How'd you like to listen to dot net rocks with no ads? Easy? Become a patron for just five dollars a month. You get access to a private RSS feed where all the shows have no ads. Twenty dollars a month, we'll get you that and a special dot net Rocks patron mug. Sign up now at patreon dot dot NetRocks dot com. Hey guess what it's dot net Rocks twenty twenty five edition. I'm Carl Franklin, I'm Richard Campbell. Happy New Year everybody, and what's how did you? How'd you do your New Year's?
We did New Year's up on the coast because we went down to the city for Christmas, right, spend time with the girls. The youngest is pregnant, so called the baby bump. Christmas very nice and uh, and then New Year's we came back up. I had a party in the boathouse with a little mixing of the worlds. You know, we're building friendship groups up on the coast, so a few of those couples showed up and a few from back on the mainland.
And yeah, I saw the pictures on Facebook. Looked a lot of fun. Yeah, you know, the best thing you can hope for in the new year. Belly laughs, like laughing so hard you're crying. That's what we did overde That's great. Yeah, awesome. Well we just had a quiet New Year's here. It was nice, you know, just the fam. It's always good. Yeah. So let's start this show with a better know of framework, which is interestingly about last year.
Roll the music. Interesting, man, what do you got?
It's a video by the Visual Studio team our favorite new visual studio features of twenty twenty four. So they took a look at all the things that were added a visual studio in twenty twenty a lot.
Yeah, right, Like it's constant.
Yeah, And you know, if you're like most of us, you don't know about these new features until somebody says, hey, you know Scott Hanselman says, hey, do.
You know this?
Yeah, and then you look it up.
People make whole talks about this, just showing features in studio you didn't know about because it's just so darn many of them. Right.
Yeah. It's because the word of coding, hasn't it the Microsoft word?
Yeah?
Yeah, Well, anyway, it's a good video and if you have how long is it? Well, Anyway, it's a good video and if you got about twenty minutes to kill check it out. It's good stuff. So that's what I got. Know it, learn it, love it. Who's talking to is?
Richard grabbed a comment of a show nineteen sixteen then when we recorded with our friend Rendel called how simple is as simple as possible? Yeah? And another friend of ours, Richard ra Chima, also known as codeputer, had a good comment here, mostly about Blazer, where he said simplicity and web application development it's not a simple conversation Blazer server and transitioning to WASM is also not simple. So why are we doing web app development anyway? App distribution to
maintenance is why we wanted it. Web apps over the client side app installs. Yeah, definitely always running the current version, don't need to deploy running in a sandbox like there's good things makes sense. Yeah, So we leveraged the distribution model of the browser and then got busy building apps, this time without easy acts to what we had before the mistley client side resources. Nothing that fifteen megs of JavaScript can't fix with a sprinkless CSS to really challenge that street.
He's our age too, so he's talking experience here.
Chris Love would have words with you. Oh yes, So let's take application development, make the super puke it all into a browser box and give it a solid stir for fun, while dumping in new app features or ingredients, more JavaScript frameworks and such as the goodness and all the joy of distribution. HTML was never supposed to be used as a U as a user experience interface for application development amidst the formatting and presentation of electronic documents.
It was about scientific papers with iperlinking.
Right.
Then there's Blazer server. Let's not worry about WASM for a now, just about using the server environment to create HTML documents that are rented. A browser does not even need a page refresh, It builds the page. Choosing marshal or components. Click in the browser is meaning less without the context of where it happened. A keystroke has the same requirements in both cases. The event requires context. That is simple. HTML is simple, but stupid. Why are we
making it more complicated? Somebody says server overload. Server load for service side rendering using cloud resources not the same as a web farmer's service. True applications never used servers that use service is provided by servers. Your application needs hosting resources from the cloud. Why are we thinking about servers? And once again, Blazer server framework is nothing more than moving the context of the client side event and exposing
that in the context of the application. Mix it an application code, update the component of the dock, and render just tailored bytes back to the document being viewed in the client. It's nice, simple and with a minimal byte count to transport. And that's my definition of web simplicity. Yeah, I can't argue that. Shame we make such a messment.
I've seen the light.
Yeah, he's trying to anyway. Yeah, i'd rich thanks so much for your comment. I'd send you a copy of music Code by but you got one. Yeah.
He's going to be a guest in a couple of weeks at fall as well.
And he's going to be a guest to shortly too, so we'll be here from anyway. I'll tease them the more.
Yeah, very good.
And if you'd like a copy of music Code, I write a comment on the website at dot NetRocks dot com or on the facebooks. We publish every show there, and if you comment there and I read it on the show, we'll send your copy. Music code by.
And music to code by, of course, is a series of twenty four minute twenty five minute music pieces that you can listen to while you code that helps you focus and keep you from getting distracted. And if you want to pay for it, it's at music tocode by dot net. Also, you could get a copy by sending us a message on the social media's other than facebooks. We've been on ex Twitter for the longest time. We're also on masted On and Blue Sky and some variation at Carl Franklin and at rich Campbell.
Right right, yeah, blue Sky seems to have all the action these days. Man, Like I realized I was. I was scrolling through Blue Sky today and it looks so much like the old Twitter. I forgot I was on blues guy.
Yeah, except it's not dumb.
Well, it's kind of the like the old Twitter.
