Episode 793 transcript

FLOSS-793

Jonathan: Hey, this week, Aaron joins me and we talk with Jay Khatri about Highlight. That's an open source project to do monitoring for your web applications. They have some interesting tricks up their sleeves. You don't want to miss it. This is Floss Weekly, episode 793, recorded July 23rd. Keeping an eye on things with Highlight.

io. Hey folks, it is time for Floss Weekly. That's the show about free Libre and open source software. I'm your host, Jonathan Bennett, and we've got, we've got quite the show today. It's going to be a lot of fun. So first off, I have a co host and it's surprise Aaron Newcomb. Hey Aaron, long time no see.

Aaron: Yeah.

Hey Jonathan. I can't do the regular scheduled shows on Wednesdays when they happen on Wednesdays. So. So I can't join as often as I'd like, but really

Jonathan: glad to be back. Well, so we may see Aaron in more rotation because as everybody knows, we have moved at least for now to Tuesdays for doing the show and oh my word, that is better for me too.

So I've got the security column that goes out live on Hackaday on Friday morning, which means that I have to have it done before I go to sleep on Thursday night. And recording Floss on Wednesday, editing Floss, publishing Floss, and then turning around and immediately starting on the Hackaday security article and having that done, it's a very busy 36 hours.

And I finally said, I'm the boss of this show. I'm the head honcho. I'm going to move it. And it's, it's been great. And so now I have, you know, almost 72 hours to get everything done. And it's so much better. Yeah, that's awesome.

Aaron: That's awesome. I mean, the original reason why it was on Wednesday was just because that's where it fit in with the twit schedule.

I mean, there was no, no real rhyme or reason to it. It was just like, Hey, we've got this open time slot on Wednesday. And that became the time slot. So it's not like there's anything else really locking us in except for. You know, people obviously, long time listeners and viewers will, will know that it's usually on Wednesday,

Jonathan: but that's easy to change.

Yeah. And if you can't watch it live, that's fine. We still got you on the download. You can still catch it.

Aaron: Yeah.

Jonathan: All right. Well, hey, we've got, we've got somebody fun today. We are talking with Jay Jay Khatri about Highlight, Highlight. io. And this is kind of a, it's like a, Browser monitoring, monitoring software for web apps.

I think I'm not entirely, I don't have a whole lot of, of inspection into this. But the idea of it sounds really cool. Have you, have you gotten to dig into this yet?

Aaron: A little bit on the website. I don't know a whole lot about it. But I do have a background in monitoring, right? So I've been doing monitoring for forever working at companies like New Sysdig.

So I was always involved in monitoring. So I am kind of, that's one of the questions I have definitely. And so we'll have to spend some time up front on the, what is this part, right? Defining what it is and what it does, because, you know, Traditionally, there's been for, for me anyway, working in the industry, there's, there's two or more schools of monitoring, right?

There's application monitoring and then there's infrastructure monitoring. So I'm kind of curious to see which one this is. Maybe it does both. How are they using AI as part of their product? Because everybody's using AI nowadays. AI is, AI is, is huge for this because it can, you know, detect anomalies and, and, and look for patterns and, and then tell you what to do.

Hopefully, if it's intelligent enough, what's going on. So, yeah, I'm super interested and really excited to hear what they're doing and how they're doing it.

Jonathan: Yeah, this is, this is one of those, I'm, I'm a little not necessarily AI skeptical, but I have AI fatigue because It's a new tool that like suddenly got better.

And so we're trying to do everything with it. But like, this is legitimately one of those cases where it may really fit because what are you wanting to do with monitoring? You're doing pattern matching. You're wanting to show me these sessions that are outside the norm. And like AI is just. So I, there might be a, a really neat synergy, not to, not to buzzword it up, but there might be a really neat synergy there with AI.

So let's not guess anymore. Let's bring the man on himself. And Jay, welcome to the show.

Jay: Thanks for having me guys. Yeah. It's kind of fun to, kind of fun to be seeing y'all Guess what Highlight does, but I feel like I really want to talk

Aaron: now. Yeah, you're chomping at the bit.

Jonathan: Go for it. What did we get right?

What did we get wrong? What is, what is Highlight for? What's the point? Yeah, for sure.

Jay: So y'all, I think y'all got it pretty well. I feel like the way that Aaron mentioned the difference between application monitoring and infrastructure monitoring is a good way to start. So we, first of all, Highlight doesn't focus on infrastructure monitoring.

We're sort of explicitly at the application level and essentially what we focus on is connecting your client. And like your browser application with all of your server side, like more traditional observability resources. So the way it works is like you install our client bundle, which could be in React, in Angular or whatever your like client application is instrumented with.

And then you install a handful of SDKs on the server side. And then we do the work to basically piece together what happens all the way from the client to the server. And the way we think about it is like, it's essentially like one big trace, right? If I can see what a user's doing. doing and what they're clicking on, and then what error they might be interacting with being able to connect that to what happens on the server is very powerful because then I can actually go that way, right?

I can know, okay. An error was found. What database call on the server caused this error in the first place, or I can be looking at some database errors on the server in the first place, but then saying, Hey, how is that impacting my user base? So that's kind of what we focus on the application level, but really tying everything in the observability world to how it impacts your downstream users.

Does that make sense?

Jonathan: Yeah, I think so. I think it does. Are we talking primarily about in the browser applications? Is this primarily web apps or does this apply to different kind of applications?

Jay: Yeah, yeah, for sure. We definitely focus mostly on web apps. And the reason that that's the case is like our client side SDK that I'm mentioning.

It runs in the browser in JavaScript. So essentially you install our SDK, you initialize it, and then we're basically monitoring the DOM. Like we're monitoring the UI to be able to then replay it after the fact. So it's pretty cool. You actually install it and then you can actually see that, okay, Aaron clicked on a button.

And when the button clicked, a new page was rendered or new text was rendered, and you actually can see it at that level. So With almost like a video, like replay of what's happening, but then it all gets connected to all of the server side logs, traces, so on and so forth.

Jonathan: Yeah. Okay. So the, the first thing, and this is, this is just.

I guess sort of a weird thing to say, but I'm, I'm reminded of Microsoft recall and that little, that little bit of a creep factor that people got when it's like, what do you mean Microsoft is spying on me? So you're, you're, you're kind of spying on users.

Jay: Right. Yeah. Yeah. So, so we get that, we get that a good amount, right.

And I think for our, for our customers, like they often ask that question, but that, that. The, the, the default thing there is that first of all, when you install highlight by default, we actually mask a lot of the data. Okay. So any of the texts that gets rendered, we actually will obfuscate it before it gets sent over.

So things like emails, things like numbers, like if it's a series of numbers, we just won't record it. And there's like a, kind of a long list of things that we just don't record by default. And then obviously if folks want to go even more strict. There's an option called strict privacy mode, which it actually obfuscates all text.

And so we try to keep like sane defaults that way. We're like really respecting our. Customers, end users, privacy, right? And that's kind of how we think about it. And all of that obfuscation is done on the client, right? So it's never, it never hits our servers. It's not like it hits our servers and then we mask the data.

We do that all on this, on the, on the browser side.

Jonathan: What could possibly go wrong with the idea of masking things on the server side?

