From Informix-4GL to the Future: A CIO's Guide

Kiz — We're getting pretty close to the quorum. Just a few more seconds, and then we can hit start. And I think we're there. Okay. Welcome, everybody. Thank you for being here with us for Informix 4GL into the future, a guide to modernization without rewrite. My name is Kiz Syed. I am part of the Four Js team, and this session is for any CIO, CTO, CFO, CXO, anybody working with mission critical applications and thinking about, hey. How do I modernize this? Should I modernize this? I'll be joined in just a second by my colleague, Mike Folcher. But first, just some ground rules for this session. Number one, all lines

Kiz — have been muted just this is just so we can keep the flow going. We will have a recording that we will share of this session after it concludes. And then lastly, please get your questions ready. We've got our resident expert on everything from 4GL to database migration, Mike Folcher here, and and he will answer your questions either as we get them or if we have time at the end. And I think this is a good time to introduce to introduce you, Mike. Please go ahead.

Mike — Alright. Thanks. Yeah. So I manage the professional services business in US and Canada for Four Js. So excited to be here and talk about modernization.

Kiz — And I think your viewpoint is going to be very helpful for the folks on the call, Mike, because you've had experience as a developer, as a develop man development manager, R&D product strategy. So you're really able to marry the technical side of things, but also the business impact. And so really interested to hear your perspective on a lot of the topics we're gonna talk about. Now on the right hand side, you'll see the agenda. Really three big things we wanna cover. Number one, what are some of the pain points with Informix 4GL right now, especially as you think about a modern

Kiz — tech stack and user expectations for what a tech stack should be and should feel like? Number two, what are your options to modernize? And then lastly, what are your next steps? How do we move forward? Okay? Before we get to that, just wanna give you a quick overview on Four Js and who we are. For the last thirty years, we have been helping organizations modernize their mission critical apps. Any device, anywhere, zero rewrite. We are a part of Constellation Software, and we've got a customer base that is international. We've got a team that is international. You'll see just a

Kiz — sampling of our customers here, everything from health care to Wall Street, everything from professional sports leagues to retail, from ERP to WMS. We've got it covered. We've got teams in Latin America, in Europe, in APAC. As you can probably tell, Mike and I are based in The United States and Canada. Alright? So that's Four Js in a nutshell. Let's get into it, Mike. What are the big pain points with Informix 4GL?

Mike — So the I see really two major pain points within Informix Informix 4GL. The first one is so most of the these applications are legacy applications, which in turn also means that they have legacy business processes. Like, I've seen a lot of companies where they're running reports and then they're pulling the reports off the printer, and then they're doing the next step in the workflow. And I think one of the big things is because there's a lack of data and there's a lack of UI experience in Informix 4GL n order to really streamline those business workflows and processes. So that's one

Mike — of the things I see is a lot of inefficiency from a business workflow perspective. And then the other thing is just resourcing. Right? It's difficult to find and train Informix 4GL developers. You pretty much have to build a lot of your own training materials. And then the other aspect is that then users as well. So you have to train users. They're using a green screen application most so when you hire new developers or new you hire new users you'd have to do a lot of training because they're not used to the tools that are being used

Mike — to build and use the application.

Kiz — Interesting, Mike. So you're almost saying, hey. It impacts the entire funnel or the entire almost product life cycle from the creation of it all the way down to the user. And so, really, you can think about it as time to market for creating an application and then time to value for either the customer or the end user.

Mike — Correct. So, yeah, I mean, you have a lot of these organizations. Right? And they have people that have been there for years and years. But, eventually those people are going to retire, and you have to bring in new resources and you have to train them. And so the training process is a lot more difficult, and then you have retention issues, right, because you're using really old technology and the user interface is really old. So you have the kinda it's both sides, both the development side and the user based side.

Kiz — That's and I know you're gonna talk in in your case studies around a real life example that you've had with this. Just really briefly, what are some of the impacts you're seeing for the business from this kind of Informix 4GL reality? Because I know we've got some business folks on the call right here.

Mike — Yeah. So I think the biggest thing is you have a lot of inefficiencies. Right? You can't show a lot of data on the screen and then the workflow from going from one state of a of data to the next state sometimes has multiple steps and processes where if you had a modern infrastructure and a modern user interface, you could do it all in one place from one page or one screen and execute and achieve the same goal. Right? So at the end of the day, it's about making your staff more efficient and streamlining your processes.