Yeah yeah, all right, let's bring on our guests. Nathan Westfall. He is a senior no history. Oh you know what I so, here's the thing I looked at nineteen thirty five in the history and it's so jam packed, like we can't even cover it all. But maybe I'll ask you Richard what is your.
I only pick one thing, you know me, I'm gonna pick one thing. Yeah, and the one thing is James Chadwick's proof that the neutron exists.
How amazing.
Yeah, well think about this. That's nineteen thirty five and within ten years the first atomic bomb would go off. Yeah, and you need neutrons like that's how fast?
Yeah, So it's amazing the things that happened in this relatively short time period. It was just explosive.
Yeah, literally in so many different ways, right like between between the Great Depression, the rise of fascism, the explosion into World War two, like that was a tumultuous ten plus years and it just happened to be doing shows in those show numbers right now.
And a byproduct was a lot of innovation, a by product of the war and all of that stuff.
Yeah. Well, for better or worse, the military spends a lot and a lot of that funding was military.
Highway system because of the military fifties in America.
We got a little time yet for that. Yeah, it'll be later this year.
Yeah, yeah.
Yeah.
Still, it was Eisenhower's idea because he wanted mobility. Yeah, all right, enough of that. Yeah, that's coming, Let's bring on Nathan Westfall. As I started to say before, he is a senior software engineering manager at Tyler Technologies, leading the Cloud operations and mobile development team for student transportation. He started development journey with Php and Ruby on rails, but fell in love with the dot net ecosystem once
leaving college early for his first job outside development. You can find him mixing audio at your local bar or fishing on a frozen lake. One of those things.
I do all the time.
I wonder which one it's not fishing.
Welcome, Yeah, thanks for having me.
So you've been listening to the show for a while, you said.
I have been. Yeah. Yeah, kind of the way I learn is podcast everything else, and you guys are fun to listen to, So that's good. It's kind of as Richard was saying earlier. It's nice to now see how the sausage is made.
So yeah, well it's good to have you at the table. Nice past the sausage.
So what have you been working on, Nathan.
Yeah, so Tyler Technologies. What I work on in student transportation is you know a yellow school bus you see on the road in the US and Canada. You know, we interact with so we're in one of every four
school districts uses Tyler system. And the app that I here to talk about is the public facing app, so the one that any parent who has a kid on the bus that we integrate with can use called myrid K twelve And it's an app that we rewrote earlier last year, so in the beginning of twenty twenty four it started to roll out, and it is a Blazer Maui hybrid application works on web iOS, Android, and it is much Blazer as you could possibly imagine shoved into
this app to have it run at scale and everything else.
I've already got a man crush on you. Man that is that is you're speaking my language here.
What was it before it was Blazer?
Yeah, good question. So it was examinant Forms. The original version of this app actually came out a couple months after dot net one, so dot net Core one was released. That's how early we were to dot net core and everything else. And what we learned early on was I can do examer reforms pretty well. But then when I have to make those same features on web in dot net core and Angular JS at the time, my feature
drift became insane, right. It kind of became worse and worse over time to try to do something in forms for iOS, maybe something a little different for Android, and then also then change languages entirely for Web. So we got the opportunity to rewrite it in twenty twenty three, released in twenty twenty four, and went full blazer, so we could share about ninety five percent of our code between iOS, Android, Web is all the same.
That's fantastic.
Yeah. What also means you battle through the tough times at zamorin too, right, because that was oh yeah, those were early bits.
Yeah, correct, Yeah, when I even used examined it before I joined Tyler, when Microsoft had only had acquired it, like very very recently. So I mean I've been using examined for almost as long as it's been help.
Yeah.
We I thought it was going to be a lot smoother ride than it actually turned out to be. I was so excited when when Zamoran you know, came on the scene. Yeah, and the promise of jam forms and all that stuff was just amazing and then just just warm me down.
Sorry, yeah, Well, and I wrestle with is it the problem with the tool learning? Is it just a problem with building mobile apps like this separation between working on a machine different from the day app with the app environment like that, cycle's just hard and it's diverse like that, that combination just doesn't matter what the tools are. It's a battle compared to any other.
It is, and whether the tools keep you up to date or you have to keep up to date, you're always going to end up doing some kind of tweaking on one platform you're not doing on the other. And that's why I think the road that you took, which is Blazer Hybrid, gets rid of all that because the cross platform this is done in the browser, which is had many many years to go to get its cross platform story done, and it's there.
No, I'm with you. I think the biggest part of Exhibit Forms that was tough is even that you had all this shared code and c sharp code at the time. I'm fighting iOS at the time, you know, releases were in June. Android was later in the year. I'd have to update all my Android dependencies just so forms would work right. And to your point, Karl, if I'm just using a browser which I know will render my CSS the same way, my dependency is just shrunk. And you know, so small.
So yeah, so if you can do a web app in Blazer, you can do a mobile app in Blazer, right, yep, that's what it comes down to.
And we're testing it both because we even went the whole route. We're not using Blazer server at all. We went full web assembly. Yeah, I went the full If I don't have to host the server, why would I? So, yeah, sure that's what we're doing. Yeah.
Well, and you've got a I mean, it's not really public facing app in the sense that it is a known set of users. It's parents of students, but it's still a lot like the server load on that. I think Blazer server makes a lot of sense with an internal app with a certain number of employees done. Yeah, to make that public facing it's really hazardous.
