)

My C# class

So I wrote my C# extension methods as follows:

public static class StringExtensions {
    public static bool ToBool(this string value){
        return value.ToLower() == "yes";
    }

    public static DateTime ToDate(this string value){
        return DateTime.TryParse(value, out DateTime result) 
                    ? result : DateTime.MinValue;
    }    
}

You can see a working .NET Fiddle below:

Equivalent TypeScript extensions

So I wrote the equivalent extensions in the TypeScript playground

...
String.prototype.toBoolean = function () {
    return this.toLowerCase() === "yes";
};

String.prototype.toDate = function () {
    return this != "" ? new Date(Date.parse(this.toString())) : 
        new Date("01-Jan-2001");
}
...

Interesting, this code gives a compile time error. But still runs 😱

Property 'toBoolean' does not exist on type 'String'.

Property 'toDate' does not exist on type 'String'.

How did I fix it?

I needed to declare an interface for String first as you can see in my playground script and the code below:

interface String {
    toBoolean(): boolean;
    toDate(): Date | undefined;
}

String.prototype.toBoolean = function () {
    return this.toLowerCase() === "yes";
};

String.prototype.toDate = function () {
    return this != "" ? new Date(Date.parse(this.toString())) : 
        new Date("01-Jan-2001");
}
...

This TypeScript code compiled and ran without any issues. Note that when you do this in your local environment, add the interface code to a String.d.ts file to keep it neat.

This SO Answer helped me to write the code.

Why did it work?

This SO Answer gives the explanation as to why adding an interface works:

Declaring the new member so it can pass type-checking. You need to declare an interface with the same name as the constructor/class you want to modify and put it under the correct declared namespace/module. This is called scope augmentation.

I tried extending my custom class from a List<T> with T being a custom class.

My C# Class

Below is my Team class in C#. At runtime, I expect to get a count of 2 for the number of enabled team members with my sample dataset.

public class Team: List<TeamMember> {

    public Team(List<List<string>> rows) {
        this.AddRange(rows.Where(x => !string.IsNullOrEmpty(x[0]))
          .Select(x => new TeamMember(x)));
    }

    public List<TeamMember> getEnabledUsers() {
        return this.FindAll(member => member.enabled);
    }
}

You can see a complete working sample in this fiddle.

TypeScript with a runtime error

I converted that C# class to a TypeScript class. This Team class compiles without any issues. But if you run it on TypeScript playground, you will see a runtime error.

...
class Team extends Array<TeamMember> {

    constructor(rows: Array<Array<string>>) {
        super();
        rows.filter(row => row[0] != "")
            .map(row => this.push(new TeamMember(row)));
    }

    getEnabledUsers() : Array<TeamMember> {
        return this.filter(member => member.enabled);
    }
}
...

Misleading runtime error

The error on the playground or even if you run it locally says:

[ERR]: "Executed JavaScript Failed:"

[ERR]: rows.filter is not a function

So that means that I have done something wrong in the constructor where rows.filter is. Trying to debug what happens in the constructor made me waste a few hours on a Sunday.

But, if I take away the getEnabledUsers() method, this error goes away. 🤯

Reading through multiple SO questions and blogs etc finally led me to Object.values. The Object.values() method returns an array of a given object's own enumerable property values, in the same order as that provided by a for...in loop.

TypeScript which works

Object.values allowed me to reference this as an array in my Team class. So I changed return this.filter(member => member.enabled); to return Object.values(this).filter(member => member.enabled);. This made my code work at Runtime and gave me the necessary output as seen in this playground link.

Here's the Team class for comparison

...
class Team extends Array<TeamMember> {

    constructor(rows: Array<Array<string>>) {
        super();
        rows.filter(row => row[0] != "")
            .map(row => this.push(new TeamMember(row)));
    }

    getEnabledUsers() : Array<TeamMember> {
        return Object.values(this).filter(member => member.enabled);
    }
}
...

Is using Object.values() the correct thing to do?

Yes, using Object.values() fixed the runtime error. But it was extra computation and memory utilised to create an array from an already existing array. So IMHO, it seems like doubling the work.

What's the real error?

Thanks to my colleague Ynze who tracked down this FAQ answer from Microsoft on "Why doesn't extending built-ins like Error, Array, and Map work?".

So essentially, it's not supported in TypeScript for now due to how an Array for example uses ECMAScript 6's new.target to adjust the prototype chain.

TypeScript - Best Practice?

Speaking with another front-end developer colleague Ben resulted in him sharing the following code as better way of defining this Team class. You can see a working version on this playground link.

...
class Team {

    members = new Array<TeamMember>();

    constructor(rows: Array<Array<string>>) {
        rows.filter(row => row[0] != "")
            .map(row => this.members.push(new TeamMember(row)));
    }

    getEnabledUsers() : Array<TeamMember> {
        return this.members.filter(member => member.enabled);
    }
}
...

Verdict

TypeScript is not like C#. So proceed with caution 😉 Hopefully this article will help the next TypeScript noob!

Google Scripts Platform

I have been playing around with the Google Scripts Platform (GSP) both at home, for my parish during COVID lockdowns and at work. With the power of collecting data via Google Forms, storing data in Google Sheets and visualising data in either Google Sheets or Docs or even exported PDFs, GSP is pretty powerful. I wrote quite a bit of JavaScript to automate so many things. (Maybe I will blog about them gradually with some useful code snippets in the future.)

Why TypeScript?

So after writing C# for around 20 years, why did I want to learn TypeScript?

• Write JavaScript code that compiles
• Build and share some nifty GCP utilities
• Build a static website with Next.js
• Build a static website with Gatsby

The Transition

This is how I started on my journey from C# to JavaScript to TypeScript.

