APEX Instant Tips #200: Live with Mike Hichwa!

Added: 2026-05-23

[00:00:19]Hello and welcome to Apex Instant Tips, episode number 200, brought to you most Fridays at 12:05 Eastern time. We're your hosts. I'm Anton >> Anton Marua.

[00:00:31]>> And many of you may have heard there's been big news in the Apex world uh recently. Um I already alluded to it.

[00:00:38]The big news is of course that this is our 200th episode. Um so um Mike, I think there's some other things going on as well, maybe not quite as important in the Apex space, but that that happened in the last couple weeks. So we invited you to talk about those sort of other other things going on in in the space.

[00:00:55]Um so as our gu as our viewers know both of them uh um Mike Hitchwa is the father of the creator of Apex um from back in like what was it 200 >> it was first started in 99 and and first GA in 2004.

[00:01:15]>> Okay. So all the way back then. Um so um Mike thanks so much for joining us on this episode. Uh you were our one EP our guest on the 100th episode as well. That tip was about how to add an icon to a popup LOV. Um so is this tip as groundbreaking as that tip?

[00:01:37]>> It's a little different. It's going to be less practical. So it's going to be more more it's just it's just not as practical. So I don't know if I'll be able to live up to that that that the vision that I had then. But we'll we'll give it a try.

[00:01:48]>> Okay. Well, um we're going to let you jump right in. Marwa, please jump in with any comments and other folks if you have comments along the way, feel free to ask them to Mike. We'll try to keep an eye on the comments um and and feed them to Mike as uh as we see them. So, it's off to you, Mike.

[00:02:06]>> Hello, Francis. Um uh okay, so today's Apex instant tip. So, it's going to be four words. Um but it's going to take me some time to kind of work up to it and set the stage. And so I'm going to start um uh and so um when I first envisioned Apex, I it was a um in reaction to Oracle Forms, I loved Oracle Forms, but I I I realized that the the world was going to the web browser. So I wanted to have um an entire experience to build an application inside of a web browser. And so uh um the the idea was that you would use a universal client called a web browser versus doing client server and you would be nothing to install. You just come to this web browser and you could design an application, deploy an application and people could use the application and the developers and the end users all they had to have was a web browser. Everything would be great. And of course when I built that application when you build the application it would be easy just like forms was easy. um it would have a stateless architecture as as opposed to stated like Oracle forms would have a a SQL net connection. Um and so the only reason it had it had to have a stateless architecture is I asked Tom Kite um how does the web work and he told me it has a stateless architecture.

[00:03:22]I asked him to describe what stateless meant and that means you got no persistent connection. I was like oo that's a that's an issue and so we had to work around that and build an application server and we had to to do that. So I didn't I did it out of necessity. um which we had to build session state management things like that. So I knew it had to be dataentric.

[00:03:39]I worked for Oracle and and I believed that most people wanted in the enterprise space to have access to their enterprise data as as I told Larry Ellison that if you could um see the data in your database that people would actually find the value of the Oracle database greater if if they see and and can touch and feel and query their database would be greater. um >> which was the original or one of the original names HTML DB was >> yes that was Larry's name uh so so it was it was get you know put your data in HTML so you can see it yes and that was that was that was it was released as an as HTML DB and and um at the time as well like our apps group was into these flows things so I I was chatting with Ron Wall who ran ran our apps division at the time and and uh so he was all about these flows like you know you start here and you go here and so you basically when people would interact with computers in their job, they would want to do something, wanted to accomplish a task and the task was kind of a flow. And so if if you noticed in internally maybe a lot of the APIs, they were all W flow and that's that was sort of the internal name uh there. But the idea there was that that that that since um everything was stateless that that Apex would would remember things. And what's interesting of this whole uh session state management is that it's very similar to what AI is now, right?