Kiz — Well okay. So bay basically, doing more with either the same or with less. I'm sure CFOs, like hearing that. Let's move forward then, and Mike can talk about alright. Here's some of the challenges with Informix 4GL. What are the options to modernize?

Mike — Yeah. So we really have three different buckets here. So we have the manual rewrites, then we have kind of a hybrid approach. Right? Rewrite, but AI assisted, and then we have using Genero.

Kiz — Okay. Alright. Let's double click now into into each of these options. Let's start with manual rewrite first link.

Mike — Okay. Yeah. So the manual rewrite is basically gives you a chance to start over. Right? And you get to use the latest technology. Developers love this route. Right? They wanna work with the latest, greatest technology and the the newest things. You get a really sexy interface. But from a business perspective, this is very costly and it really incurs a lot of risk. You'll need to reengineer basically all of that domain specific business logic into a new technological framework. So a lot of times I'm assuming most of these applications, you don't have documented requirements for all of the features and functionality in your application. So you basically have to reverse engineer what's

Mike — in the application, write it down into specifications, and then reengineer it in a new technology. And this also takes a lot of iterations really to get it right. And Oh, okay. So you're not gonna hit the mark the first time. Right? So you're gonna need a lot of iterations, and that really takes time and, obviously, then money as well.

Kiz — You know, Mike, I know you're a developer at heart because of the excitement that you started talking about doing everything from scratch. I I know where your interests lie, but we've seen this with so many customers that that do try to do a rewrite. It is often over budget, over time, and not just by a little bit, usually by an order of magnitude. We're talking multiyear, multimillion. Okay. So those are the risks. Right? What are some of the benefits or reasons why you might wanna think about rewriting?

Mike — Well, it gives you I mean, it gives you a chance to start over. Right? So if you can improve some of those maybe data structures or business processes, And if you so if you have a very limited set of business requirements, then it's probably not as much of a challenge to rewrite. Right? Because that's really where the complexity comes in. If your system is super complex, right, you have to almost like I said, you have to extract all of those requirements out. So if you don't have a lot of requirements to extract, it'll be you'll be able to quantify it much more. See,

Mike — the projects that when you're doing a rewrite, they're super hard to estimate and to quantify because these applications have really evolved over thirty plus years. Right? So people don't even realize all of the baked in requirements and business rules that are in these applications.

Kiz — That's a great point, especially for mission critical applications. They've been around for so long, and and you're afraid to rearrange any of the chairs on the deck because if something might go wrong. You know, I think of the example of Bret Taylor rewriting Google Maps in essentially a weekend. I assume that is not the case study for most folks. You think that's accurate?

Mike — Oh, yeah. I mean, too is you're looking at probably an application that has very rigid and strict requirements. It's probably developed in a framework where everything was documented. So you had all of that knowledge already in codified, and then you could just take it and reengineer it.

Kiz — Yeah. And you make a great point around documentation. I think we've seen countless examples of not having documentation in place, and that makes this manual option even more challenging. Let's move forward now to the topic du jour AI. How can we leverage this to do modernization?

Mike — Yeah. So, I mean, AI obviously is the big buzzword these days, and it can really speed up the development process, and the AI engines are really getting better every day. Right? I think there are some challenges here, though. I think the first thing is with just the Informix 4GL language itself. Right? So AI needs large amounts of data and datasets to really be able to work effectively. Right? And there's just not these robust dynamic examples of Informix 4GL out there. I think that's really concern number one. And then the other concern is that you're so, basically, Informix 4GL

Mike — is a procedural language. And the chances are almost all modern languages, right, are object oriented. So this you get a real problem here because if you're going one to one, right, which AI is going to do is you're gonna end up with object oriented code, but it's gonna be structured like procedural language. So you're gonna get into a situation where it's gonna be really hard to extend and maintain this code because it's gonna be object oriented code that no programmer would ever write because you would have these huge monolithic functions because it needs to map the way the procedural application worked.

Kiz — Okay.

Mike — So the yeah. I think so the more likely scenario, I think you'd go back to kind of the what you do with the manual rewrite is that you kind have to pull out all the requirements for that application, and then you would need to feed them into an AI engine. That's really the only way that that situation would work.