Jay: Yeah, exactly.

Aaron: I was just looking as you were talking in, you know, certifications and compliance, because that's obviously a big part of this too. I mean, there's just general privacy concerns on the customer side, but then there's also like payment information and all that kind of stuff that you have to obfuscate.

So yeah, I was just looking at what your Compliance certifications are so,

Jay: yeah. So we have all like the, the, the default, I guess, compliance certifications. We also have a handful of customers in like the healthcare sector and kind of more, more more regulated industries. I think the, I think the other thing about it, Aaron, is that like, because our product is open source there, it opens up a lot of possibilities for our customers where they don't actually have to send us data in the first place.

And they can almost pay us like us. Like a traditional licensing fee run the product in their own VPC. And then it's as if they're just sending data to themselves and it's totally fine. Right? Because. Because anyway, they're still storing that data that's being rendered in the UI in their database.

And so that kind of like solves a lot of those problems, but right, you don't

Aaron: have to worry about sending it to a third party. Oh, I'm using this service. I'm just running it myself. And then, yeah, yeah. That's a very common situation to run into companies that are like, no, we can't use a SAS service, for example for those reasons, right.

We have to, even as easy as it would be. And as, as much as we would love to hand you our credit card and let you bill it monthly, We have to have access to the bits because we just, we just aren't allowed, you know, you can think of government, obviously government Situations, healthcare. There's just a lot of companies, financial.

They just, their, their policy is we can't do it. So we have to run it ourselves. So that's cool. And open source provides for that. And of course that's, you know, that's going to be a big part of our discussion today, but we'll get there in a minute. I want to talk, I want to talk more about the product first.

Okay. Okay. Let's do it. So, so who, who who would you say your competitors are in this space?

Jay: Yeah. Our biggest competitors, our guests, I guess, are the other companies in like the application monitoring realm. Because they have recently come out with, I guess some of the competitors have come out with session replay.

Some of them don't do it in like the same way we do, but they do have it. One of the companies in that space is like new relic, for example, they just came out with a session replay offering. There's another cup, I guess if you're familiar with log rocket, They have like a session replay offering, but they focus more on like analytics and product analytics and stuff like that.

I think the big thing about what we do differently though, is like that whole connecting the client to the server, right. Being able to like piece together everything that happens for almost from like a user journey. And so if you look at New Relic's offering, it's pretty, like, isolated, I guess, and I think a lot of the big players are like that.

And so we kind of sell our product as, like, one big product, rather than, you know, multiple SKUs, and that's kind of how we think about how people should use this sort of tool. Does that make sense?

Aaron: Well, totally, since I used to work there. Yeah, we had it all, we had, we had all those products separated, and they combined them into one UI.

But I think they still sell them differently. So you had, I work specifically on infrastructure monitoring,

Jay: but then we

Aaron: had APM, right, which was your application monitoring, we had synthetics. Right. Which was like your testing and then we had user experience. So you could get your, so exactly what you're saying, like you have it all streamlined into one thing.

And the nice thing about that is you have a nice journey for what your application is doing. So here's what happened at the front end. They click this button. And then here's what happened in the middle as it was, you know, traveling to the cloud or wherever your infrastructure is, here's what happened on the back end.

And then all of a sudden there was this error. I'm assuming you can show that because you've got metrics here. I see. Yep. And then that's why this particular thing was so slow. And that's why your users saw the little circle going around for a minute before, you know, that's the thing that drives us all nuts, right?

Yeah. And I.

Jay: I think the interesting thing there also, like on the business end is we've, we initially built this for engineers. Like our original idea was like, Hey, we just want to build like a session replay tool. That's built for the engineer. Like how, how can we like reproduce the dev tools of the browser, but then give them the visual power that they would get with one of these like session replay marketing tools.

Right. That was the kind of original vision. But I think Over time, we realized that like, once people get this in the hands of their business, it starts to spread in kind of weird ways. So for example, you mentioned like being able to understand why the spinner doesn't stop spinning, right? That's actually very relevant for like a support person, right?

So if I'm a support person at a tech company, and I get like an inbound ticket that says, whatever spinner wasn't spinning, right? I can actually use highlight to help The engineering team triage before they even get the issue in the first place, because perhaps it's not actually an engineering issue, you know?

And so I've been in, we've been in like demos with support and engineering to kind of teach them how they can work together to kind of fix issues faster with a tool like highlight, which is kind of cool. And I don't know, it's kind of awesome that like the tool itself. Broadens the market in some sense, right.

Where we can kind of start to sell to other, other personas throughout the organization. Yeah, exactly. And there's a

Aaron: huge business benefit too, right? Cause you don't want customers abandoning their session. You want to be able to actually show, no, we're making these transactions faster. We're doing more business, you know?

So once you get past the technical part of it, then especially if you can show that, Hey, we, you know, accelerated or, or you were able to fix your problem faster or whatever. And that means you can do more business. Then it becomes a no brainer. It's like, yeah, we need this because yeah. So

Jonathan: yeah, I'm real curious.

Is there any way our people do you using this and like their continuous integration, their test suites is I can, I can imagine. And I, I sort of have this past weekend's debacle on the mind where someone, a very large company pushed out an update and, you know, soft bricks, like 8 million windows machines around the world.

And I'm just, I'm just thinking now, you know, more of us are thinking about maybe we shouldn't test in production and we should have continuous integration and do some of these tests. So I'm curious, is this, is this a piece of that? Can I, if I have a web app, can I put Highlight in my CI pipeline? And then, you know, if something fails or maybe even it doesn't fail, but Highlight says, This took 30 milliseconds longer in this run than it did last run.

You know, are those, is that sort of tooling there?

Jay: The tooling is there, but it is not, I would say it's not a common use case. And I think it's just more of like, we're a pretty small team. So it's just like what we focus, like what we want to focus on in the short term. We have customers though, that do run highlight in their testing environments and then kind of like tag specific data based on the test that's being run.

But I would say it's like not super purpose built for that. Yeah. And I think Aaron was mentioning synthetic monitoring earlier. Like I think actually a long term vision of connecting highlight to your synthetic monitoring infrastructure, whether we build that or whether we partner with someone else could be a really cool.

Product, you know what I'm saying? Cause most of these synthetic monitoring tools essentially like run puppeteer or like a headless browser and then like just record the browser and then maybe record some logs and all that fun stuff. But the cool thing about highlight is it's, we're just replaying it.

Right. So you'll actually get all the Dom insights and then you get the connection to the backend trace. And so, yeah, I do think there's a world where, where that could, that could work out really well.

Jonathan: Yeah.

Jay: Yeah. How long, how long has highlight been around? Yeah, so we've been around for like two and a half years now.

And our team's about 10 people strong, strong and mighty 10 people.

Jonathan: That's a good, that's a good size team. You know, you can, you can manage a team like that fairly easily. You get too much bigger than that. And it like, it gets, it gets sort of top heavy trying to keep everybody on the same page. I think, I think, like about 10 is, is really for, for something like this, especially is, is really pretty ideal.

Jay: I agree. I think it's like a. You always think that, Hey, we throw more engineers at this problem and maybe we'll, we'll be able to build more. But I'm, I, every day I believe more and more that that's really not the case. You know, it's like two, one to two engineers on a given project is like the, is the sweet spot.