[00:04:54]where everyone talks about their conversations with AI and they wish that they would remember or have a memory bank or remember what they're doing. And so it's very very similar. And so what what what we envisioned back then was uh um that that that you know Apex would automatically maintain the session state. So as you go through your application or step to step, it remembers what you typed in the fields and it could be used. And so anyways, we have these validations and processes, flexible security, everything conditional, it would be extensible. So the extensible point was I knew that a platform the idea was to do standard components like charts, forms, reports, but sometimes your application wants to do something special and we didn't want to put you in a in a box. And so I felt as a forms developer, I couldn't do something special. I was kind of limited. I I could do, you know, a data grid and I could do some things, but I really couldn't do some things that I might want to to express my in my application uh creatively. And I I was blocked. And so I wanted to make sure that people were never blocked and they could always uh uh uh um uh express what what they wanted even if they had to do it procedurally and extend it. So it might not be low code or or or no code but it but you would not be prevented from doing it. So th those were the original founding design principles. And Mike, I want to interrupt and say it it it didn't make your list, but one of the things that I've always been so struck by and impressed by is that Apex is built with Apex. All of the visual components are built with the tool itself. And >> and that has to me been a huge selling factor because a I know that the developers are using the tool, right, in order to develop the tool, right? And so things get things that we need as developers are being used by the Apex team all the time. Um uh but but B just that that so much of it became metadata driven because of that I think and that metadata is a huge driving aspect of of Apex. So >> well that's that's that's a great transition into my next slide and and and uh um so so in this slide we talked about like like the designing principles but but not the implementation and so some older people might remember before uh HTML DB and what's now Apex uh there was something called web DB and what web DB was it was it was a generator so it would generate PL/SQL this was an era in Apex in the 19 sorry in Oracle in the 1990s where we had something called designer generator and designer generator would generate forms but it would also generate PL/SQL web applications and so um I I thought that was the way to go and I started doing that but then in 1999 I had this realization that it was absolutely silly to generate every application from scratch and every application would have a reporting application every application would have a charting system. It just made no sense that you should just use a shared uh infrastructure. So, so we came up with an idea to have a model driven execution engine. That's the formal academic name.

[00:07:58]But basically what it says is that that all the as aspects of your application are captured in metadata in an Oracle database in a table in in in rows and columns. you decide that my application is called FU, that my my my I have page one is called home, and that that there's a a region called dashboard, and that there's a a chart called, you know, sales by by by country. So that that the idea is that you just enter in in capture and metadata everything that needs to be um uh uh um itemized by the application and an application engine in real time. When you request it, you're saying, "I want to see application 100 and page one." It would go read the metadata, say, "What is application 100?

[00:08:44]What is page one?" And then it would say, "Oh, it's a dashboard with a chart.

[00:08:48]I've got it." And it would go run the SQL statement from data in an Oracle database in a row and a column. There's a SQL statement. It would execute that SQL statement and then return you the result. And uh um what's what it turns out is that in in in uh uh building applications a lot of the time is spent networking navigating uh the the uh you know making a network call getting a response making another network call and since the apex engine was also is included within the Oracle database there is no network so the network is eliminated. So when you want to ask SQL statements or questions or conditional SQL statements, you can run 20 30 40 SQL statements or more in less than a millisecond. And and that allows you to eliminate uh network traffic to to make a network traffic call. It has a specific overhead. We can make probably a hundred SQL statements in the same time it would call you to make one mid-tier to to database tier. So by by encapsulating the the Apex engine or the application engine, it as metadata and rendering it in real time from the database makes your application just data. You you just add data to the database and it's just data. So um apex.oracle.com has over hundreds of thousands of applications. One database, hundreds of thousands of applications, millions of pages. um uh uh uh and so so the the model scales very well and and everybody is running the same code. You upgrade one thing, you've upgraded everything. So the the founding principle of of of of the Apex engine and of the Apex application server and of our low code framework is this model driven. So it's all about the model that if the model is good, your application's good. If if Oracle does not do a good job with the model, then you could have criticism like maybe our workflow doesn't have everything. We're working very hard to to bridge gaps where we're basically in enhancing our model in different places. I think we have very good models for interactive reports for faceted search and and many other things. Okay. So what does the model look like? So if it's all about the model, the model defines my application.