Yeah, yeah, And our traffic's really predictable, right, Like we can see different time zones come online because buses normally roll in the same morning the same time East Coast and then Central and then west.
Right, right, you know when it's going to be busy. Right.
Parents are going to check the f twice a day, when their kids going to school and when their kids coming home, and everyone's doing it at the same time.
Almost, so aren't you then making an argument in favor of server sign.
Ah kind of you know, but the costs at the end of the day for web assembly to support all of this is I have we're partnered with AWS here at Tyler, so we're using dot net on AWS. But the whole web server for hundreds of thousands of parents is an S three bucket and then a CDN and I'm done.
That's it.
That's all I got.
Yeah, that's great.
Nice.
So I got to ask, what are you using for authentication?
Cognito Actually, huh, tell.
Me about Cognito.
Yeah, so Cognito. We we're going back and forth what we should use. There's a whole bunch of obviously authentication providers you can use Cognito. We just wanted to stay within the Amazon ecosystem because we don't have huge requirements for authentication. They have a nice portal for users self registering, which is our big use case. It's not like we create these accounts for parents. They can self register and
give us the option to do social logins down the road. So, you know, we're over six hundred and fifty thousand users that have been created in Cognito, and then even so Cognitio create a nice pipeline because we were not in cognito common examinineforms and they let you actually in real time, So once a user tries to log in for the first time, we can migrate those users in real time. So no password change, no nothing.
If you're going to run on AWS, you should use AWS as authenticator, just the same way that if you're going to run on Azure you should be using Azure AD, right.
Like, yeah, I'm a big fan of peeping it together.
Yeah yeah, yeah, it's just not worth the pain. Yeah, Like, it's very challenging to build a unified authentication strategy across clouds, fighting that a lot, right.
And so I imagine that all the credentials are kept out of your reach, Yeah, so that you don't you're not liable for you know, break ins or anything like that.
Correct, which is another big benefit. We used to host it ourselves with dot Net's identity server and handled all that and that's all gone now and all everyone got migrated over and again that real time is nice because you can use landa when they log in, Oh, this user doesn't exist, let me check your existing user store and the user never had to change their password. It just moves right over and that's great. Nice starts working.
So you did that, very nice migration. They never knew you changed authentication strategies. Doesn't matter.
Yeah, no idea.
Then and then one day nobody's hitting that identity server anymore. You can turn it off one day.
Correct, Well that was I had the pleasure of a couple of months ago finally turning everything off, and it felt so nice to go through and web servers off, APIs off, everything got shut down.
Why do we feel good when we turn stuff off?
It's like calming, right, Yeah, No, it's.
The money goes back into your wallet, reversing the flow of money.
It's also it's also an end of pain, right, Like, that's weary. It's not going to hurt you anymore because it's the stuff.
I know. There was a there's a brief period where we were rewriting it, and I always hated getting a you know, an issue come through and I'm like, I got to open examin forms again and I got to go through and fix this. And after using Blazer for so long, rewriting it's it was hard to go back.
Oh yeah, wandering back in his zamm reforms like walking into a dark forest. Yeah, I remember why I left this place. I appreciate that.
I it's on my list of things to do to open up uh MAUI in dot net nine and take a good hard look at it.
Again.
I don't because ever since I found you know, the Blazer hybrid MEI hybrid stuff, I'm I just don't. I haven't gone back.
Yeah, basically, so yeah, we've done. We've done a few to play around with it, but I every time I do it field and need of like I could just do this and blaze and I'm used to it and it's works just as good.
Right yeah, yeah, well yeah, if you're getting the results, if the customer is happy, which is parents, which are who are never happy, then you're doing all right right right.
Yeah, you can definitely say that again.
So yes, yeah, yeah, don't mess with it.
And how many users did you say you had?
We currently have over six hundred and fifty thousand uh registered in our active counts daily active counts prett high on the top of that.
So wow, you do not want to do tech support for six hundred fifty thousand people. That's not good. No, they got to be all self service, They got it, but also they don't care. They just want to know where the kid is.
Correct, which is a whole other set of problems that we had to do in a way that's helping both the district, who is really our customer, right, The school district is the one that uses our technology.
Those the ones you're paying you, right, But.
They're servicing the parents, and we also have to serve the parents and the district. So how do I explain to the parent, hey, this bus is out of commission, right, and this is why you can't see where your kid is, right, and solving all these problems to someone who's not in the operation of school busing, right, making them not come
back at us. Obviously, we want to protect our operatings as much as possible and be like, hey, your app's not working, But at the end of the day, it's you know, your bus didn't drive today because it's a holiday. True story, right.
So is that what your app does primarily as track buses?
Yeah, so it does two different things. So you can find your kids' schedule all through self registration. So if I had my kid was in elementary school, I could register them to the app and see, hey, the bus is going to pick them up at this corner every Monday through Friday at eight twenty in the morning. This is my driver's name and what bus is coming. And then if the district allows it and has the capability, you can actually see when your bus is coming in
your real time. You get notifications kind of like Huber if your bus is about to arrive, it's five minutes away, it's running late, and kind of provides that experience. So parents can just really quickly say, I mean, it was a cold day up here in New York. I was outside and realize how windy it was. I would hate if I had to go stand at the bus stop for twenty minutes hoping the bus would show up. No, it's nice if I can get that notification just walk outside.