And I don't know, I, I'm scared of the day when we, we want to become a bigger team because I just don't know how the management structure will look and all that stuff. But for now, yeah, I'm really happy with it and it's been awesome so far.

Jonathan: So what, what was the what was the, the problem space originally that you guys wanted to find?

Like, were, were you seeing, you know, application errors? Were there security problems that you were trying to catch? Was it just performance? Like, it seems like you can do all of these three things. I'm just curious, which one was the first focus?

Jay: Yeah. So, so a little bit of background on the company is like, I, before Highlight, I actually went through Y Combinator with another company.

And At that time, we were working on kind of like a, maybe it's not worth going too deep into actually the startup itself, but there were a lot of companies with us in our batch that were using a like marketing tool, like hot jar and connecting it to like a error monitoring tool, like bug snag or century or something like that.

If y'all are familiar with those tools, right. And. It was kind of like hacky, right? Because it was like on in, in the bug snag tool, you can get the client ID of the session and then in the hot jar tool, you can also get the client ID of the session and you have to sort of piece that together. But it was very like, Hey, a customer comes up to me and says, I have an issue.

And I, as a startup founder, want to be able to fix it really quickly. Right. Cause I don't have many customers in the first place. How can I do this? And so that was kind of the problem statement, honestly, honestly, initially, and then I think from that, that's where we started to kind of find the enterprise value among larger customers over time and that sort of thing.

So that was kind of the original idea. And honestly, it hasn't changed much, right? Like. Connecting the client to the rest of your observability resources is, is our motto, like that's what we want to focus on. And now only recently, I guess the last year or so we've gone more into like server side monitoring and all that stuff.

But I do think that the general principle of wanting to connect all of these source sources is it remains true, you know,

Jonathan: So I'm curious, you had this problem, you put together your own little tool to solve your problem. At what point did you say, we should just release this open source? Was that like from the beginning you wanted to do it that way?

Or was there kind of an aha moment of, this could be useful for other people, let's set it free?

Jay: Yeah, that's a good question. So, we, we, originally it was not open source, it was closed source. But I don't know, I've, I've worked in a lot of companies that were very pro open source. And so even all of our infrastructure components, we never really relied on like managed services and things like that, which looking back on it, it's like, I guess, convenient that we did that.

Right. But yeah, the first year, year and a half, it wasn't open source. And then we started to kind of go more into enterprise. Like we got a few customers that were like, you know, like more than 200 person teams kind of thing. And at that point, I think we realized that it would be beneficial from more of like a convenience perspective for these customers to be able to self host, highlight, manage versions of their own highlight instance.

And then I think the other thing that like me and my co founder did is we reached out to all of our customers and we were like, Hey, if you, if we, if we open source this, would you self host it on your end? In other words, we were kind of trying to figure out whether they would stop paying for our SaaS.

Yeah. Our SaaS provider. And none of them said yes. So none of them, none of them wanted to actually open source it. So to us, it was actually pretty it's pretty de risked, right? Like we were like, okay, if we open source this, it's not going to change our existing business among all the startups that use Highlight.

And if anything, it'll just increase trust, right? Cause they're seeing the open source project and they know what we're working on at any given time. And then it gives us an opportunity to really focus on getting in those larger accounts and kind of proving the more self hosted software licensing kind of model, you know what I'm saying?

Jonathan: So I'm, I'm curious. You went, you went open source. What has the buy in been since then? So as, as some of your customers pushed code back in have you had in, in, in the interim, have some of them gone to self hosted? Have you, has open, has open sourcing the product lost you customers or has it been all upsides?

Jay: Yeah, good question. I mean, I don't know if I can say it's all upsides. I think there's, There's a lot of work when it comes to an open source project, just in maintaining it and accepting PRs and all that stuff, right? Like you kind of open up another level of, of committing to the project, I guess, right?

But I would say in terms of like business value, it's definitely been all positive. Like we haven't, yeah, we haven't lost any customers on the lower end for sure. And then on the higher end, like we work with some of like some very large health and health insurance, health care providers, we work with very large hedge funds and banks.

And so I do think that that we've proven out the fact that that's been worth it for us. And obviously, you know, we can always be putting more into the open source project and marketing it and getting more contributors and things like that. But so far it's been working really well. And then on the contributing end.

We've had maybe like 20 30 contributors at this point outside of our company.

Jonathan: Oh, yeah

Jay: So it's been a pretty healthy honestly growth on that on that sort of end as well

Jonathan: have have any of those contributors been like Continuing repeat contributors to it or they mostly drive bys because I would imagine I would imagine on a project like this You would get a lot of drive by like High quality, but drive by contributors where it's a business that's using it.

That says, wouldn't it be nice if highlight would also do this, or we, we have this specific bug that we need to fix. I imagine there's a lot of that.

Jay: Yeah. Yeah, there is, there is, which is kind of cool. Right? Like if you're, if you're a potential customer of highlight, you can be like, Hey, there's this issue with this SDK.

Can I come and fix it? And we're happy to happy to accept it. Right. So that's kind of cool. But, but on the on the contributors end for like drive buys, I would say maybe 80 percent of folks that come in are more drive by and are not like repeat. But we do have like a good group, small group of folks that are pretty consistent.

And we've put in a lot of work to like incentivize them to keep helping us. For example, we have like a bounty program. I don't know if you've heard of Algora. No. Okay. It's not particularly. It's like a SAS product that you can connect to your GitHub. And then we basically pay Algora a percent of every transaction that we pay to our contributors.

And so our contributors can land on a PR and they'll see that there's a bounty on the PR and then they can help us with that particular feature. So it also helps us prioritize what we want to be contributed to, because a lot of the repeat contributors, they're just working on what they want to work on, but if we can incentivize them in the right direction, it's kind of cool We can align it with our product roadmap too.

Aaron: And that's great for, for people just getting started. I always yeah, you know kids I'm old enough to call college age kids, kids getting out of school there. They, a lot of times they'll ask, how do I get started even in high school? I mean I did a, I did a high school. Talk, you know, about startups and working in technology and things like that.

And that was one of their questions is like, we're really afraid because we're going to go invest all this time and money into college. And we, you know, we keep hearing that it's really hard to find a job. It's like, well, that's what, you know, open source is great for that. Right. Yeah. Yeah. And it sounds like this is even better because now you're getting paid and you have a system instead of just going out and searching GitHub for what you want to work on or, or areas where you can work.

Now you've got a nice system where you say, I want to work on, Monitoring, right? And now you can find those problems and try to solve them and get a little money and get some experience on your resume. So I think it's awesome.

Jay: Yeah. Yeah, it is really awesome. And that Algora team I mentioned is, they're also a startup, so they're very open to feedback.

Like we've kind of worked with them to, to get some features in that would help us kind of manage and triage and all that stuff. So I do just a small, small shout out, like to check out, check out that project. I know there's a lot of open source folks here. But it's a, it's a pretty cool project for sure.

Yeah. I'm going to bookmark

Jonathan: it. I'm, I'm looking at the Algora page and get hub too, because there had, there have been groups that have done this, that have done bounties over the years. And some of them kind of eventually dry up and I forget the name of the one, but there was one in particular that just sort of quietly stopped paying bounties out and I don't know that we've heard much from them ever since.