[00:10:59]What does that model look like? And so here's my page I want to create. And so this is this is the this is this is the end user. this is what the Apex engine renders to the the the to the user. So all good there, but what is actually in the back end or what what what is what is what does the metadata look like? So if I was to export the metadata uh for that page and lines uh 5,857 through 5,861, they are my SQL query and it looks like this. And so we've got large globally unique identifiers, key foreign key references, we've got unintelligible parameter names, a bunch of W flow references which has really no context.

[00:11:48]We have no safe way to edit it. If you try to edit this, uh uh uh we won't support you. Um uh it's not documented and there's limited validation. If you want to corrupt the Apex dictionary, all you have to do is change some of these numbers and it will be all corrupted and you might break stuff. And so it's basically uh a um what we have today and it's very similar to what everyone has. It's just an export import format. You're not supposed to look at it. You're not supposed to know about know about it's not important what it is. It's just a a way to get something out of one system and into another system. It's a interchange format. And so, um, on May 14th, just not a few days ago, we released, uh, Apex 26.1 and, um, we've changed, we've adjusted our export import format, and we made a big change to the export import format. We changed it to not be more readable. Sure, we changed it to actually be an open application specification language. So as opposed to making it a data dump, we decided to export it as a language as with a full grammar and everything about it. And so we decided to go all in on making our export import as elegant as we could. And so here's what it looks like. the same page, the same application and there are no global identifiers just just the hierarchy uh the the the way where it fits. You know, page five on top is is is indented underneath the uh the region, the sales overtime region. And so it doesn't need to have a parent because the indentation within the language uh uh tells it where it is. Um all the parameter names are documented. If you hover over them, you can hear you can read about them.

[00:13:40]They're they're they're easily understood. If if you make an error or you type the wrong thing, it will get it will turn red and and and it will be wrong. Um it it tells you your error. Um it it's supported and safe to edit. If you want to edit this in an editor uh by yourself, you're more than welcome to.

[00:13:57]And if it doesn't work, you can call up Oracle support and we're obligated to fix it. Um, and uh, it's it's fully parsed. And so if you make a change that that you reference something you're not supposed that that is that is doesn't exist or no longer exists, it it will return an error. So it provides that that that uh a part parse base validation. So it's strongly typed, if you will. So in my opinion, it's a it's a uh it's it's a metadata zen garden.

[00:14:26]It's very well organized. It's very clean. It's very tidy. Everything is great. And so if we compare the before and after, we have a data dump which is the raw foreign keys, primary keys, uh d the the execution engines uh uh uh uh column names, just kind of the raw stuff. And then on the right, you have the the the view optimized for the uh for for a developer, for a human. It's it's it's it's in a much more readable format.

[00:15:00]It's the same information. The the information on the right is just it's just they're both imported and they both result in exactly the same data and exactly the same tables. They just do it one elegantly and one less elegantly. So if you compare the two, you have a data dump, that's the the the graphic on the left, and you have a Zen garden, which is the elegant metadata that's in a in a language and and not in in a data dump format. And so it's much more elegant.

[00:15:27]And Mike, you mentioned that um you can uh you can edit this and reimpport it, those kinds of things. Is it is it can you create an application from scratch this way or should you always start it?

[00:15:37]>> I'm getting there. I'm getting there.

[00:15:38]You have to wait for slide 15 and I'll get there, but ask another question about the the >> Okay. So, um so we've got a couple of these that are the similar things. Um that's that's an answer. So, um we got we got questions and answers coming in in the comments.

[00:15:54]>> Okay. Okay.

[00:15:55]>> Good. So, >> so, so, so this is these are they're all kind of setting the stage, right? And so, um, uh, uh, uh, so why do I care what my export import format looks like or whether I can edit it? Like, I don't want to edit it anyway. I that's why I use a low code system. I drag and drop.

[00:16:12]I point and click. I go through wizards.

