A Jon Skeet Meetup retrospective

This week we hosted Jon Skeet at the Adelaide .NET User Group. Jon demonstrated some of the new language features coming in C# 8, and it was probably the biggest attendance we've had in a really, really long time.

Because we had so many registrations, I lined up some extra help with my two oldest kids (conveniently on school holidays). They helped set up the room, liaised with the pizza delivery guy, and helped pack up everything. I think they even found a few things familiar with Jon's use of Fibonacci sequences in one example and similarity of programming language features (my eldest daughter has been doing some Python coding at school).

Jon lives in the UK, and we're in Australia, so whilst we would have loved to have Jon in person, the next best thing was to have him present remotely. We've had remote presentations before using Google Hangouts, but this time I opted to use Skype (and made use of the new recording feature).
I don't use Skype a lot, but we got the call up and running without too much difficulty.

I set up a webcam in the meeting room, along with a boundary microphone (an MXL AC404 USB Conference Microphone). The intention is that the presenter can see the audience, and the boundary microphone allows people to comment and ask questions from a fair way away (eg. right at the back of the room) and still be heard. This seemed to work pretty well - people up the back were able to ask questions and Jon seemed to hear them ok. Jon's video feed was pretty good. I think our feed back might have been a bit jumpy, but for the most part I think it was ok.

I left it to the last minute to arrange for access to the WiFi network at our venue. I ended up using my phone's 4G data for the call, which worked well. It was only after we'd finished that I discovered that an email had come through just before the start of the meeting with the WiFi details. At least I've got them for next time.

We also picked up a meeting sponsor this month in Simon Cook from Encode Talent Management. Being able to not charge attendees (to cover the cost of pizza) was great, and hopefully this relationship might continue in the future.

Skype notes:

• The recording doesn't include the extra buttons/overlays/controls (which is good)
• The recording will record a split-screen of web cams if they're both active. If you just want to have one webcam in the recording then you'll need to disable the other one. When screen sharing, then just that is recorded.
• Taking a 'snapshot' follows similar rules as for recording. eg. a snapshot would include both webcams if they're active, or just the shared screen if one is being shared.
• If the presenting person isn't using a headset, then to avoid echo it's probably best to mute the microphone (and just un-mute it for questions)
• Making the overlay controls slide out of the way was inconsistent. Sometimes moving the mouse off the screen worked, but one time it didn't. Would be good to figure this out.
• F11 makes Skype go full screen
• Skype now supports NDI. This means it should be able to talk to software like OBS Studio if you wanted more control over broadcasting or recording. I don't know if the feed it sends is just the video/shared screen or if that also includes the control overlays etc.

Other notes:

• Our regular room can hold a decent crowd if necessary.
• People will eat as much pizza as there is available to eat
• People don't drink much water (but this might change if it was warmer weather)
• Organise wifi access earlier!
• Windows 10 has simple video editing via the Photos app

The recording of Jon's talk is up on YouTube. I won't be giving up my day job to become a YouTube broadcaster anytime soon, but it's nice to have a record of a great presentation.

Speaking at .NET Conf - Put your C#, VB and F# projects and packaging on a diet

I'm really exited to be selected as one of the community speakers for .NET Conf

.NET Conf is a free “virtual” conference organised by Microsoft and the .NET developer community that is streamed live around the world. Being virtual, it means organising travel and accommodation is remarkably easy!

My talk is titled “Put your C#, VB and F# projects and packaging on a diet”, drilling in to the new project system for .NET, and how you can use it even with old projects that target .NET Framework and it starts at 04:00 UTC on Friday 14th September (check local times). Go to https://www.dotnetconf.net/ to watch the live stream.

All the demos from my talk and links to other resources can be found in the Github repo https://github.com/flcdrg/project-system-diet

Converting a SQL Server .bacpac to a .dacpac

Microsoft SQL Server has two related portable file formats - the DACPAC and the BACPAC. Quoting Data-tier Applications:

A DAC is a self-contained unit of SQL Server database deployment that enables data-tier developers and database administrators to package SQL Server objects into a portable artifact called a DAC package, also known as a DACPAC.
A BACPAC is a related artifact that encapsulates the database schema as well as the data stored in the database.

When they say related, they're not kidding! Both of these files formats are based on the Open Packaging Conventions (a fancy way of saying it's a .zip file with some other bits), and cracking them open you discover that a bacpac file is basically a dacpac with a few extra files and a couple of different settings. Knowing this, it should be possible to manually convert a bacpac to a dacpac.

First, unzip the .bacpac file (using 7-zip, or rename to .zip and use Windows File Explorer’s Extract Archive).