Oh, I see. I see.

Jay: Yeah. That's pretty tough.

Jonathan: That's tough. But

Jay: it is, but I mean, it's, it's, it's kind of cool, right? Like it, yeah, it's, it's pretty awesome for us that. We can we can help those folks out because I think it also Even beyond like helping that small group of folks that are repeat contributors it also grows our Pool of contributors that just want to be repeat contributors, you know, so it's been awesome for us and I really like that model that's been working out so far

Jonathan: Yeah, and so when when you guys decided to go open source, what was the what was the license that you went with?

Yeah, we went with apache

Jay: 2. Which we're, I think, still pretty happy about. I don't think we make plan, have any plans to change things. The only thing we've changed, I guess, on the licensing model is being a little more explicit about what we charge for on our open source license. Our, our, our open source distribution versus what was, what is free.

So when we started the project, you could just install highlight. And get it running in like a Docker container and all the features in our SAS product were accessible in the self hosted distribution. But now what we've done is we've actually, it's a little more open core. If you want to think about it like that, where certain features like, like RBAC and SSO and a lot of the like more super enterprise y features we've gated on the open source distribution.

But You as a team, if you want to just try it out and even if you want to use it long term, we still have a lot of traffic, honestly, on our open source installs. You can kind of get going with just the core functionality. Does that make sense? Yeah.

Jonathan: So I was, I was going to dive into this. We can go ahead and talk a little bit more about it.

Like what, what the. What the pricing model is, like how, so as we like to say here on the show, programmers need to eat, even if they are open source programmers, they have rent to pay as well. And so, yeah, I'm, I'm curious. Let's talk a little bit more about that. Like what, what is the revenue model? Do you have, are most of your customers still sass where you host the product for them?

And and let's, you know, let's dive into that.

Jay: Yeah, I would say more than 80 percent of our revenue comes from sass still. I do think that the pool of self hosted folks is starting to grow. Just given the investments I've, I mentioned the investments we wait, we've made in the, in the sort of open core self hosted kind of model I mentioned before.

But yeah, the way we charge is like on the SAS product, it's like 50 a month for like an indie person. That's like just getting started. It's like 50 a month. And then you just pay a usage fee on top of that. So if you're sending. 10 million logs, you'll pay an additional a hundred bucks a month or something like that.

And it's pretty simple. So that's kind of like the, the base tier that we offer. And then on top of that, we have a couple more tiers for like larger businesses and larger business sizes and things like that.

Jonathan: Yeah. Yeah. And I, I

Jay: assume that

Jonathan: you, you make use of the highlight client yourself in, in all the, you, you eat your own dog food to some extent.

We

Jay: do, we do. Yeah. And, and that's. That's been pretty cool. Actually. We actually funny story is like early on, we used to demo highlight installed on highlight and it would just confuse people so much because it's like, you're looking at the highlight UI and then the highlight UI is the highlight UI and you're clicking play on it.

So that's maybe a bit of a regret early on. Cause like people were just, didn't even know what they were looking at. But yeah, we do use highlight on highlight and it's been pretty cool. We use it for obviously all the features and we monitor our own app. We actually have our other, like right now well, yeah, right now we use we use the session replay tooling and we connect it to a lot of our like server side tracing and things like that, and it's been pretty cool to demo to our customers that we do this ourselves.

And we almost use that as kind of like what a demo environment looks like, even though we kind of replace what the screen looks like. So it's been, it's been awesome using highlight. Cause it's like. We can, we can just prove to folks like this is how it should be done and what it could be like if you use something like this, you know?

Jonathan: Yeah,

Jay: absolutely.

Aaron: How hard is it to get started though? Like, is this something that you could go in and do a POC at a customer pretty easily and just say, Hey, pick one of your applications, we'll get it installed and we'll see how it goes basically. That's always well, that's one of the ways I find that, you know, when you're trying to sell this stuff and get people to buy it,

Jay: if

Aaron: you show them how much value they can get out of an application they're used to, to your point, nobody knows highlight on highlight.

Right. Then you know, it just, it just makes a lot more sense to them at that point.

Jay: Yeah. Yeah. It's very easy to get started. And a lot of our early larger customers came in, we did like not even reaching out to us. Like they just came to our landing page, made an account and started sending data through.

And then we kind of went through more of an enterprise buying cycle or whatever. Right. So yeah, a lot of our customers do the POC without even us knowing, and then they come out to us and they ask more questions and that's great. But, but when it comes to more. Larger customers at the get go. Yes, we get a POC going and we can just have them install it on their own application or one of the applications they have.

And because it's so simple, it's like at the application level, it's very easy to get things going. Whereas I feel like with a lot of infrastructure monitoring, you have to install agents, DevOps needs to be involved. It can be like kind of very hands on. For us, it's pretty nice. Cause it's like, Hey, these are the three code snippets you need to add throughout your application.

And you'll be able to see a session connected with a log on the backend. And then from there, they, the, that's like enough to get them to a point where their imagination can kind of take them to what, what, what capabilities they want, you know what I'm saying?

Aaron: Yeah, absolutely. Yep. And it can be difficult, I think with, cause you're kind of like, You're up against a few different areas with this product, it seems like.

And what I mean by that is like, I'm just going back to my time at New Relic where you would have to install multiple agents to get the full experience, right? I mean, you would have to, okay, we're going to use, you have to use it. You're writing your application in Python, so you have to use this Python library.

And then for Synthetix, you have to set up this whole environment and install software there. And then for the backend stuff, you have to install another agent on your server. And To log what's going on. Oh, and by the way, there's logging. Are you going to send that to Slack? What are you going to do? And so all of a sudden it seems simple at first.

And then as you go through, in this case, let's say a POC, cause that's what I'm used to thinking about this in the customer kind of gets a little bit of buyer's remorse because they're like, Oh, this is actually a lot harder than you told us it would be. Or you made it out to be, you know?

Jay: Yeah. Yeah. I think, I think the, and this is like maybe more of a philosophy, philosophical thing as well.

I think. The future is going away from infrastructure monitoring. In my opinion, like I think, you know, with, with serverless concepts and the fact that, you know, now you're not managing your own containers, any, any given startup today, right. That starts, they're not dealing with Docker containers in production.

Maybe they do it locally or whatever, but I think that's, That is maybe the the argument for why application monitoring is a future and us kind of focusing there is where, where we'll like start to, to build more and more value for the market. On the, but yeah, at the same time on the POC end, I, I, I, it makes sense that like for infrastructure, it takes a lot of work and there's like.

An SDK in this case, cause tracing, you need application level stuff, but resource consumption, you don't need application stuff, so it's just an agent. And it's kind of like a lot of random stuff. I think the nice thing about what we're doing is it is very focused, right? And people can install and send a little bit of data to get, to have an idea of what the full value picture looks like there.

Aaron: We should talk a bit about AI too, before we get too far into this and how you're incorporating that. I mean, everybody's mentioning it on their pages, but then I also want to just to set up a further topic. I do also want to understand like other, like you mentioned that you work or it seems like you're, you're, you're open to working with other open source.