[00:16:14]Like, I I don't I just don't care. And you're right to care. And no one cared.

[00:16:19]I didn't care. Um Um, until AI changed the world. And now we want to leverage us as developers want to leverage AI agents to help us uh build out our applications. And if if the AI agent is going to specify my application, if my AI agent is going to help me fix my application, if it's going to translate my like my application to French, if it's going to to change uh labels on my on my buttons, whatever it's going to do that, I should be able to work with my favorite AI agent and I should be able to ask it to do something. And so what can I ask my AI agent to do? And if it's Apex, anything because everything you can do in Apex is exported and imported via Apex Lang. So if the AI agent can edit the Apex laying, which it can, uh uh uh then anything Apex can do, you can ask an AI agent to do. Now the AI agent might not have the proper skills or your question or or your request of your AI agent, your command might not be clear enough. So it might not work but it is capable of doing it. And so now our job in the community as well as here here at Oracle is to give the AI agents. So we're working to to to to to train the AI models like chat GPT. We're working to to improve all the skills to make it so you can do everything. You can always do a lot of things today but in the future you'll be able to do much more.

[00:17:50]So the you why do you care about uh the export import format? you care because you want to delegate certain um uh uh application u development responsibilities uh uh uh to your uh agent. So, >> and Mike, as you mentioned to me earlier, this is a different approach than some folks are taking. It's not that you've exposed a bunch of APIs that an agent can invoke an API to do something. You've exposed the entire application as a language of intent. Uh which is like like like we were discussing a different approach to this.

[00:18:24]Um >> that that's right and so we we think it's unique. I haven't seen anything. Um maybe someone else will copy us, copy what we're doing. And and uh so if you look at um uh uh um uh domain specific languages so so so Apex Lang is a domain specific language. That's the academic term. But there's another popular domain specific language called SQL, right? And so the analogy to SQL is remarkably similar. So you say select star from M, right? That's a SQL statement of create table. um those SQL statements actually just are a in in a language format right there's a SQL parser the SQL error messages and things so SQL gets parsed uh create tables let's take you say create table employees it goes and it inserts rows into tables inside the Oracle database they're tab dollar sign and call dollar sign and and they're they're just rows and tables and they got ugly table names and ugly foreign keys just like Apex has like just a table. And so you have a domain specific language called SQL that inserts rows into into tables uh that that uh uh that that constitutes an Oracle database that have been working since uh 1979. So that was the model by which I I modeled everything after as I built Apex at just as Oracle built the database and then the it just took me a long much longer time. Uh I didn't have cod and date write me an application specification language that I could copy. I I had to we had to create it ourselves. Uh a lot of people contributed to making this language what it is today. A lot of really solid work there. Um and and uh so so anyway, that's that's the that's the genesis of the language. But the vision here is that you should be able to as a human go in and look and and edit just the way you do now through an IDE. And you can change labels, you can add pages, you can do everything. And we're going to continue to enhance that.

[00:20:17]We we we like this approach, but we also feel that people are going to want to generate applications or they're going to want to make changes to their applications um conversationally and we want to let you do that in an open way.

[00:20:29]And the most open transparent way we can do it is you is is to allow anyone to do everything and have it all out there in the public. And so by by open application specification language, the open means that that the grammar is open, that we're giving it to everybody.

[00:20:45]We're not hiding it. There's nothing proprietary. There's nothing you need from Oracle to generate Apex Lang, right? When you if you ask a LM to generate a SQL statement, you don't need the Oracle database. You just need to know what an Oracle SQL is. And the LMS know that to generate Apex Lang, you don't need an Oracle database or anything from Oracle. You just need to know the grammar of the language, which of course we publish, >> right? And theoretically, somebody could write a um a converter from Apex Lang to anything else, right?