Kiz — Wow. So this almost feels like a red herring or in a deceptive option where you think, hey. It's gonna be fast. It's gonna be overnight. But in reality, it's a pseudo manual effort.

Mike — Yeah. I mean, you're still got I mean, I think AI is gonna get you maybe, like, 70% of the way there, but you're still gonna have to have developers then go in and make changes to it to get it to be production ready.

Kiz — Alright. Let's talk now about the third option, which is which is Genero. This

Mike — is this is really gonna be your most cost effective and risk averse methodology because you can really preserve that business logic, those domain requirements that are already baked into the application, and then you can just really modernize just the user interface and the user experience. You can also modernize incrementally. Right? And retraining your staff isn't as big of a lift because Genero is really just a super setter. It's really an evolution of Informix 4GL. Right? It's a modernized version of that. And so in addition so Genero is baking in AI capabilities as well, so you'll be able to leverage AI with the Genero

Mike — platform.

Kiz — What does the timeline for this look like, Mike? You know, I know in the past with customers, we've seen something on the order of three months to to nine months. But what do you think is a reasonable timeline? You talked about maybe something being iterative. Help us understand what does implementation look like.

Mike — Yeah. So I think a lot depends on what is the most important thing for you to modernize. Right? If you want the UI UX to be modernized, that's gonna take depending on how big of an application you have, that might take six months to a year. If it's the if if you just and these steps can be done incrementally. Like, typically, what we would do is come in. We do an assessment. Right? But we can we can switch you over to Genero and staying in terminal based, we can do that in a matter of weeks. Right? Okay. And then it's really a matter

Mike — of, okay. Let's take some of the pain points. Let's take some of the hard trafficked application areas that you have, and let's really modernize them, make the workflow streamlined, and really improve efficiency.

Kiz — So you can take iterative steps at your comfort level in order to get to an evolutionary outcome somewhere down the road.

Mike — Yep. Yeah. So you start pulling off pieces of the application incrementally and modernizing them.

Kiz — Mike, what happens we have a question here in the Q&A from Kathleen. What happens if there is any kind of issue and you do wanna roll back? Is that possible?

Mike — Yeah. So the nice thing is so typically in this model, right, you're staying with the same database. Right? So you can run both systems in parallel. So you can have the the green screen Informix 4GL application running, and then you can modernize that into Genero. And then if you need, for whatever reason, that there's a bug in the Genero application that you need to fix, you can roll back and use the Informix 4GL application that you still been using, fix the issue, and then move users back onto the Genero application. So the idea is that you can

Mike — start moving users over slowly. You can take subsets of users and point them to the new application. So there's a bit there's a lot of different ways to implement the solution.

Kiz — How would a rollback type option work with either the manual or the AI options?

Mike — Yeah. So I think you're with both of those options, if you're staying with Informix, okay, you probably have a little bit more flexibility. But if you're changing databases, which a lot of times you are, right, because most people wanna if they're modernizing the application, they probably wanna get off of Informix as well. So you're looking at basically, like, a no turn back option. If you wanna try to sync the data back and forth, it gets super expensive, and it's very, very risky.

Kiz — And on that topic of database, I know a lot of folks will either look at some of the gold standards like Oracle, especially if they need to have high fidelity or they'll look to the cloud. Any issues on database with Genero?

Mike — No. I mean, Genero works with all your modern relational databases, and we have a lot of that's usually the two big pulls for going modernizing with Genero are user are is user interface, user experience, and then we wanna switch databases. And so that can also be done, in an incremental fashion. We've worked with a customer where we had the application work against both an Informix database and an Oracle database, And they had a period of time where they're they're live on Informix and then testing with Oracle. And then we switched them over, and then they were live with the Oracle. And

Mike — then but the data the application was still compatible with the Informix as well.

Kiz — Okay. Let's keep moving before. Just two quick questions that I saw. One is, hey. Can we help with documentation since that does seem like a core part of just having any good software life cycle? Yes. We can. Another question just around AI. Mike, can you just mention anything else that you have on AI or, Versus Code?