Now do the following actions (you could do these programmatically if this is something you need to do repeatedly):

1. Edit model.xml
   
   1. Change [email protected] to 2.4
2. Edit Origin.xml
   
   1. Change ContainsExportedData to false
   2. Change ModelSchemaVersion to 2.4
   3. Remove ExportStatistics
   4. Recalculate the SHA256 checksum for model.xml and update the value stored in Checksums/[email protected]=’/model.xml’
3. Remove directories _rels and Data

Now re-zip up the remaining files and change the file suffix back to .dacpac

To verify that the .dacpac is valid, try using SSMS with the Upgrade Data-tier Application wizard. Run it against any database and if you can proceed to without error to the "Review Upgrade Plan" step, you should be good to go.

Create a temporary file with a custom extension in PowerShell

Just a quick thing I wanted to record for posterity. The trick is using the -PassThru parameter with the Rename-Item cmdlet so that this ends up a one-liner thanks to PowerShell's pipeline:

$tempNuspec = Get-ChildItem ([IO.Path]::GetTempFileName()) | Rename-Item -NewName { [IO.Path]::ChangeExtension($_, ".nuspec") } -PassThru

Microsoft LifeCam Studio stops working with Windows 10

I have a Microsoft LiveCam Studio webcam that I bought a few years ago for the Adelaide .NET User Group for when we have remote presenters. It's been pretty good (although not long after I bought it, Scott Hanselman tweeted that actually the Logitech 930e was worth considering with possibly a wider shot).

I went to use it the other day, and it just plain refused to work. My laptop has a builtin webcam and that was showing up, but using any app (eg. Microsoft Teams or the Windows Camera app) just wasn't showing the LifeCam. It was strange as it did show up as an audio device, but not video.

I brought up Device Manager, and looked in the Cameras node, but it wasn't there. I tried unplugging it and re-plugging back in (and rebooting Windows) to no avail.

I then tried the webcam with a different PC, and it worked, so at least I knew the device wasn't faulty. Firing up Device Manager on the second PC revealed something interesting though. The LifeCam wasn't under Cameras, it was listed under Imaging devices. Who would have guessed!

Switching back to my laptop, in Device Manager, I went to the View menu and selected Show hidden devices. Looking under the Imaging devices revealed something unexpected. There were two device drivers listed for the LifeCam! I right-clicked on both devices and selected Uninstall device.

I then plugged the webcam back into the laptop, and now Windows registered that a new device was attached and indicated it was installing the device drivers. After a short wait, it was now working correctly!

Mystery solved 😁

Creating VSTS Service Hooks with PowerShell using the REST API

Service Hooks are Visual Studio Team Services way of integrating with other web applications by automatically sending them events when specific things happen in VSTS, like a build completes or code is committed.

These are what I used in my earlier post about integrating VSTS with TeamCity. If you just have one service hook to set up then using the web UI is fine, but if you find yourself doing something again and again then finding a way to automate it can be really useful.

Interacting with service hooks via VSTS REST API is documented here. Web Hooks are a particular service hook 'consumer' suitable for sending HTTP messages to any web endpoint.

I'm going to create a PowerShell script which requires the following parameters

Param(
   [string]$vstsAccount,
   [string]$projectName,
   [string]$repositoryName,
   [string]$token
)

Using the VSTS APIs requires authentication, so the first thing is to encode a Personal Access Token (PAT) so it can be set as a HTTP header. (You create PATs from the Web UI by clicking on your profile picture and selecting Security)

$base64AuthInfo = [Convert]::ToBase64String([Text.Encoding]::ASCII.GetBytes(("{0}:{1}" -f "",$token)))

There's a whole lot of VSTS events that you can choose as a trigger when creating a service hook. In this example, I'm interested in being notified when a Git pull request is created. In order to use this particular API, I need to also know the ids of the VSTS Project and Repository that I want this service hook associated with. I'll use API calls to find those out.

$uri = "https://$($vstsAccount).visualstudio.com/_apis/projects?api-version=5.0-preview.1"
$result = Invoke-RestMethod -Uri $uri -Method Get -ContentType "application/json" -Headers @{Authorization=("Basic {0}" -f $base64AuthInfo)}

$projectId = $result.value | Where-Object { $_.name -eq $projectName } | Select-Object -ExpandProperty id

$uri = "https://$($vstsAccount).visualstudio.com/_apis/git/repositories?api-version=5.0-preview.1"
$result = Invoke-RestMethod -Uri $uri -Method Get -ContentType "application/json" -Headers @{Authorization=("Basic {0}" -f $base64AuthInfo)}

$repositoryId = $result.value | Where-Object { $_.name -eq "$repositoryName" } | Select-Object -ExpandProperty id

As far as the REST API is concerned, we're creating a new 'subscription'. Create operations use a POST and usually also require JSON data to be sent in the body. We'll use PowerShell to model the data and then convert it back to a JSON string:

$body = @{
    "publisherId" = "tfs"
    "eventType" = "git.pullrequest.created"
    "resourceVersion" = "1.0"
    "consumerId" = "webHooks"
    "consumerActionId" = "httpRequest"
    "publisherInputs" = @{
        "projectId" = $projectId
        "repository" = $repositoryId
        "branch" = ""
        "pullrequestCreatedBy" = ""
        "pullrequestReviewersContains" = ""
    }
    "consumerInputs" = @{
        "url" = "https://servicetonotify"
        "basicAuthUsername" = ""
        "basicAuthPassword" = ""
        "resourceDetailsToSend" = "all"
        "messagesToSend" = "none"
        "detailedMessagesToSend" = "none"
    }
}

$bodyJson = $body | ConvertTo-Json

Obviously you will need to customise the url value to point to your particular web service that should be notified. If that service requires authentication, you can supply a username and password in the basicAuthUsername and basicAuthPassword values. You can also control what detailed information VSTS will send to by setting the three *ToSend values. In my case I only needed resourceDetailsToSend but not the other two.

$uri = "https://$($vstsAccount).visualstudio.com/_apis/hooks/subscriptions?api-version=5.0-preview.1"
Invoke-RestMethod -Uri $uri -Method Post -ContentType "application/json" -Headers @{Authorization=("Basic {0}" -f $base64AuthInfo)} -Body $bodyJson

Using the VSTS REST API is pretty straight forward, and gives you great access to query and modify your VSTS environment.

Boxstarter and Chocolatey tips

Two big things happened earlier this year on the Chocolatey front. First off, Boxstarter (the tool created by Matt Wrock that allows you to script up full Windows installations including handling reboots) is now being managed by Chocolatey. Boxstarter.org still exists, but the source repository is now under the Chocolatey org on Github.

The second is that Microsoft are contributing Boxstarter scripts in a new Github repo – https://github.com/Microsoft/windows-dev-box-setup-scripts

If you’re looking to use Boxstarter to automate the software installation of your Windows machines, there’s a few tricks and traps worth knowing about.

Avoid MAXPATH errors

It’s worth understanding that Boxstarter embeds its own copy of Chocolatey and uses that rather than choco.exe. Due to some compatibility issues Boxstarter currently needs to embed an older version of Chocolatey. That particular version does have one known bug where the temp directory Chocolatey uses to download binaries goes one directory deeper each install. Not a problem in isolation, but when you’re installing a lot of packages all at once, you soon hit the old Windows MAXPATH limit.
A workaround is described in the bug report – essentially using the --cache-location argument to override where downloads are saved. The trick here is that you need to use this on all choco calls in your Boxstarter script – even for things like choco pin. Forget those and you still may experience the MAXPATH problem.

To make it easier, I add the following lines to the top of my Boxstarter scripts

New-Item -Path "$env:userprofile\AppData\Local\ChocoCache" -ItemType directory -Force | Out-Null
$common = "--cacheLocation=`"$env:userprofile\AppData\Local\ChocoCache`""

And then I can just append $common to each choco statement. eg.

cinst nodejs $common
cinst visualstudiocode $common 
choco pin add -n=visualstudiocode $common

Avoid unexpected reboots

Detecting and handling reboots is one of the great things about Boxstarter. You can read more in the docs, but one thing to keep in mind is it isn’t perfect. If a reboot is initiated without Boxstarter being aware of it, then it can’t do its thing to restart and continue.

One command I’ve found that can cause this is using Enable-WindowsOptionalFeature. If the feature you’re turning on needs a restart, then Boxstarter won’t resume afterwards. The workaround here is to leverage Chocolatey’s support for the windowsfeatures source. So instead of this

Enable-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V-All

Do this

choco install Microsoft-Hyper-V-All -source windowsfeatures $common

Logging

If you have a more intricate Boxstarter script, you may run into some problems that you need to diagnose. Don’t look in the usual Chocolatey.log as you won’t see anything there. Boxstarter logs all output to its own log, which by default ends up in $env:LocalAppData\Boxstarter\Boxstarter.log. This becomes even more useful when you consider that Boxstarter may automatically restart your machine multiple times, so having a persistent record of what happened is invaluable.
The other things you might want to make use of is Boxstarter-specific commands like Write-BoxstarterMessage (which writes to the log file as well as the console output) and Log-BoxstarterMessage (which just write to the log file)

Find out more about these and other logging commands by running help about_boxstarter_logging.

My scripts

I keep a few of my Boxstarter scripts at https://gist.github.com/flcdrg/87802af4c92527eb8a30. Feel free to have a look and borrow them if they look useful.

Find out more

If you’re really getting in to Chocolatey and Boxstarter, you might also be interested in Chocolatey Fest, a conference focusing on Windows automation being held San Francisco on October 8th.