[00:21:11]>> And it's it's out there like you can convert Oracle Oracle SQL DDL. create tables to Postgress or to anything else and and you know we might even do it too, right? Like so so if if if you want your Apex application to run on your iOS or Android device and and it might not make sense to run an Oracle database. I I've been working with the database team and they're talking about making a smaller tighter version maybe but I don't know like it seems overkill to put an entire Oracle database with the Emon and the PON and the flash cache and everything on your iPhone. it just seems a little big but you know maybe maybe we should you know build something that's a little different with SQL light you know maybe it's the right engine so we we can we can leverage the same thing and so it's just building a an alternative engine but the specification what you know the description of what type of chart you want is common right it's so the specification's there it's just the implementation that that that that is different >> excellent >> okay so now the real question.

[00:22:14]Why should I use Apex when I can generate an app with an AI agent? You could just go generate any application you want willy-nilly and it will work like um you don't you could just say you know build me an application that tracks this and it will generate a react application or TypeScript application whatever application can be done and so yeah you you can do that um and so but a better question is can I get better results generating an Apex app because it can generate an Apex app too than I can generating direct implementation. So I believe that's the real question. Um there's a subtlety to this and that in today's world we're like May 22nd 2026 you will get better results direct generating direct react because all the LMS are trained on it. But once the LMS become trained on Apex Lang then they'll be able to generate seamlessly and and and and with greater proficiency everything that that Apex can do. So if if your solution set is within Apex's sweet spot, it's within its capabilities, then um you you will be able to generate anything Apex can do super fast and super efficiently. But it does take a little bit of bootstrapping as the as the u um AI systems uh train and and they've only had a few days to do that. And so it's going to take a a little bit of time for us to get there, but we're working very hard to get the skills and to work with the LMS to to get them trained up uh on on Apex.

[00:23:47]>> Okay.

[00:23:48]>> Game of leaprog. Yes.

[00:23:50]>> Yep. And so the um uh so if you want to answer the question like is it assuming it's equally efficient to generate an Apex app as it is to do direct generation of a React app or or what.net net or whatever. I don't want to pick on any particular framework. These are all very very good and very robust frameworks with with a tremendous amount of computer science behind them, but what what do we recommend the Oracle community and and what people use? We we think that that that we want to look after our customers best interest and it's rooted in in what do you want to own? Like what do you want to be responsible for? And so um that gets me to today's tip which I said was four words.

[00:24:32]>> You only have five minutes to give the tip. So, let's see if we can do the tip in five minutes.

[00:24:37]>> Tip in five minutes. Okay, four words.

[00:24:40]Um, generate what you own, own what you generate. So, there's only four words.

[00:24:45]>> That's it. We're done. That was that met the five minute mark.

[00:24:50]So, so, so and what what this means I'm going to translate the words uh uh uh is that uh no one if you ask application developers like no one wants to generate an operating system and and and and redo the way they do uh Linux does CPU management or flash cache or network access or memory management inside the the operating system. Most people don't want to generate databases. they don't want to redo the the 40 some odd years of concurrency control and atomic transactions and data integrity. So there's a lot of value in the platform.

[00:25:23]So all all these platforms provide tremendous value. So what what company the the what the world's best companies use virtually every single I I want to hold every single bank uh uses an operating system uh uh they use a database and then some applications are built directly directly in the core functionality and some leverage an abstraction tier. So an operating system is abstraction on hardware. A database is a an abstraction on top of data and files. An application server uh an application platform is an abstraction on application capabilities. And so by using an application platform like Apex, you are abstracting the need. So as opposed to to generating and asking your LM to generate every detail of your forms, all your session state management, all your security, all your your um map and spatial capabilities, authentication, authorization, and do the entire stack, your the idea, the the tip is to say just generate what your company needs. Generate the unique attribute that that's unique to your organization. don't you know take the risk of generating a bunch of stuff that's AI slop if you generate less if you can generate the minimum you you can then uh understand it better you can better better maintain it and you can you can if something goes wrong you can have a better chance of fixing it and so uh uh uh uh trusted platforms provide this governance and so um u I think I feel proud that Apex is a trusted application platform it's used by intelligence communities worldwide military is worldwide. Um, it's used in governments. It's used in banks. Banking is our top community. And so what what my tip is is that in the era of AI, embrace it. Don't embrace it blindly, but but understand clearly what you want to own. Do you want to own the implementation of session state for a low code applica, sorry, for for an application platform, or do you want to uh to own your business logic? I think you want to own your business logic. You don't want to understand how do you draw slices of pie charts. And so, do you want to own the slice drawing or do you not want to own it? And so, that's my tip of the day. Generate what you own and own what you generate.