JavaScript to TypeScript

Coming from a very OOP (Object Oriented Programming) background, writing code in JavaScript felt a bit dirty. Don't get me wrong - JavaScript is a pretty powerful language and how it works on GSP is terrific. But I missed the strict types and compile-time error-checking capabilities of C#.

I heard about TypeScript a long time ago, but that was people complaining about how hard it was. But in recent times, that sentiment has changed and developers want to work with TypeScript.

TypeScript turned 10 years old as I completed learning the basics of it. I started off with learning TypeScript coming from a C# background and then moved on to the TypeScript Handbook.

I would recommend going through the handbook from end to end to grasp the concepts and the differences from other languages such as C#. Variable scoping and function overloading were a few things that I learnt along the way that were completely different from other programming languages I have used.

TypeScript in Practice

Once set up properly, the intellisense in an IDE (Integrated Development Environment) like Visual Studio Code makes writing and compiling TypeScript a breeze. In addition, using a unit testing framework such as Jest makes your code so much more robust.

Next steps

I want to quickly utilise TypeScript in frameworks like NextJS and Gatsby. But I wanted to complete something useful for others and utilise my newly acquired TypeScript knowledge before I went there. So you will see a few nifty little utilities I wrote on GSP with JavaScript converted to TypeScript.

You will also see some weird errors/issues (and my fixes to those) as I encounter them.

Wish me luck and let's see how I go!

Read the full blog post (9 min read) - Umbraco for Enterprise Clients: An agency perspective ↗️

This is a stub for my blog post on umbraco.com/blog ↗️

As my day job culminates with building websites on a CMS (Content Management System), I was also tempted to roll out an open-source CMS such as Umbraco and build my own blog on it. It's easy enough. But then I would need to host the CMS and take care of upgrades. So once again, I just continued on Blogger.com.

This was until I came across my former colleague Thai Tran's Blog. I was impressed by what I saw and as I scrolled to the bottom, I came across the Hashnode tagline.

Powered by Hashnode - a blogging community for software developers.

That sounded perfect. That was way back in June 2022.

Fast forward to 17 September 2022, I had ten days of annual leave booked in and a few tasks on my to-do list. One of which was to consolidate articles across different websites where I had written as a guest, MVP or employee.

I went back to Thai's Blog and scrolled down to the footer to remind myself what he was using.

I read the hashnode.com homepage and was convinced this was the way to go. Some of the features I really liked were:

• No Ads
• No Paywall
• Map your custom domain
• Free SSL
• Markdown for authoring

What I couldn't find information on was how Hashnode was monetized or its plans for monetisation. I had to dig a bit further to figure this one out. But Hashnode's co-founder Sandeep Panda says the following in this public question to the Hashnode team.

PRO features and monetization plans: Although we have raised sufficient funds to run Hashnode, we are looking to introduce a revenue model for long term viability. We have a few ideas in mind and I'll share them in a separate post.

So I signed up for an account and started to migrate my old blog posts from Google Blogger to Hashnode. While doing that, I thought it would be good to document my journey to inspire others as as I was inspired by Thai.

Read the full blog post (7 min read) - .NET Core 3.1 approaches end of life ↗️

This is a stub for my blog post on luminary.com/blog ↗️

Read the full blog post (5 min read) - Umbraco cDXP targets enterprise market ↗️

This is a stub for my blog post on luminary.com/blog ↗️

Read the full blog post (8 min read) - Why you should be modelling content ↗️

This is a stub for my blog post on luminary.com/blog ↗️

Read the full blog post (12 min read) - Scripting an Azure deployment of Umbraco 9 ↗️

Or listen to it on YouTube. Read by Paul Seal

This is a stub for my blog post on 24days.in/umbraco-cms/2021 ↗️

Read the full blog post (3 min read) - Luminary achieves Microsoft Cloud Platform Gold Partner status ↗️

This is a stub for my blog post on luminary.com/blog ↗️

Read the full blog post (7 min read) - Choosing a DXP: Kentico Xperience or Umbraco CMS ↗️

This is a stub for a blog post written by Andy and me on luminary.com/blog ↗️

Read the full blog post (18 min read) - How To Choose A Headless CMS ↗️

This is a stub for my article on smashingmagazine.com ↗️

Find out how to rate the digital experience maturity of your client and to guide them to the best solution.

The trailer

The trailer I posted on LinkedIn and Twitter before the talk.

The talk

This is the talk I gave at Umbraco Codegarden 2021.

Q and A (Unconference session)

This is the combined Q&A session for all the talks in that time slot

About Codegarden 2021

Umbraco Codegarden 2021 was a virtual tech conference held on June 9-11th 2021 and organised by Umbraco, the company behind the open-source Umbraco CMS. Codegarden is an annual conference held in Odense, Denmark and the home of Umbraco, but due to the corona pandemic was held virtually for the first time in 2021.

Over the course of 3 days, 2000+ attendees watched experts from all over the world, and all different areas of tech, come together to discuss everything new and upcoming with Umbraco, as well as the trends in the industry. Codegarden 2021 also introduced Unconference sessions for the first time - where speakers and hosts could get together to discuss the topics and ideas from the previous talks. Talks covered everything from accessibility to DevOps, best practices to career development, UX/Ul to Cloud hosting and SO much more.

Read the full blog post (1 min read) and watch the video - A/B testing, personalisation and marketing automation with Umbraco 8 ↗️

This is a stub for my blog post on luminary.com/blog ↗️

Read the full blog post(3 min read) - Packages in Umbraco 8 ↗️

This is a stub for my blog post on luminary.com/blog ↗️