Projects and things like that. So I wanna talk about ecosystem as well. Mm-Hmm. . But let's start with, let's start with AI since that's the the, the, the site guys. Elephant in the room here. Yeah.

Jay: Yeah. I actually, I actually have a, a, a, a a a a twofer for you on Aaron, Aaron on that front where we actually did recently a collaboration with an AI related project with highlight.

So I'll tell, I'll, I'll tell you about that after. But yeah, on the AI end, I feel like I may be in between y'all two where I'm not too much of a skeptic, but also like, I think we tread a little lighter than maybe most of most of the startups in our, in our, in our world. But the, I guess there's two ways that we've used AI.

So one. Is in ter in terms of like enhancing the existing experience of highlight users. So for example, if you're on a session, we have an AI feature that will actually like summarize the session for you. And we send like an email digest every week by default. But you can turn it on such that it actually will describe what a user did in that session before you click into the email, for example, right?

Mm-Hmm. . Mm-Hmm. . So there's a lot of things about like. Understanding the user journey and the model actually can tell you that, Hey, user clicked on this button, but then had some user frustration at this timestamp. So that's kind of cool. It makes it easier for you to triage issues when it comes to your sessions.

The second thing that's kind of similar to that is like on errors. So if you have like an error and a stack trace, we will actually suggest a fix to your stack trace. So we'll say, Hey, on line five of this stack trace, maybe Instead of using nil use actually initialize the value or whatever it is.

Right. Because that's probably where that air is coming from. So those are kind of things I guess. And there's a few other examples that we've used to kind of augment the existing spirit experience of highlight. And then one thing that actually we're launching next week is more of like making highlight, creating highlight power users faster.

And so one thing we've added is the ability to convert. Plain text queries into structured like into our structured query language. So in our log viewer, for example, if you say all logs with level error from three days ago to yesterday and click enter, we will actually convert that plain text query into like a structured query that you can then like run against your log viewer.

So yeah, lots of cool things on the AI end that we're thinking about. Thanks a lot. And I guess, yeah, they're, they're, they're though in two places, right? One is like making it faster for you to get value out of a highlight early on and learn the query language and things like that. And then the second is more like augmenting the experience.

And I actually think that maybe, and maybe this is a little controversial. I think the former is more powerful at least in the monitoring realm right now, because I think that the whole monitoring world is a lot about, it's a lot about accuracy and exact, right? Like you want. You don't want people like summarizing your log lines, right?

The reason you log something in the first place is because you want to know at what timestamp it happened and what exactly was being said. So I, I think it's, we're still early on in figuring out exactly where AI lands in this, in this world. But I do think we're making a lot of progress. So that's pretty exciting.

Jonathan: Yeah. I'm curious what what language is Highlight built on? Please, please don't tell me it's Java.

Jay: It's not Java. It's not Java.

Jonathan: Somebody emailed me. I said something slightly derogatory about Java, and somebody emailed me like, hey, by the way, we do Java if you want to talk about it.

Jay: He's picking a fight.

He's picking a fight with you. No, no, we write Go and JavaScript. So, like, our front end's all in JavaScript, and then our back end is, I think, exclusively in Go. With a little bit of Node. js here and there on certain testing things, but yeah.

Jonathan: Is it, is it mainly built on, on Node. js? Like, what's the, what's the framework that, that highlights built on top of?

Jay: Well, yeah, I actually wanted to talk a little bit about that too, but, The framework, like the web framework, is that what you're talking about?

Jonathan: Well, I mean, so the whole thing, is it, is it primarily a Node. js project? Or is it like written from the ground up with the Go backend? Or, you know, is it a, is it a module that runs on top of Apache?

What have you? Got it.

Jay: Yeah, yeah. It's written from the ground up in Go, actually. So it's essentially like a, a few services written in Go. We have like a main service that collects monitoring data and things are like buffered in Kafka. And that's kind of one of our biggest, like, I guess databases that if you want to call it a database that we use before things go into the rest of our system.

And then we have like a set of services that read from Kafka and then write to ClickHouse, which is basically our actual data store that we use at the end of the day. So those are maybe the two big projects that we use and depend on, on the open source world. And then I think one thing that's might be worth notice note noting on the open source framework side of things is like are you, are you all familiar with open telemetry?

Yeah. I was just going to, that's what got

Aaron: me down the road of the next, how you're doing integrations and does this work with Grafana? Can I set up dashboards anyway? Yeah.

Jay: Yeah. So that's one thing actually that maybe Aaron, you know, a lot about because I think a lot of the bigger monitoring vendors have started to kind Talk a lot about open telemetry, right?

It's like one of their biggest marketing marketing verbiages when it comes to putting the word out. But I think actually it's really beneficial for us in particular, a small startup, because it's made it really easy for us to be able to support so many languages. So for example, a few months ago we were at a conference out here in Seattle called Microsoft build.

And obviously there's a huge ecosystem around T net, right? Mm-Hmm. And we don't have a T net SD K. So we showed up to this conference kind of a little like, let's see what happens. And turns out, one, we got a lot of meetings from the conference in terms of conversations and getting POCs up and running.

But two, we could just kind of point them to the open telemetry documentation. And now if you look at do nets, open telemetry docs, and highlight. It's actually just a few small steps above the OpenTelemetry documentation to make it like an easier experience on top of Highlight. So that is one project that I think we're really thankful for.

And yeah, I think the, the, the, the team at OpenTelemetry has been doing a great job in terms of Getting all of these languages supported and we're trying to help a lot on that front too, with kind of contributing back to the client side SDK as well, and kind of I guess, yeah, promoting that environment for the most part.

So it's been awesome. Yeah,

Aaron: that's cool. For those that don't know, I mean, open telemetry is a, is a project that actually started out as open tracing, I think, but it was an open census. Oh, OpenCensus. I forgot about OpenCensus too. Yeah, Yeah, yeah. They combined them. So anyway, it was, it's really an open source way of getting information about well, really anything.

But, I mean, things that you want to use tracing for. So mostly, like, applications and things like that. And so, again, it provides this common standard, right, that you can use. So it's like, oh, your application supports OpenCensus? telemetry. Great. We support open telemetry. We can ingest that data or whatever, do whatever with it.

And you can do it in an open way and you don't have to, Oh, I've got to pay a license to, I'll just call it one of my former, you know, app dynamics, let's say, right. Yeah. Which is Cisco. I don't know if they've renamed it now, but you know, I've got to pay a license to them in order to get that data into another application so that I can use it or read it or massage the data or whatever, analyze the data.

So it's just a, it's just a great leveler, right? And levels the playing field for people that are doing that. So that's great. I'm glad it's working out for you. I know that we didn't use it in the companies I've worked for in the past. We didn't use it for so long because of all the turmoil and customers are like, we're really excited about open telemetry, but we're not quite ready yet because there's things going on.

We're not sure what's going on. There's a lot of uncertainty.

Jay: And I feel like you're always kind of going to have that problem to an extent, right? There's always be like the long trailing. That's the long trailing elixir and Erlang and all these languages that like, you know, they're a little far behind on the open telemetry bandwagon.

But I think for the most part, actually, it has a lot of coverage nowadays and you're seeing it in new rel, a lot of new relic and what they say. There's a few other companies that are really kind of milking, milking that the concept of open telemetry and helping educate people. And honestly, it helps us too, because it's like, if you want to try out something.