[00:27:45]Excellent. Um, so, uh, Mike, we do have a couple of questions, uh, that came in along the way. Um, and, uh, one of them, uh, is, are there plans to improve the design, uh, design time environment? So outside of Apex Lang in the builder itself um >> yes we want we want to make everything easy and I will um tip my hand of my vision here and so >> I don't know how many people on this call go to the Cheesecake Factory um it's a it's a restaurant >> but they give you this visual menu and and and >> it's got 400 items on it >> and but but you don't have to read anything. There's nothing needs required translation. You look at it and and and there it is. And and let's say you like hamburgers. You can have your with cheese or without cheese. With with Swiss cheese or with proolone, you could have it with pickles or without pickles.

[00:28:42]Okay. So imagine that in the world of of of web applications that there were design patterns, common best use patterns, but today's patterns are today's patterns. Tomorrow's patterns evolve. And so my vision is that that when you build applications as opposed to specifying the details, you know, even though it's abstracted, you know, which slots going to be in, how many spans do I have, which template would you like to use? You know, you don't you don't really want to do that. What you want to say is I want a faceted search and and I but there's different types of faceted search. You know, sometimes on the facets, they're like on the top, then you click on the thing called filters and a filter thing pops up and you do your facets. Sometimes like Amazon style, the facets are on the left and some people like their facets on the right. So in your uh cheesecake factory menu of everything possible that you might want to to eat, you you just say you say, "Oh, I'd like to have a search." Oh, what type of search? Oh, I'll take a fasted search. Okay, great.

[00:29:43]What which which pattern of fasted search would you like? And so what we want to do is to innovate in a way that's that makes it easier for the developer and easier for AI to generate something because what you're you're you're you're capturing because Apex exists to put UI on top of data. Um and and and uh uh but what does the UI look like? Well, it falls into let's say 400 primary patterns. So we want to document those patterns and then when we implement that pattern maybe there's some JavaScript involved maybe the pattern is like a conbon board and so um as a pattern evolves the the implementation detail could change and so the idea is that that they'd be versioned and if you want to update to the next version of the pattern you wouldn't have to do anything you would just upgrade the to the pattern and so the idea is to is to is to make application easier by by making UI pattern patterns based. And I think that anyone that's used the internet for for anything real like banking or shopping or um uh uh analysis or whatever, they see these patterns. Of course, a modern pattern is a a um chat assistant where you you ask questions and natural language comes on the other side. And so there's probably five of those patterns and we want to provide all of them out of the box. And so that's how I see um the the the evolution of of user interface that is uh pro-human and and that humans can select the patterns and also pro AI and that AI doesn't have to have the burden of getting all the detail right and they can just take the pattern and then when someone improves that pattern that everyone's uplifted and everyone's running the same code >> and I think this uh this really falls into what you were saying you know when you say generate what you own what you generate You don't necessarily want to own those patterns. Those patterns are they come and go throughout the, you know, >> you might want to own two of the 400 or three because you have a vision that no one else has or you have a uniqueness, but I don't like do you really want to own the like, you know, faceted search, you know, like we've got eight faceted search patterns. Do you have a ninth that is absolutely critical? You might, but chances are you don't, right? Like, >> right. And and to your point, what what you as a business uh what makes you unique is what what makes you successful. And so you need to concentrate on those things. And and this gives you the the way to concentrate on those things that are unique to your business. And if there's nothing unique to your business, well, um maybe you have nothing.