Mike — Yeah. So Genero has an AI interface — since renamed **Genero Intelligence** — and it allows you to basically give instructions to an AI engine and have it give you a prototype of an application. So you can do it with different sections of your application or a whole portal application that you would need to build.

Kiz — Ok. Let's talk now, Mike. What the we mentioned earlier the iterative steps you can take. Let's walk through each of them.

Mike — Yeah. So this is your Informix 4GL screen. So you can do a straight recompile in Genero and pretty much have the same functionality.

Kiz — Wow. Okay. And that's what that looks like?

Mike — So you can run it in terminal mode, but then if you wanna use the browser client or desktop client, this is what you're gonna get. So you can see here that all of a sudden now you have fields in a user interface that you can copy and paste.

Kiz — Yeah. And, again, just going from terminal straight to this. Wow. Okay. And what did you say that timeline usually looks like for that?

Mike — Like, if you just if you wanted to have the same look and feel and behavior but be in a graphical environment, I mean, we're talking about weeks.

Kiz — Wow. Okay. And then what happens what how do we get to here?

Mike — Yeah. So this is obviously a little bit deeper surgery. Right? So what we what we typically do when we get into this situation is that we analyze the Genero or the 4GL code, and then we pull out the business logic, and we kinda isolate that. And then we start tinkering and modifying the user interface and the user experience. Right? The user interface is mostly I mean, it's still in the same our similar format that Informix 4GL is, and then the user experience is really just a combination of dialogues, to to control the user navigation to the application.

Kiz — So this, does seem like a relatively low risk option because you can go backwards if you do need to, while at the same time just getting incredible, bang for your buck because you can go from, hey, two week to this relatively fast and with some pretty good historical guardrails on how long that's gonna take.

Mike — Yeah. And it really then gives you an opportunity typically in your Informix 4GL application, you have four or five or six different, screens that users have to navigate to where now you can kinda sew them all together in one page of a of a web application. So think about the user experience in that situation. Right? They don't have to go navigating all around through a menu options. Right? They have everything, one screen where they can do their work.

Kiz — Well and these users, they're used to the UX UI for Amazon or for Salesforce, and they're so they want something that looks similar to that.

Mike — Because then they don't need as much training. Right? They know how to use web applications. Right?

Kiz — Okay. Mike can you just walk us a little bit through next steps here and with for folks I think especially around the database migration part of it.

Mike — Yeah. So, I mean, typically, what we would recommend is, okay. The first step is we just take the application, and then we go to Genero.

Kiz — Mhmm.

Mike — Right? So that's like I said, that's a couple weeks of process. Right? And then we really we talk to the users and really talk to the business too. What is the most important thing for you? Is it getting off of Informix, switching to another database? Is it really the UI pain points that we need to improve? You know? And then and then digging deeper if it's a UI thing, what are what are the areas that the users really struggle with the most? What are the business processes that people are doing day in and day out in the application? And then taking those and

Mike — modernizing those, really. You wanna you wanna give carrots to your user base. Right? Give them reasons. Okay. We wanna go on this new platform. And then once you start there and you get start getting momentum, it's really just then taking those same lessons learned and then pulling the rest of the application up so it's at the same standard.

Kiz — And I think you raise a really good point, which is getting feedback, getting connected to the business. Right? Because, ultimately, these all everything that we're going to be doing on digital transformation is going to have a business impact in making sure that you have a team that can really understand what does the business need, how will we get there. It's not just about the tech stack.

Mike — Like, I mean, at the end of the day, these are solutions. They're solutions to business problems. Right? And so you can't you can't ever lose sight of that. Right? It's not about using the latest, greatest technology. It's really about solving business problems, making business more efficient.

Kiz — Okay. Talk to us about an example when you had to do that, Mike, when you were in that hot seat and you had to do this migration.