What about kids fourteen and older, Let's say, can they use the app as well?
They technically can. We obviously try to restrict it as much as possible as it's meant for the parent, But there's parents have the ability to share that information with any other account, right, so they can even say, hey, for the day, the babysitter can have it to make sure that my kid gets off the bus.
Nice the kid wouldn't necessarily be waiting at the bus and it's late and be able to flip up the app and say, oh, yeah it's ten minutes late.
Yeah, we have some kids that do it, but primarily it's the older kids are kind of left on their own. As you know as a parent, they kind of you're becoming more of an adult. It's a lot of kids in elementary school, right, because even in some states, depending on your age, you have to be at the stop for the kids to get dropped off. True, so I don't want to wait outside.
Plus there's the Kappa thing, right, Yes, the Kapa against kids using software. If you were going to allow kids I think it's fourteen and younger or under fourteen.
Yeah, it's somewhere like that. Thirteen fourteen, Yeah.
Thirteen and under, if you're going to allow them to use it.
Not that there's any sensitive information here, it's the location of a bus.
Yeah, you'd have to go through hoops yet that Yeah, And.
We happen to be recording this when there was a pretty big snowfall through central US. So I got to think that impacts bus as a whole bunch. And that's when everybody's going to be hammering on the app, like, have we seen the maps?
Correct?
He's the bus running?
Is it?
Is it moving?
Have you seen the maps? Like this white band all the way through the Midwest from the from the west coast to the east coast. Unbelievable. Yeah, quite a piece of weather anyway.
So do you present it as a map or is it just a table of this is the bus, this is where it'll be?
Yeah?
No, good question time. So it's it's all configured by the district because different districts have different policies. And you know, some districts have a policy where they don't want to show the parents the bus just don't it's about to arrive soon because they don't want angry parents getting in their car and chasing the bus down because they miss their kid and so on and so forth. Right, But if they allow everything, Yeah, you can see your bus on a map. We use signal R to broadcast that
information and then show it live within the app. You can interact with it, follow it in three D all this other fun stuff.
Yeah. I like that. Yeah, I know that's yeah, so you can go but very that's very like uber esque in that sense. You can literally watch the vehicle emotion. But I can also see why certain districts would say, you know, that's probably a little too much information, that's a creepy Yes, exactly.
You get both extremes. I want to see everything and I want to see nothing, right.
Yeah.
The Uber thing is is when we've been trying to deal with for a while. Because Uber is obviously great, right. I mean I've gone to the airport so many times and I go to a conference, I'll pull up Uber get where I need to go. But people don't quite understand all the time, is that Uber is nice because you as a driver, except that I'm carrying my one passenger somewhere else. While a school bus is if it's all configured properly, because the driver can't do any interaction
while they're driving. They can't have a screen up, they can't have all this other technology that a taxi driver UBI driver can have. I can't just notify one parent. I have to notify sixty kids maybe and their families, and not also share where those fifty nine other kids also live, right? And how do I do that in a different way?
Challenge?
Yeah, No, it's great, and it's great. Question. It's like what's the right thing to do there?
So you have let's count the clients. You have a web obviously and mobile. Do you also have Windows and Mac clients.
We don't have Windows because actually all of the development team, including myself, we only run max. I do have it running on a Mac locally, though, because that's how I do development is mostly in the Mac.
Right okay, because you're using the Mali hybrid model.
Right, correct, there's nothing stopping us.
It's more uh, nothing stopping you.
Yeah, it's more because of you know, got to go through enterprise hoops. I don't want to fight that, yeah, exactly.
I mean that comes back to Richard Rukima's point is that we love these mobile things because you can just pull them up, you get the latest version, you don't have to install them and ye WPF who needs it? For this guy?
Apps that? Yeah? Yeah, that's you don't want to run WPF on now? That would be gosh.
Did you also take a look at some of the other sort of you know, multi client frameworks out there before you decided on this or was it a natural progression because you were always using zamor and already.
Yeah, it was obviously we were leaning this way. We were using Blazer already, Uh since what was it done in a core three three one depending on whethery're using web assembly or not, so we kind of were already
familiar with Blazer. Tyler also has a internal UI library that we use, and of course it's web components only, so it's do I want to port that something else or keep using web We did look at things like Flutter and Ionic for a little while, but really we're a dot net shop through and through, even outside of mobile development, so it kind of just made sense for
us to stick with dot net. And I've never had any issues with Blazer, which I know people have their opinions on necessarily using web assembly and all this other stuff, but it's at least I know what can be wrong, so it's familiar.
Yeah, But I mean, MAUI was supposed to be the logical successor to zamorin forms like you are making a bigger leap here to jump over to Blazer.
We are, and we definitely walked through that. So we've actually had two apps and we did them differently. So ride through sixty now known my ride K twelve, we did that whole Blazer rewrite, right, We took things in and forms through the way it did everything a blazer. We have another application called my Stuff, which is for a different application that we support kind of same functionality, but for a different on prem solution basically uh bus srouting.
That one went from examer Informs to Maui. There was no blaze or anything we write. That one went through what Microsoft wants you to do with progressing straight over to Maui and it because of how that app operated. It didn't have anything overly complex. It wasn't difficult, but it definitely was not as easy as it is made out to be.