[00:32:14]>> You need to make your business run better, right? like like if if you can do better, you know, better faceted searches or or better uh uh chat dialogues or better forms or better select lists, well, why don't you go in the software business and open a software company, right? But if if if you're better at actually if you're a bank at making loans that that that that don't fail or I if you're a government at providing services to the community, if you're a hospital at making sure that your patients have the best care, that's what your thing is. not necessarily building the select list or the or deciding you know which slot that particular uh region needs to needs to to go into. And so that's the the idea is that that that that that um yeah it's to it's to leverage the the the world's best practice the world together you know all these innovative companies Apple and Google and and Microsoft many others they they all kind of synergistically iterating the patterns of of human computer interactions and the best of the best we want to surface as patterns I in in Oracle and allow our customers to build their own if if somehow we miss something.

[00:33:26]>> Excellent. Well, um we have one other question that I see that um I see a couple that have been answered and a couple that haven't, but there's one about um where can developers find the Apex Lang specification um itself.

[00:33:39]>> Um so, so uh uh we we'll we'll send a followup, but there are skills that define Apex Lang.

[00:33:48]So there's github.com/oracleskills.

[00:33:53]Underneath there's a some Apex skills.

[00:33:55]Um uh uh and then there are other places where we we document other aspects of the language. Not everything has been released. We're going to be releasing it little peace meal. We felt like we didn't want to wait until we had uh every every last piece of documentation done. We have a lot of documentation, but there's a lot more coming. And so um but right now we do have skills and we expect um our skills to be updated on an almost weekly cadence to to to get better and better and better to to help us and uh on this. And so if you're doing generative development with Apex today, it's the worst day ever. It will only get better. It's going to get better every single day as the LMS get better, as our skills get better. But what's important is that we have the right architecture, the right abstraction, and the right the the language-based ability to to to to do it. But it's spectacular today, but but but the but the future is is is even brighter.

[00:34:50]>> Well, Mike, I can't I can't let you go without asking this essentially the same question that you've heard for the last 18 months, and that is when is Apex 26.2 coming out?

[00:35:02]uh it much much le less time frame than than uh um the uh uh uh than the last release took and and and the last release took a long time because the industry is going through its largest upheaval ever in my opinion and and I felt that Apex had to be positioned for this otherwise it would be dead and and and that Apex Lang has done that and so now we're back into iterating back onto regular cadence and work less some conference in some desert someplace out west and and and I'm hoping that it will definitely be done before that conference.

[00:35:39]>> Excellent. And any um of course there's all the safe harbor you can whatever you want but any thoughts on what what that might include the big things that that are going to be >> the big things are are are um improving and hardening the ability to perform both generative development as well as it'll be a hardening uh uh uh no no major earthshattering releases. is the earth shattering ability is to be able to to specode vibe code conversationally develop um in addition to the traditional IDE based development and be able to use those best practices uh source code control diff merge work in professional environments um scanning tools so so it'll be a boatload of features be a long list of features but they'll be kind of supporting they won't be you know a gamecher type things, things that you'd want to to run a a security scan against your Apex lang files to to to or you know, all all those types of things. So, a lot of iterations on those areas.

[00:36:43]>> Excellent. Um, well, I think that covers it with this is perhaps our longest Apex instant tips episode ever. Um, >> it's a good tip though. I like even even if you don't use Apex, it's a good tip, right? If you're just generating anything, you should you should generate what you own and you should own what you generate. Like you should, you know, >> absolutely solid tip and the tip is is was five minutes. I it was probably 15 seconds. So, um so I think we we we met the goal of the tip happening in in that time period. Um so, uh Mike, thanks so much. Congratulations to your son on his graduation. And uh thank you.

[00:37:22]>> And uh if anybody else has additional questions, put them in the comments later in the show and uh we've got a direct line to Mike. We'll uh we'll make sure we get them them answered. Um Mike, thanks so much again. And um for our vast audience, uh thank you for joining today and do all the things that you're supposed to do. Subscribe, like, uh send your mom a letter about the show. She'll love it. So, all right. Appreciate it.

[00:37:49]Thanks everyone.