Mike — Oh, yeah. So this is an example. So customer in the public sector industry, and they had really two core applications. Right? So one was a student information system or as it's called in the industry assist. And then there was an ERP system, right, managing human resources, payroll, budget, accounting, accounts payable, all the financials that go into to running a local government or school district. So the so the idea was okay. It was at the time, it was the late nineties. And so since they were a software vendor, needed to compete in the market space. So they needed a at the time, it was a a

Mike — desktop application. And so they decided they were going at the time, they decided to go with VB6 because, well,.net hadn't been out that time. And then so they estimated, okay. It's gonna take about two to three years with about 15 developers. So they did a complete rewrite and ended up taking about five to six years. And then they ended up having to go to about, at the peak, I think, about 22, 23 developers. So then it they wanted the plan was they were gonna first do this the student information system and then do the ERP system. Well, they sat back and said,

Mike — we can't afford to do this again. Right? So they took their ERP Informix or yes. Informix 4GL application, and then they said, let's use Genero to do it. Well, Genero took nine months with five developers, and they were able to get to market. So in the meantime, the student information system hadn't penetrated the market as much as they had hoped, so they ended up taking the student information system application that was originally written in the Informix 4GL and converting that to Genero as well. So they had two student information systems.

Kiz — So I just wanna make sure I got the math on that right, Mike. You're saying they were able to do this migration within, what, a third or a fourth of the time with a fifth or fourth of the people? Yep. Okay.

Mike — Yeah. So it's and then so at the end of the day too, the student information system was in VB6. Well,.net had come out towards the end of the process. Right? So now they were in a legacy application with Microsoft. So they were kind of in the same or same or similar boat that they had been into before.

Kiz — Oh, so again, we talk about time to market and time to value, especially for anybody that's managing a P&L, that's managing budgets, whether, again, you're in IT, you're on the business side, being able to reduce both of those things, huge value unlocks for the business. Question just that just came in, Mike. You know, public sector, regulated industry, how does Genero deal with regulatory environments?

Mike — Yeah. So we actually have a lot of customers that are in regulatory environments from school districts, local governments, public safety. So and Genero really does well in those environments. I think it's because of the stability of the code. Right? You have applications that have thirty plus years of business logic and domain logic because they haven't been rewritten. That's the big thing. So you take all that to the those years of domain expertise are baked into the product, right, versus having someone you build it from scratch. You know, you're gonna have to reengineer and refigure out all of those same solutions that you already have into

Mike — the product.

Kiz — Yeah. Okay. Thank you for that, Mike. Thank you for walking through a day in the life for you. What for the folks on the call, what could an action plan look like? What would it look like if they were to engage with your team or with our teams in Latin America or in Europe and APAC?

Mike — Yeah. So, typically, I mean, this is a this is a great slides. So, typically, we would we talk with the customers and do an assessment. Right? What what is the most what are we trying to :31,1 00 --> 00:27:33,440 achieve here? What are your pain points? Identifying those and figuring out what what to fix first, essentially. So after that assessment, right, typically, the next step is then we do a prototype. So the prototype is gonna help both parties. Right? It's gonna help us understand really what the scope of of changing this application is. Because all these applications out there, right, it's code. People have

Mike — their own styles. It's their own they're pitching the solutions in their own way. Their businesses are unique. So doing a prototype that takes a month, two months really helps everyone understand, okay. This is the target that we're shooting at, and this is the amount of effort it's going to take. And then from there, it's really coming up with a plan and giving you options. Right? You can take that prototype and go and and run with it yourself. And you can partner with us, and we can help you along the way. And then building and also building out a plan like we

Mike — talked before about doing it incrementally. What makes sense for your business? How do we wanna do this incrementally? How much risk are you are you willing to incur based on the type of business that you have? So it's really understanding the business and then building a plan that fits with the business.

Kiz — And I know we mentioned this a few times. Just wanna mention it again is just being super focused, hyper focused on what is the solution we're trying to provide? Who are we trying to provide it for? What do they actually need? What does the business actually need? And it sounds like what your team and the other professional services teams can do is provide flexibility for folks thinking about this migration to do what is best for their business.

Mike — Yeah. I mean, that's the key at the end of the day. Right? We're not building applications for fun. We're building them to make the business more efficient to solve business problems.

Kiz — Yeah. Okay. One thing that in addition to the recording that we'll send out at the at the end of this call is an effort impact matrix so that you can decide for those folks on the call, hey. Where am I on this spectrum of manual rewrite, AI modernization, going to Genero? What makes the best sense for me? Right? So we'll send that out. And what you can also do next is after reviewing that is schedule time with us. You'll see the contact information right here. You can talk to us to get a free diagnostic on, hey. Here's my current setup. Here's my current business challenge.

