 Oh, I've been trying to record this several times now, and that's been getting frustrating. It seems like there's something wrong with the SD card and the camera that I've been using. So I pulled out my older one, because I know that definitely works, especially since there's no SD card, it's just the main memory. So, yeah, frustrating. Two days ago, when I meant to first record this, I had put out the article for the F-sharp Advent. I'll have a link down in the video description for that. If you've watched the, at this point, the only F-sharp video I've done on this channel, you've seen some of the content in there. I do provide some additional information and it'd be worth reading, even if you've already watched that. There is quite a bit of additional information. The update, the tutorial video that I have planned for Stringer that I've kept mentioning, I've run into so many ridiculous snafus on that. Oh my God. So, originally, as I had it planned, it was supposed to use a database and do just simple crud stuff where the text was validated on the client side using Stringer. So, set up the database file, connect to it using Visual Studio, get the connection string from Visual Studio and put that directly in. Of course, everything works. Those of you who are more familiar with using databases than I am, because I don't really do much that involves persistent data, you know what the issue is there. I didn't, but you do. You can't do that because the connection string winds up being a fully qualified local path. So, if you try to move that onto a different machine, unless you have the exact same username and checked the repo out in the exact same location, it's not gonna work. Like, all right. There's gotta be ways of making this path relative. So, I look into that and from quite a few answers on Stack Overflow, it seems like it is. I couldn't get it working. I don't know if that's a case of me not getting the stuff right or Stack Overflow answers being wrong. Both of them happen quite often. So, I don't know. I just know that that's not working. But then, as I'm looking into it more, realize that there's a bit of an issue if you don't have the libraries for working with that specific type of database installed on your machine. Oh, and like, all right, this, the point of this really isn't the database. The point of this is just that there's some kind of persistent storage and you're using what I developed for the input validation. You don't care how it's stored. This isn't an actual business app. It's just showing off how to do the validation so we can do a different type of persistence. Okay, no, XML is one option, but oh my God, is that crufty and hard to work with and you'd have to design a whole schema and I just know, there's another thing that XML stuff will apply to JSON and other stuff which are better to work with but you still have to supply your own schema. There's another approach which is definitely a very niche thing. I'll have some stuff to say about that in a second, but file system as a database. The idea that you're using the directories and files specifically in a way that it works like a database. It's niche. Please do not do that willy-nilly. Even if you look into research around it, you'll find that it's very niche. The benefits that it's able to provide are definitely edge cases. The way you do it properly is not at all obvious. There's a lot of areas to trip yourself up and there's good reason why you don't see FAD commercial products as it's just so niche. But for this kind of thing, you're not really concerned about how the data is actually being stored. It's not a business app. You don't care if you lose data. You don't care if things get malformatted sometimes. That's not the point. It's demonstrating something else. This kind of thing would work perfectly since obviously file system paths can be relative. I've never seen an operating system where they can't. So that works. I go and implement that and the whole thing's working. I'm ready to go. I'm getting ready to record it and the announcement for the release of .NET Core 3.1 happens. Now, I'm not gonna forget to record my video but I might forget to do that update. So I do that update before I forget because it's not gonna take terribly long. Go to record my video and it just stops working. What in the fuck? So I don't know if that means that there's a bug that was introduced in .NET Core 3.1. The runtime, I didn't update the project dependency but the runtime could have introduced a bug. I find that unlikely though. The other possibility is that I was depending on behavior that should not have been working and only was working due to a bug in the previous runtime. We're dealing with file operations and not any special kind of file operations. So I find that unlikely. But regardless, it's not working. I'm getting these exceptions about how a process is already accessing the file when I go to access the file and I can restart the machine and all sorts of other stuff to make sure nothing's happening. I can spin up a VM that is completely brand new that has nothing going on in it. Clone the repo into there and it says the same thing. What the actual hell? So I've been trying to figure out what is going on there for a while now and just yesterday towards last night I finally figured it out and it is so frickin' stupid. Remember how I said FAD has, it's very niche. It has a lot of pitfalls. There's a lot of things about it that are a bit weird. One of those things is that you cannot, under any circumstances, operate the thing with a debugger attached. What happened is I know that. So when I was doing the development originally, I made sure that it was always in release mode so that any time I went to run the stuff, the debugger would not attach and I could go in and check the files manually and it worked. It's a bit tedious but it works. I winds up not being as big of a deal as it might sound in practice because unit tests, make sure you have a lot of unit tests when doing it but again, it's very niche thing. Don't do it unless it's like a research thing. What are you doing? But what I realized is when I went to do the update, I sort of reset everything when I went to open up my projects again. It went back to debug mode which is the default. So I didn't notice that and kept trying to run the things with the fricking debugger attached. So yeah, it works now. I made sure that you can put it on different paths within the machine, on different machines, all sorts of shit as long as you do not run it in debug mode, it works. So I'll have a video on that fairly soon, especially since it works. I have some stuff to take care of today so I don't know if it can be done today but I will get that out as soon as I can. That's pretty much it, this little update. Again, if you are interested in that article for the F sharp advent, it's down in the video description. That's yeah, have a good one.