On the application monitoring end and kind of go away from your existing provider. It's it's a good option to kind of try things out So

Aaron: yeah, 100 percent avoids that lock in. Like I was saying, we're exactly exactly

Jonathan: So this this is a good place to jump in and ask something it I I want to kind of dig around a little bit in the the different things that you support in that you can put highlight on top of and I'm gonna, I'm gonna throw out some probably weird examples, but I think it'll, it'll kind of maybe put the boundaries on the space of what you're, what you're looking to do.

So, can you install highlight on top of a, WordPress website and get information about people's interactions with WordPress. Can you do it on top of something like a Tari app? I'm not sure if you're familiar with that framework. Can you do it with something that's running like a flask, Python flask back end?

Can you do it like, so just some, some weird things that I'm curious and maybe, you know, maybe OpenTelemetry helps a lot with all of this, but I'm just kind of curious of like where the boundaries are, where this would make sense to try to use.

Jay: Yeah, I'm not familiar with Tari, but I feel like, is it T A T A U

Jonathan: R I?

It's, it's taking a web app and making an application out of it.

Jay: Yeah, makes sense. So I guess what I would say on that front is like the, the, anything that runs in the browser can be, can install our client SDK. So we do have customers that use highlight in like Weebly and WordPress and Squarespace type applications.

That might hit like a, a, a proprietary back end that maybe another team at their company is using. And the power there is that like, if I have a front end team that's like using this no code tool but someone clicks a button and something breaks, they can actually attribute it to what might be happening on the back end that's going wrong, right?

So yes, the answer is yes to the first WordPress thing. Tari, I'm not too familiar with.

Jonathan: There's a, there's a, there's a more common solution than Tari. And I, I am absolutely blanking on the name of it. Is it

Jay: Electron?

Jonathan: Yes, of course. Electron.

Jay: Yeah. So we do, we haven't, we have actually Electron documentation.

And because that runs a web browser in a desktop app, it's a very good fit for highlight actually. And so, you know, companies like Notion and. linear and those types of tools could definitely get benefit from highlight. And then the third thing was a flask app and yes, that's our bread and butter.

So we, our, our CTO is a big Python guy. And so we have a lot of, we have a lot of tutorials on getting this up and running and fast API flask. And a few other like common Python web frameworks.

Jonathan: Cool. Is, is there any, is there any of this that you don't support that you're kind of thinking, man, it would be nice if we could run on, and I don't know this space well enough to even throw a name out.

Yeah, I, I threw my, my big guns out and you, you already support them. I passed the

Jay: test.

Jonathan: Yeah, I guess so.

Jay: No, no, I would say yeah. One thing that we're eyeing a lot is like mobile, mobile recording. It, I think the reality is it is a very different ballgame though. Like, you know, with our 10 person team, I think we would need to have actual iOS engineers on our team or, or Android engineers on our team to be able to even kind of start to, to build that.

So we are thinking about that. And I think every once in a while we'll close a customer that is like, Hey, When you turn, when you, when you get mobile recording working, let us know, you know, and so we'll see, we'll see if we get like how fast we get there, but I think eventually we'll get there and I think that's the next maybe milestone that we're looking at for sure.

Jonathan: Yeah, that's interesting. Is, is there anything else on the radar that you're, that you're looking to do either short term or long term? Want to chat about?

Jay: Yeah, I mean, I feel like the big, the big thing is maybe mobile recording, the other like smaller things like. I guess we have like a big launch week next week.

And I'm actually working on recording a bunch of videos for that. So y'all are helping me practice for the video. Yeah. But yeah, like we're, we're launching a bunch of really cool things. For example, the AI stuff I mentioned for building queries. We're launching a like metrics drill down product or feature.

Where like on an actual metric, you can actually click into it and see all the sessions related to it. So if you're like visualizing users grouped by a specific URL, right. If you could click on one of those bars that you've, you've created, and then you can see all the URLs or all the sessions of those users on those URLs.

It's pretty cool. Same with logs and all those kinds of stuff. So we're kind of really milking that like cohesion feature, like connecting everything in your, in your product. And then I think even more longterm, the other thing we're really excited about is like more product analytics and helping people understand user engagement in their product and still sticking with the application level.

But, but going more towards like support and the, the product type folks at an organization. So yeah, a lot of, a lot of fun things. And if anyone's interested check out our Twitter channel or our Twitter account for next week, cause we're doing a bunch of launches and that could be fun.

Jonathan: Yeah so if somebody wants to get started, wants to get it installed, is there a quick start guide somewhere you would point people to?

Jay: Yeah, I mean honestly if you just go to highlight. io slash docs That's the best way to start. There's a get started button there, which will take you to all of our language specific docs.

And then you just pick your language and get started. Alternatively, you can just go to our actual app and click sign up at the top. And that will also, it also embeds the docs. So it's pretty easy to get started. And, and if anyone has questions, we also have a discord that they can kind of jump in and ask questions about.

Aaron: Yeah, that's nice. It's always kind of a pain for me when you go. You want to do like a free trial of something. I mean, I'm assuming if you sign up, I mean, you just get the free version or is, or do you get a trial of some of the more advanced features?

Jay: You actually get everything. Yeah. The only thing you won't get is like, yeah.

Like you only, you won't get like our back and SSL. So yeah, you'll just be able to try it out. And it's very hands, hands off in terms of having to talk to us and things like that.

Aaron: Yeah. But anyway, what I was saying was easy, easy access to documentation as you're going through the process, like the onboarding process for people is super important.

Yeah. And I hate it when I do, Oh, free trial. And then I have to open up several browser windows and try to figure out, you know, and you're going back and forth. Oh, here's a PDF. Now I've got to like read a PDF. Yeah. So yeah. Yeah. No, it's nice that you've done a little document integration. That's great.

Jay: Yeah. Yeah. And we're, we're always trying to make that better actually. Yeah. We're this upcoming few months, we're going to be doing some work on, on the docs to make it a little more language specific rather than product specific. But I mean, nonetheless, it's very easy to get started. And hopefully folks can kind of jump in and get some data flowing.

Nice.

Jonathan: Aaron, is there any, any other questions you want to cover before we start to wrap?

Aaron: I mean, I could go on. I mean, you know, I could go on for a few hours here. We could talk about customers. We have a couple, we have a couple of

Jonathan: more minutes before it's really time to wrap. So if you, if you have a couple more queued up, then go for it.

Aaron: Well, I was going to ask, like, is there, I was looking at the customer page cause that's always, it's a good place for people to start. Like what are customers actually saying? And I'm in marketing, so I know, you know, there's, you know, some marketing magic that happens there, but But I do have like a customer story, I guess, like that where someone has used this and then come back and said, wow, this really helped us do X, you know, we couldn't do it before we were having this problem.

And wow, this really helped us get over just to kind of illustrate to people, like the potential, I guess.

Jay: Yeah. I mean, honestly, if you go to that page, Aaron, that first customer's testimonial from that healthcare company is a great example of like a very large company that I guess historically has been a little more slower in their ways that has kind of used highlight to actually get a lot more visibility into those older applications.