Kiz — Could Genero help me here? And we're happy to just walk through things that you should even consider, Not even related to generic, just other parts of your enterprise architecture. So you can do that diagnostic. As I mentioned, we've got teams in US, Canada, and Latin America, Europe, Asia Pacific. And so language barrier will not be an issue. We'll find somebody that that can talk your language as well. That's what can happen next. You know, let me pause here. I know we're coming close to the conclusion of our time. Please put your questions into the Q&A. Let

Kiz — me just take a look right now, Mike, if there's anything here. We've got a quest ion on database migration. What is you mentioned earlier, Mike, some of the challenges that can occur when doing a application migration. What are some surprises you've seen with the database migration?

Mike — Yeah. So the database migrations are very interesting. So there's a couple things you have to consider. Right? First is that what's the delta or the difference between Informix and then your target database that you're going to? So the Genero has a piece that's built into the software that will translate or we call it the ODI layer that will translate Informix SQL instructions to whatever the target database is. Right? But there are some limitations because sometimes it's not one to one. But when it is one to one, it will do the translation for you. So some of the challenges are really what

Mike — other applications not what non-4GL applications are talking to the database, and then what kind of triggers and store procedures that do you have. Now we've helped customers migrate the triggers and stored procedures. That's not a problem. But if you have, like, if you have Java applications or dot net applications that are that are talking to the 4GL or talking to the Informix database, that's when you get a challenge, and we we need to work together to to develop a plan for that.

Kiz — Okay. Okay. Another question around data. You know, Mike, you think about Meta's Facebook's recent pseudo acquisition of Scale AI and really focus on the data layer. That that's what that is. How do you and your team think about ensuring data integrity during any of these migrations? 73 2

Mike — Hey. Yeah. So we basically do an apples to apples comparison between the data source in Informix and then the database in the target database. So we have a period of time when we're basically running the application against both databases to compare and making sure that each of those transactions that they have the same net effect on the database. That's really that's really the key when you're doing a a database migration because, obviously the data is Fort Knox. Right? That's what you wanna protect.

Kiz — Yeah. Speaking of protection we just got a note from regarding monitoring and observation of Genero absent production. I know that's something that is on the on the roadmap for our R&D team. I don't know if you have anything else to add there, Mike.

Mike — Yeah. There's so the next version of Genero is gonna have integration with Prometheus, which allows you to monitor activity in the application. So it adds observability and monitoring features to allow you to then to hook into a lot of popular third party applications to monitor or to visualize that of the application. Yeah.

Kiz — And I think the question was actually very specific to Prometheus, so thank you for mentioning that. Okay. Let's give it maybe one or two more minutes to see if we get any more questions. I know we've gone through a lot today. Mike, any last points that you really wanna emphasize to the folks on the call who do have Informix 4GL applications?

Mike — Yeah. So I think it's important to note that so from my perspective, right, Genero is really just the next evolution of Informix 4GL. Right? So it's this the base language is the same, but then Genero adds so much more to it. So right now, people are solving a lot of, I'll say, problems. Right? Like, if you need to do XML parsing, if you need to do JSON parsing, if you need to be make HTTP requests, you can't do that in a Informix 4GL. Right? So there's always either have to use a C library or you have to use scripts and do

Mike — something fancy where you pull in the data. Like, that's all built into Genero. That's what you would expect out of a modern framework. Right? So the blend here is that you get that same core functionality that's in Informix 4GL, but then you get so much more because you can do so much more thing. You know, you can you can do all these low level technical things that you would need to do and see, but then you also get all of the user interface capabilities that Genero has as well. So it really is the most effective and streamlined way to modernize your application.

Kiz — And as the final word here, we will share that effort impact matrix so that you can determine what is the best path. And, again, we're happy to talk to you as well to say what we think are the best path maybe based on your current setup. But what Mike just said is Genero gives you the least risky option with the highest reward. And it really allows you to modernize to get to a place where your applications are future proofed. Right? And with that, we will conclude this session. Thank you again, everybody, for joining. We do have another session coming up on UX UI

Kiz — design for mission critical applications. We just concluded one a few weeks ago on enterprise reporting. Let us know if you'd like to talk to us about that or if you'd like to go through the recording for that. We're happy to share it. Look out for an email from the team. And, again, thank you for joining.

Mike — Yep. Thanks.