Well, both products had their challenges, right, Yes, Yeah, and I would argue if you still had a Zamora Informs app today moving to Maui on done at nine, he's a heck how a lot less painful? Totally agree. Yeah, definitely if you jump when they wanted you to jump, you jumped into a pifle of glass like.
Correct, correct? Yeah, we only did that migration from Forms to Maui within the past six months, right, so when all that stuff got worked out and learning curves and all that.
Yeah, if you can wait, wait you it will. Yeah, it's a benefit without it's like, if your app's not currently on fire and people aren't going to kill you, you know, stay with what you got and let the bugs be worked up by somebody else.
Yep.
Well, and also you know, you don't really he's not really going from Maui to Blazer. He's using Blazer inside a Maui app, So so it is Maui Blazer essentially. But I was thinking of something like Avalonia. I mean, do you look at that?
Oh yeah, yeah, and that we did look at too, So I think some confusion has a couple of libraries that we were looking at, and you know all those cross platform cross uh you know, whether it's Web or Maui UI frameworks, and again, what it what it came down to us was we wanted to stick with the UI components that here at Tyler Tech someone has built
and they were all web components. So we were kind of, yeah, do we want to take that and replicate it somewhere else or do we just want to keep using web components.
And it totally makes sense, as you said, because you know, the cross platform part is already done. Yeah, yeah, you're not worrying about this platform that.
Plus you've got six hundred thousand users that are used to a look and feel correct. Don't mess with that. That's just cost for you. The more you mess with that, the more trouble you're going to get.
Yeah, and we want to be good partners at Tyler. Right. So Tyler does everything public sector right, whether it be courts injustice or public safety or healthcare to student transportation. You know, there's bunches of apps that Tyler reproduces, and we want to make sure that if a parent happens to also be in a district where I can pay my electric bill through Tyler Tech, right, that that look and feels kind of the same.
Oh interesting, Right, So literally you can have multiple products in this space and they're like, oh I know this already.
Yep, Yeah, that's cool.
That makes sense.
All right, that sounds like a good place to take a break. We'll be right back after these very important messages. Did you know you can lift and shift your dot net framework apps to virtual machines in the cloud. Use the elastic beanstalk service to easily migrate to Amazon EC too with little or no changes. Find out more at aws dot Amazon dot com slash elastic beanstock and we're back.
It's at Rocks. I'm Carl Franklin, It's Richard Campbell, hey, and that's Nathan Westfall and we're talking about his application hosting six hundred and fifty thousand plus users with an S three bucket and a CDN all in Blazer for mobile clients and for web and not looking back.
Not at all, not at all, not at all.
So no web forms here, no Windows forms, no WPF, no native client code for iOS or Andrew, you don't have.
To deal with that.
Yeah, definitely not UI native code. Right. So, I think the only things that we got stuck with, and I say stuck with, they were not major to begin with at all. But in terms of either you know, browser specific, iOS Android specific, you got push notifications.
Yeah that's a problem, right, always.
Right, push notifications and authentication actually had to have a nice little hack to get to work in all of them because we wanted to use what Blazer does for authentication. But how I log in auto mobile device is not going to be the same on the web. So how do I make sure I can still use Access Token Provider? How do I still use that on MAUI?
So let's talk about push notifications. Yeah, so I'm not talking about signal ar can send you a message and say, hey, how you doing. We're talking about when your app is minimized and you're getting messages in the background and you want that number with a red circle with that number to increment that. That is not trivial to do and an Android. I figured out how to do it by doing an actual fore ground service that works as a
background service. Yep, but that adds another service, and on iOS forget about it, like you have to, you know, draw three courts of blood and sign over your firstborn and right, and then they don't even have it for Blazer as far as I'm concerned. They didn't last time I checked. So how how on earth did you solve that problem?
Yeah? So we this is I guess saying outside of the ecosystem. I know we talked about being inside of it, but we use ad your notification hub for push notifications. Part of the reason was because the older application Examined Forms already had it because it was Adure first, and we wanted to just keep the same system in case parents had both apps. Right, So it was a decision we made to just just keep it around.
But how does that notification service interact with your dormant yes app.
So we had to write code specific for Android and iOS to tie into both push notifications. So on Android it's firebased FCM now right, they officially retired GCM unfortunately, and then iOS is always messing with delegates and all this fun stuff that if it doesn't work how as you said, it doesn't work how the docs work, I'm kind of out of luck because it's not like I'm getting an help from Apple Lane time soon. So tie into both and just kind of have the native push
unification system work. And then if they happen to click the application, you know, we're doing some things with session state too, kind of if Blazer then wants to detect that I have a message safe, then I can prompt that in a totally different way or manage it totally separate.
Yeah. Wow, but now you're using notification AUPs like you now describing a cloud hybrid app too, because you're also using is three.
Yes, yeah, a little bit of both on that one.
Not complicated at all their data Imagine you have any problems? No, no, but no, I see how you got there. Totally makes sense, right, and there's nothing bad about notification apps they were.
Yeah, I've honestly out of alved the Azure services. I've used them. We used to do Azure probably at this point seven years ago or so, right from everything that they have to offer, and I've never had any complaints iuthentification of it. It's honestly nice how easy it works
from an API perspective. If I have to generate a notification, how to send it to them, and then how to managing devices, which I've had to do before in terms of my you know, P and S token on Apple and my fire based token on Android and which messages can you get and can't you get. I've written that code before to manage it, and I'd rather never do it again.