And I think it's actually a function of the fact that they can install this at the application level. And they don't really have to touch a lot of the infrastructure stuff. Right. And in those cases, I feel like it's awesome because they from, from having just like logs, for example, just logs and metrics.

They're being able to kind of move to using more modern technologies like next JS and all of like sort of like JavaScript isms that are kind of becoming a lot more popular today and be able to monitor that monitor that with something like highlight. And then at the same time, actually connect that to those older services that these, these, these applications depend on, right?

So that's kind of like, I guess, a very, like a, almost like a perfect enterprise story in terms of folks being able to get value. And then on the earlier end, I feel like with, with smaller companies, we do have a lot of customers in like the modern JavaScript ecosystem that use us. And it's a lot more, it's a lot, a lot, it, a lot of the value comes around connecting the client to the server, like I mentioned earlier, right, where if I use another tool, I'm using a bunch of like, I'm using four siloed tools that are not very well integrated.

And with highlight, I can kind of get a more comprehensive picture of what's going on when I'm debugging an issue.

Mhm.

Jonathan: Yes. So we do have a bit of a live chat room and we got a question just now from I happen to know a flask developer that will probably be very interested in this. And he wants to know, are the front end javascript plugins trapped or blocked by default by any other plugins?

And he mentions privacy badger, ublock, etc. And that's, that's, that's interesting. Like what's the, what is the interplay between the highlight tools and some of these other like privacy focused tools?

Jay: Yeah, that's interesting. So we were, I guess we, when we started highlight, we were we were not looking forward to the day that you block considered us a, a, like yeah, like an ad snippet or whatever they call it, right?

Unfortunately, you block now considers us an ad snippet and blocks us. But to, to, to solve the, for that, like we have a few options, right? One is you can actually like. Proxy your highlight requests through your domain. So if you want to run, let's say you're, you work at acme. com. You can proxy all of your outgoing highlight requests through highlight.

acme. com and you just need to share with us some DNS credentials and we can get that working. And then we even have like a more hands on approach where we will actually proxy it through like a cloud flare worker, all the data. And that's a kind of an even better way to, to sort of mitigate these blockers, but it is a problem.

It, it, it really is a problem. And we are we think about it a good amount, but it's honestly, at the same time, I get why a lot of these blockers do this, right? Like I think more often than not, they're doing it for a good reason. Right. And we can, we can solve this, especially for the larger customers that have a lot of traffic on those, those two notes that I just mentioned.

So, okay. So I'm just saying, so if

Jonathan: I'm, I'm real curious with, with something like uBlock I don't know much about the people that run the, the uBlock extension, but have you, have you heard reaching out to them? Like, Hey guys, I see that you're blocking us. We don't actually do advertising. Here's what we do.

And I've just, I wonder whether that would bear fruit or if they would ignore you. I honestly don't know.

Jay: I haven't tried. I haven't tried. We haven't tried. Our team hasn't so we should try. That's a good point I I feel like the only problem though is I think their definition is not super Strict in terms of why they would block something like this and I think they do it based on amount of traffic

Yeah, but

you're right maybe Some reasoning and some background on the project and things like that would help kind of make the case.

So that's a good point. It's a good point. I'd be, I would be interested to

Jonathan: hear what

Aaron: happened. Yeah. Yeah. Yeah. Would this also hold true for other ad blocker technologies and, and plugins like ad block or like I use Piehole would Piehole, Piehole block? I guess they would, would they?

Jay: I'm not sure about that one specifically.

I know uBlock we are on like a list. And the thing about these lists is they're like open source, right? So you can find them on GitHub and people basically submit them. And the thing is, it can't, it's not necessarily the company that's using Highlight that's submitting them. It's right. It's their end users.

So if a company has like 5 million users and they're all, all of those users are sending data technically to the company via Highlight. If one of them complains, uBlock might like consider that. You know what I'm saying? And a lot of these GitHub repositories that have the data in them are shared. So I do think that uBlock is not the only one that pulls from that list.

So Aaron, it might be that. Piehole, for example, pulls from the same list, and they just run a job every month to kind of keep, keep track of what's going on on that front. So

Jonathan: yeah.

Jay: Yeah. It

Aaron: reminds me. I'll just set something up and then test for it.

Jonathan: Yeah. It reminds me very much of email servers.

Ending up on spam blacklists and you know, if that happens you let's see I think so spam house is the big one and they've got it automated because this happens all the time They've got it automated you go to their website and you say here's my you know, my URL or my IP address Please remove me from your list And if you're actually sending spam, you know, they'll let you do the remove requests like four or five times and then of course you get like blocked forever.

Yeah, yeah, yeah, yeah. Makes sense. I wonder whether these, you know, advertising block lists, whether they are sophisticated enough to have some of that tooling. There have to be, there have to be domains and scripts and such that end up on there that shouldn't be. Yeah.

Jay: Yeah. It's a

Jonathan: good

Jay: point. That's a good point.

And, and I think it honestly is a good learning for you guys to tell me like, Hey, just reach out. Cause we're, we're not doing it. We're doing it in good faith. But at the same time, it's like, honestly, not a huge deal. Like it's like the companies that really care about this and they're like, Hey, we're missing these sessions.

They'll reach out to us and then we'll help them out with that proxy stuff that I mentioned. And it's not a big burden. But I imagine, you know, right now we have like 300 customers. We have 3, 000 customers. It becomes like a bit of a different story. So, yeah, yeah, yeah.

Jonathan: Okay. I've got a, I've got a similar question to the one Aaron asked.

He wanted to know a success story. I want to hear a weird story. Now, like ask, asking people, what's the strangest thing that someone has done with your product? So where, where has somebody used highlight that just surprised you or you find odd and not, not every project has one of these stories. But if you, if you have one, I think it'd be fun.

Jay: Yeah, so I don't have a story of a customer using it in a weird way, necessarily. But, one thing we used to do in the early days when we were getting our first customers, is we, So Highlight is just a JavaScript snippet, right? So you can run it in the background in the browser. And so we would actually go to our prospective customers, and we would go to their websites.

And we would inject the highlight snippet on their site. Oh

Jonathan: yeah, something like a monkey script, essentially.

Jay: Yeah, and so like, we would go to the Floss, the Floss website, and we would install highlight, and then I'd send you an email and be like, Hey, We just installed highlight on this website. Wouldn't it be cool if you could troubleshoot what was going on for all of your sessions, you know?

And so that kind of worked pretty well. It wasn't very scalable because it's just like, I think at the time also session replay was pretty new in the engineering world. So it was like kind of a ripe time to be doing it. But, but yeah, it was kind of a fun story of using highlight to sell highlight, you know, I like it.

And yeah, it's kind of a fun, fun, fun fact,

Jonathan: but okay. So we are getting back towards the bottom of the hour. I want to know, is there anything that we have not asked you about that we've not covered that you wanted to make sure and let folks know about? I know I asked, I asked all the hard questions here at the end.

You got to really think about these.

Jay: No, no, you're good. You're good. I think honestly, just I would love folks to be keeping an eye on what we're doing. We specifically on the open telemetry note We put out a lot of content and webinars and stuff like that about OpenTelemetry. So would encourage people to follow us on, on, on Twitter, if they want to kind of keep an update on that sort of thing.