So you also hint did a thing that I've talked to a bunch of mobile devs about, which is making sure the Android and iPhone versions are utterly in sync.
Like if you had a parent's one with an iPhone more than an Android and the app was much different. Yeah, like that that, I'm surprised how upset people get. Yeah, why does iPhone have this and I don't have it on Android?
Yeah, And for us, before Blazer was the web was incredibly honesting like, I can't emphasize how bad you could see your live busting on web because it was just it was never built, that never had priority of what was there. And now what's nice is because we have all of our shared code and a shared class library. If I have to add a new button, I'm putting it in one spot, I'm quickly testing all three devices and it's out to production right after that. You know, it's it's quick.
And what are the three devices? iPhone, Android and browser browser? Okay, yeah, yeah? Do you do you much phone specific testing? Like do you care which iPhone or which Android device?
We more sot Android because of the possibilities almost being endless, right.
Yes, we recall it the drawer broken dreams for a reason.
For for iOS, we only care really about the version of iOS that we support. I don't know if you know, but iOS had a breaking change between iOS seventeen and eighteen that impacted Mali Hybrid because of that IP address zero do zero do zero dot zero bug and they took that away, but Maui was using that to spin everything up, and you know that encouraged us to do more testing around that. But yeah, really, just always changes at.
That point, well, and Apple is so insisted on getting into the latest version of iOS all the time, which is I think a little annoying for consumers but pretty good for devs.
And it's good. It's good for consumers because security, you know, and fewer problems. Apple is still the most secure mobile platform there is because they own it top to bottom, and it's.
Part of that. Part of that is enforcing homogeity. It's always the latest version. Yeah, yeah, and with you and it's certainly from a dev perspective if you're willing to jump through iOS hoops because they are iOS hoops, at least it's one os you more or less have to care about, correct. Yeah. But I've definitely talked to folks who they have to have a Pixel device, they have to have a Samsung device, yeah, you know, and then it gets questionable after that, like those are those are
the two? And it's at that point you're like at seventy percent.
Yeah, and foldables whole other.
Yeah, you do you do foldable testing? Good? Lord? Really, I just held up a foldable nice.
Yeah, only because our product owner has one, so he'll tell me immediately if something doesn't work.
Because that's what he's testing on, so he cares about Yeah, well that's all that changes your media queries completely, right, Yeah, you have to change the rendering.
Yeah, and some of it's nice in terms of most of the mobile devices kind of work immediately on tablets and larger screens because I'm writing the same code for web anyway. Right, we kind of have that mindset as a developer from dam one because all of my code for IYO with Sandre and the browser are coming from
one spot. I'm already thinking about twenty device screen sizes as I'm going here, right, so you know, making sure that works, localization works, all this stuff is now up in the forefront of a developer instead of kind of sure.
And they're absolutely iPad fanatics. Yes, if it's not perfect on the iPad, they lose their minds.
Yes, yep.
Yeah. So it's just that that issue continues to exist and it's not simple, like you really do think about how you make your UI in the different board screen constrained correct, much less tested.
Yeah. No, we even on some of our differences between like mobile and web. Right, we might write a component that renders some UI content, but then we have other logic that's Hey, if you're in a browser over this size, I'm prompting it in a dialogue, as opposed to if it's on iOS, it's just a full screen new page and blazer for it to work. Right, And yeah, yeah, gets complicated, But.
It sounds like you have a tooling team because you guys have multiple products, so it makes sense to have a tooling team that just worries about Okay, I'm going to make this new collecting text from the user interface. Here's how we'll render on an iPhone. Here's how it'll render on an iPad. Here's how it'll render on on pixel and so forth, like it's and how render on web like. You are making different versions and still make it look close to the same.
Yeah, yeah, a lot of that's still on us in terms of how we want it to look. But the components are nice and effective. They all have accessibility in mind, right, they all have different screen sizes in mind and all that stuff, So how we use them. It's kind of hard to mess up, I'll put it that way, right, we can, but it is harder to mess up.
Yeah, And you can test, right, you can test the mobile device just by shrinking the browser, right, right, and it's essentially what you do, you test the reactivity of it. Yep, Richard, do you remember when we used to do shows on mobile and the Tablet Show and all that stuff. Yeah, and we would always say, you know, what is the trade off between making your UI look like the platform UI,
like iOS Android and looking like your branded thing? And it seems like people were more concerned back then about making an iOS app look like an iOS app and making an Android app look like an Android app. No, I just don't think that's such an issue, do you.
I don't think I think you're right. I think Jobs was big on skemorphic design, and he wanted you know, there was always a threat they wouldn't let it go through the store if you didn't make it look like an iPhone app. Yeah, And so there was a time when that mattered. And then Facebook came along and said nope, Facebook looks like Facebook. Yeah, and some and then corporate internal You're like, no, it looks like corporate internal. Yeah. And I mean, and I do think Tyler falls into
this category. You have a bunch of different apps and you see an advantage to a common look and feel. Absolutely, But I'm not going to let you go, Nathan on the whole accessibility thing. So do you spend some time on visually impaired or audible player, like how different folks can use this app?
Yeah, we do. We Again, what's nice is a lot of the components kind of have support for that already, right, all the area labels and what's there the other pieces. Because we're on a whole bunch of buses across the US of Canada, we see laws coming out all the time that kind of require this, right Colorado, if you've worked with something in that state recently, you kind of know that they have this initiative right now that on June of this year, people are starting getting fined if
it's not accessibility out of the box. Right, So we kind of have that in mind, knowing how public facing it can be and all that sort of stuff.
And often public sector expected to be the forefront of that.
You need to sell it first, yeah, correct, Right, the law comes from the government, the better the government better be compliant, right, So.
Totally yeah, So and yeah, when you think about phones, like, there's a lot of different ways to do that. You could you could provide this whole interface as audio only, correct theory if you think it through properly.
Yeah, And we've talked about that before. We're we were going through the app a couple months ago. We're like, yeah, you know, we can improve a few areas and really start and dial in some things in. What gets complicated is the map, Right, how do I open gl screen kind of explain what's going on? So we've talked about how we can kind of accomplish that.
Yeah, I think you would. I mean in the end, if somebody else was operating that app on behalf of somebody and describing what was going on, they don't be that precise. It's like, yeah, the bus is in motion, it is this far away from the stop is expected at this time? Correct?
Yeah, And it's it's kind of how we want to see how far we can go with. It's kind of what we're starting to play around with, right, because I might have a kid in sixth grade and a kid in tenth grade, and one's going to the middle school, one's going to the high school. How do I explain now both buses and the relationship to where I am and where they're going.
Yeah, that's actually I think it'd be a really cool project. That'd be really fun. It's like, how do I do this well where people are wowed? Yeah? Yeah, I kind of like the idea of using that interface because the parent is driving and does want to know where their kid is and you don't want to take their eyes right, Like, can I pump this through bluetooth and tell you safely? Yep? Yeah, the bus is fine, Yep, it's good.
Yeah. We have some experience in this already because another app, this one's examined because it's only for Indurid, but we actually put tablets on school buses, and once the tablet goes on a school bus, a lot of them, I say, a fair amount of states have the rule where the screen just has to be black when the bus is in motion, So we are almost audio only from this point on right, how far away am I from my stop?
Where am I taking my directions? So we basically a lot of our testing is done with just the screen off and how well can you understand what's going on with it being completely blocked out?
That's really cool?
Yeah, and have some great user acceptions testing to just watch what does a driver need in motion from my app? What will make them the most benefit?
Did you actually go around with drivers, not you personally, but somebody in the company go around with drivers and get their feedback.
I actually did, and there's some people in the DEB team that did so. I before I was a manager. I was actually one of the lead devs on that project, and I went across the US riding buses with you know, all sorts of kids on different districts, and you know, got to see it firsthand and how it can both impact.
But how my mind is, I don't know. I think did the right thing putting a developer on a bus, because now all of a sudden, I think of twenty more problems I have to fix, seeing like what can happen.
Right, I've always been a big fan of this putting putting the developer into the position of the user. Yeah, because they do come away with exactly that. Oh I've seen how they actually use this, and we've been thinking about this entirely wrong.
Yep, yep.
It's powerful and it also connects up like we can get really abstracted from our software. Yes, like when you see it in the field helping people, it's pretty powerful.
Yeah. No, it's great to see the kind of the impact you can make, right, I mean I did not realize. I you tell people, before I worked at Tyler, I did not think this much about school buses. I started working at Tyler when I was when I was twenty, so I was only in a school bus two years before this, and I still did not think of a school bus as much as I do.
That you were more school bus savvy than most. Yeah, right, that's really funny. That's cool. Yeah, it's one of those truths. Like the great thing is when we do this kind of tech, right, it's it's almost invisible. Yeah. Yeah, you know the fact that they know you have to sort of have a moment ago. Wow, I whipped out my
phone and it told me where my kid was. Yes, you know, it's seconds, including dealing with all you know, it's a it's a heavy snow day, but no, the bus is in motion and fine, yep, that's what you wanted to know, right, And that's really thinking about all the pieces that had to work together to make to give you a good answer. It just was impossible that going.
Back to the notifications thing, if what kind of notifications do you have? Obviously, like if the bus breaks down, side of the road, accident, whatever.
Yeah, so we support a whole bunch. So there's a couple of them are actually integrated directly with the tablet, which, again talking about saving computing resources, it's the tablet on the bus with the district's local map and where the bus currently is, where they're going next. Is the one doing all those ETA calculations. Right, here's the ten parents. I need to notify that I'm about to be at their stop, because I, as a driver, said it's my next one, and I got three more turns to make.
So that's awesome, right, the tablets doing all the work for the mobile device.
Yeah, the nice little localized service there, it's the best position to know. The workload's not more than a tablet can handle.
It's well, I imagineflove bus breaks down. There's could be a there could be a marriad of reasons that you won't know. So the driver can type into the tablet. I'm sure ran out of gas in a tree, dog, you know, hit a deer or whatever.
Yeah.
Yeah, now they can let dispatch. No dispatch can let x number of parents know whether they based off the bus or even you know, we track how far along the buses and the route, so I only need to notify parents after said stuff. Right, I've made it halfway through and now I can tell the other half. We're working on next, and I'm interested to see how we can do this with Maui. But we're working on doing
those live activities. Right. So if you've ever done an uber and you can kind of see live update, it's not quite a push notification, right, it's one of those live tracker ones. We're trying to play around with that to see how far we can take it. Where if you don't even have to open the app and now you're fully functioning off push notifications, that's kind of awesome, right, I kind of have less resources on my side and you're getting this up to date information without even opening the app.