And then, yeah, if any of, any of the folks from the Hackaday discord want to give us feedback on the product. I will welcome that very openly, so would love folks to, to keep in touch on that front, too. All right. Do

Aaron: you have, like other, other avenues? I mean, not everyone is excited about X, I guess, we have to call it these days, right?

For, for various reasons.

Jonathan: The social network formerly known as Twitter.

Aaron: Right, so I'm just thinking, like, you mentioned Twitter a couple times, like, is there, have you thought about, you know, You know, Mastodon or even like Hey, I'm interested enough. I want to join a Slack group, for example.

Jay: Yeah. So our discord is a good place to be if folks want to kind of just learn about what goes on, on the content.

And we do, we do pipe all of the. stuff and any of the stuff that we post on in discord as well. The second thing I would say is LinkedIn. We're pretty active on too. Those are our two big social channels. Mastodon isn't something we've looked at, but maybe we should honestly it's not too much effort to kind of pipe things to another source and yeah, that's a good call out.

Yeah. I mean,

Aaron: I don't know. I've surprisingly mastered on it. It's not. I don't think it's growing as much as they would, you know, people would like, but the, my followers on Mastodon are pretty pretty active. Okay. They're active. So that's good to know. And like you say, you know, you just use a tool like buffer or something and you just post it at the same time.

It doesn't, it doesn't really hurt. It's a

Jay: good point.

Aaron: It's a good point. And it helps support the ecosystem. I think it helps support alternatives for people that might not jive with

Jonathan: Yeah, yeah, yeah, yeah. Of course, of course. That's fair. Alright, so I've got a couple of last questions that I've got to ask before we let you go.

And that is, you personally, what is your favorite text editor and scripting language? Oh man,

Jay: I'm a, I'm a Vim person. Okay. And I'm a, specifically I'm a NeoVim person. Okay. Yeah, me too. But nowadays I do use VS Code because I don't code as much. And so I use like the Vim plugin in VS Code. Okay. Which honestly has gotten really good since I, since I looked at it, like, I guess four or five years ago.

So, yeah, that's my, that's my that's my text editor. What was the other question? Scripting language. Scripting language. Oh, I'm a Python person for sure. Yeah, yeah.

Jonathan: Yep, very cool. Nice. Alright, well, Jay, thank you so much for being here. Sure, appreciate it. And had a lot of fun learning about Highlight.

And yeah, of course. Yeah. Thank you so much. Sure.

Jay: Thank you for having me. This was really fun. This was really fun. Yeah, it was a blast. All great questions. And I, and I yeah, excited to keep in touch too. Yep. Awesome. All

Aaron: right. What do you think? That's great. This is, this is, this is fantastic. I think it's, you know monitoring, be it application or infrastructure monitoring is a crowded place that's had a history going back to the early days of computing, right?

I mean, you could, you could really go all the way back to the early hackers at MIT that were doing things with you know, PDP machines and stuff like that if you really wanted to, right? Because they had to get, you know, that same telemetry data, although it was a very different world. But so it's interesting to find someone that's doing something different in this space or trying to help solve a problem in a different way.

Right. Cause a lot of people will come along and they say, Oh, we're new and exciting, but they're actually not doing anything different. Right. Except maybe they're a younger company and they're charging you less money. So yeah, to hear someone that's tackling this problem from a user perspective and.

You know, really trying to meet developers where they're at with things like Hey, you can use this without necessarily having to bother your DevOps people or your sysadmins to install some complicated thing that they're going to be like, no, I'm not doing that because that's not in our architecture.

So, so to be able to do some of those things is actually really, really cool. And it does solve a problem that developers have. So yeah, I, I like it. I mean, I, I'll have to actually play with it. I didn't we were talking so much, you know, a lot of times we'll actually try this stuff out while we're talking, so I didn't have time to sign up and do all that kind of stuff, but yeah, I want to play with it and just compare with other products that I've either worked on in the past or used in the past to see how it works.

Jonathan: Yeah. When I, when I scheduled you for today, Aaron, I didn't realize you were such an expert in the space. That that worked out really well.

Aaron: Well, yeah. And I didn't elaborate either in my email. I just said, this is right up my alley. I think it was my response, but yeah. I mean, I've actually worked in the, in this industry specifically for.

Whatever 10 years or something at this point, which is a long time. It seems like and so, yeah, I've used, I've actually worked for these companies that produce these products and actually used a lot of the other products as well. So yeah, it really is right up my alley and you know, was something that was not on my radar, but it will be now I'm going to go follow him on on LinkedIn right now.

Jonathan: Yeah. It's fun. It, you know, for the past few hundred shows, it feels like when I'm on or now that I'm, I'm the main host, I've, I've been the play by play guy and the other person has been the color commentator. And today I get to be the color commentator and you're the expert. You're the, you're the play by play guy.

It's kind of a neat inversion of the roles. That's funny.

Aaron: Oh, I hope I didn't go too deep on anybody for anybody, but I hope it wasn't too much inside baseball, but it was, it was really interesting. Like I said, I could, I could talk and talk and talk like I'm ready to sign up for a day session, right?

Where you go through everything and talk about it. And I went to their careers page too, cause I'm looking for a job right now. They don't have any product marketing openings, but maybe when they get a little bit bigger definitely I will be checking in from time to time.

Jonathan: Yeah, interesting. That, that could, that could be something.

All right. You never know. It's happened before on the show. It is. Yes.

Aaron: Yeah. Yes, it

Jonathan: has. Yes, it has.

Aaron: All right. You have anything you want to plug? Well, of course you should go to my two YouTube channels and check those out. So there's RetroHackShack and there's now RetroHackShack After Hours, my second channel, which isn't quite as scripted, although it's still really interesting.

So for example, all of my e waste Wednesday videos, Wednesday, one of the reasons I can't do, couldn't do this show when it was on Wednesday was because I would go to e waste in the morning and I can only go on that day at 10 o'clock. To be able to buy stuff from this e waste place. And so So what I do on my second channel is I bring that stuff back and just kind of like do an overview.

Here's what I found today. Isn't this cool? Whatever. What I recently did on RetroHackShack After Hours is I actually filmed the whole process. I actually took my phone and, and, and took video and audio and stuff and showed people, here's what an e waste place looks like. Here's what you, I can find at mine.

And here's what I pay for stuff. And, you know, here's the bin of you know, cards, ISA cards and PCI cards and things like that, that they sort out just to show people what the experience is. So in either case, go check out my two channels on YouTube and subscribe if you're interested in that way, you'll always get that content in your feed.

Jonathan: Yep. Awesome. Thank you, sir, for being here. Absolutely.

Aaron: Anytime.

Jonathan: Yeah. So I do want to plug, of course, Hackaday. We've got the security column that goes live there Friday morning. And then of course, Hackaday is the home now of Floss Weekly. We sure appreciate that. We are actually looking for a guest still for next week.

So if you have someone or you Are a project and want to be a guest, let us know. It's floss at hackaday. com. And then we've got the folks from life Ray on August 6th. That's two weeks from now, looking forward to that as well. I think that is all of the housekeeping, all of the notes that we have.

And so we'll, we'll let you go. Thank you everyone that is here. Those that caught us live and those on the download and hate, we will see you next week on floss weekly.