Right.
You takes coordinates and stream them to a service somewhere and pick them up on the client, yeah.
Yep, just to be able to say, tell me when they're ten minutes out, yep, I can get there on time. That's clever. And again that would actually save resources and not leading on the app all the time. You're wasting anybody's time that way.
Correct. If I can get everything from that live activity, it's all technically feeding from that tablet device and then they never open the app. I mean, I'm like, we're not checking anything, right, it's all coming directly from the bus YEP.
All right, what have we talked about on this app? What is your testing strategy? Like? What tools are you using when you when you do a bill?
Yeah? Good question. So we do fare money unit testing. Our API is also in dot net, So that's all unit tested with X unit because all of our UI could have shared. We're using b unit to test as much as we can, right, So as much as I can test in terms of rendering or business logic I might have on the client side, testing that kind of like I would an API, right, And then we also use browser stack to kind of run into end tests on here, right, So launching a couple of devices, making
sure they can do X number of user functions? Right? Can I add my student? Can I load the bus? Can I create a new account? All that sort of stuff on number of devices and let us know. So we don't have obviously coverage everywhere. I don't think anyone ever does anymore. Just development moves so quick it's hard. But those unit tests, especially the unit knowing that I can kind of have the same rendering and the same functionality really quickly. Is good to see.
Absolutely, it's great. Anything you don't mean the same mistake again and again make new mistakes correct. It's enough testing in place to capture the last mistake so it doesn't happen again.
Correct. Yeah, we tried to raise many tests upfront, but we have a policy here where if you get a bug or something client reported, it has to get fixed with the unit test. Right, It's always that regression case.
Yeah, first make a test that fails, then fix it and the test doesn't fail. Now it's actually fixed exactly.
Yeah, we love the unit, love the gal Yep, just good stuff.
It's a great product. Yeah.
Well, and also I try I appreciate you not fixating on test coverage. Is that it is its own kind of crazy.
Yeah. If you were trying to shoot fround percent test coverage and it's kind of you're never going to get it right. Yeah. So the nice thing that developers can understand is we give them all the tools and they're able to do it upfront, because sometimes it is faster. If I'm building an API service and I don't want to keep trying to mock all this data from the tablet. Yes, I'm going to write a unit test because I can
grab something that's usable and keep integrating over it. And then the simple requirement of just if you're fixing a bug, you have to write a test to something that people can kind of swallow.
And those auld end up being the best tests anyway. Yeah, they were ones with a bug that made it to the field and you wrote and the testing for it is more complicated than you think. You would never write that test out of your head just considering it. So you write this better test and then you know, go through that process and fix it and maybe it never ever hits it again, but at least you won't have that problem.
You do mocking, We do, yes, well, we do mocking because a lot of this data comes from our main application, student transportation, right, and I don't want to have that running for this, so we mock a lot of the database and kind of just core data.
Right.
So the way this app works is a lot of it's read only anyway, parents not really obviously updating their location for the bus and all that. So we kind of can get away with a little more mocking than you might be able to.
So you have like bogus services that run and send mock data that kind of thing.
Yeah, yeah, bogus services we use. I forget that. I think it's called test containers. So even some of our implementations with redtis and how we you know, not kill our services by having one hundred thousand parents watch buses? Right, how do we kind of scale that efficiently? We can kind to do that cool mocking it, Yeah.
And a lot of it, I think a lot of Just because you've got a dozen parents want one bus doesn't meaning you can catch that once. It's also correct, very granular.
Yea, sounds like a fantastic, very successful product.
I appreciate it's It's been a fun one to work on. I I selfishly like, uh like being involved in this project. It was a really fun one for me to work on the team, to work on you know, kind of even development kind of picking this up from what we do day on to work on this. Everyone's kind of enjoyed the dev life cycle more than anything else. Right, It's kind of easy to get going. We had someone start a month ago and they were making fixes in
the app a couple of days later. Right, it was just easy to pick up.
That's a great metric. How many days to productivity for a new worker. Yeah, great metric.
Well, Nathan, what's next for you? What's in your inbox?
More Blazer? So this has been this has been a little test for us, and it's been successful enough where we've got a few more around the corner that we're working on. And I hope that, you know, dot net kind of sees the benefit of Blazer as we get more and more to this space and keeps dumping time and energy into it, so hopefully I'll keep using as long as I keep working here.
So that's great.
Well, thanks for sharing the story with us, and congratulations.
Yeah, thanks for having me. It was fun being able to talk to you guys and kind of share what we did right here. You talk about Blazer all the time at dot net rocks, so I wanted to be like, oh, I got a story for that.
He'll love you certainly, all right, Thanks again, and we'll see you next time on dot net rocks. Dot net Rocks is brought to you by Franklin's Net and produced by Pop Studios, a full service audio, video and post production facility located physically in New London, Connecticut, and of course in the cloud online at pwop dot com.
Visit our website at d O T N E t R O c k S dot com for RSS feeds, downloads, mobile apps, comments, and access to the full archives going back to show number one, recorded in September two.
Thousand and two.
And make sure you check out our sponsors. They keep us in business. Now, go write some code, See you next time.
You got jam Vans
And