Where is Web 2.0 headed in the next several years, and what does it mean to the IBM system manager?

After interviewing a number of experts in the area of Web technologies and trying to get an idea of the direction these technologies might be heading for the purpose of helping the IBM i system manager who wants to modernize, I concluded that the Web is in a state of turmoil. It's Libya in the sky. The concept of Web 2.0 itself is indefinable, so it's hard to tell where this amorphous collection of technologies is headed. If anything emerged from my inquiries, it is 1) cloud computing is here to stay, and 2) Web-enabled mobile devices will be more ubiquitous than PCs.  Other than that, it's Adobe, Microsoft, Apple, and Google wrestling around like a litter of six-week old puppies. Hurry up HTML5, pleeze! 

Below are interviews with three extremely knowledgeable people who kindly shared their views on the emergence of these truly amazing technologies with MC Press Online readers. These are not the only people with whom we spoke, and you will hear from others later. There is a lot of material here, but it's all good. Note that these interviews are heavily edited to improve their clarity. So grab a coffee and settle in for a look at our forecast of where Web technologies are and how they might affect you in the future.  

Tim Rowe

Business Architect Application Development for IBM i

Web Integration Team Lead, IBM i5 Development Lab

Rowe: There are many different offerings and options that you can use. There is no single answer when it comes to Web 2.0. There are a number of different tools that are available, depending if you want to go with pure Java. But there also are many different things that you can use that are readily available: the JavaScript Dojo Toolkit, for example. Rational has some Web 2.0 type tools you can take advantage of. If you want to stay with pure RPG, obviously it doesn't have a Web 2.0 interface, but if you want to make use of some of the Open Access interface solutions and hook into one of the handlers, for example, you can start taking advantage of some of the Web 2.0 technology that provides the look and feel and some of the Web 2.0 type attributes. Certainly, Rational has several products that exist today that can help you start down that path, whether through EGL or with HATS.

MC Press: Before we get into the how to do it, let's ask ourselves where are we going. Say I've got a 5250 application and I've got users who are suddenly buying iPads. What do I need to be doing today to make my company as competitive as possible five years from now?

Rowe: Part of it is determining where you're at today and how you want to get to a different place in the future. There are a number of different aspects to Web 2.0 that you can bring into play, whether it is some of the look and feel issues of the user interface, or whether you are making use of some of the nice AJAX support within a Web-based JavaScript type interface, or perhaps wanting to make use of some of the other Web 2.0 type widgets, whether wiki, forum, or blog type features, and how do you want to go about incorporating these into your business? For instance, if you are in a commerce type business, there are things already available through say, Websphere Commerce, to help make use of those blogs and wiki type implementation features. But if you're running a pure RPG application, then you need to start making decisions about the best direction for you to go in modernizing so that you can take advantage of what you already have.

MC Press: So it kind of depends on what business you're in as to how you might deploy a modernization solution involving Web 2.0.

Rowe: Your business, yes, but also your skill set. Let's face it; if you've got a pure RPG shop, rewriting everything into a pure Java implementation might be out of reach. But if you're looking to get there immediately, and you want to do something quickly and take advantage of some of these look-and-feel issues, then a product like HATS can get you there very quickly. It's got its limitations, but it can start you down the path. Taking advantage of some of the Open Access support that is out there today—through some of the ISVs that are providing handlers—can also get you there while continuing to maintain your core RPG business applications. You could also do it with PHP. And depending on where you're heading, if you wanted to make use of a blog or wiki, for example, there are open-source blog/wiki utilities that you can drop onto an IBM i box today using PASE and run and drop those into your Web UI [PASE for i is an integrated runtime for porting selected applications from other operating systems].

MC Press: Does your team have a vision of what the landscape is going to look like in two to three years in terms of the relationship between today's RPG applications and tomorrow's World Wide Web?

Rowe: I have to start with RPG since that is the foundational language of IBM i. There has been a lot of work over the last several years to really try to make RPG into a modern programming language, and I think we've made great strides in that area. Even in the next two or three years, I don't see our IBM i community being able to completely divest themselves of that, so we've tried to come up with an infrastructure to allow our RPG community to take advantage of these more modern Web UI types of things yet still maintain that investment in RPG. And we're going to continue to try to maintain that support for the foreseeable future. At the same time, there is additional tooling through the Rational folks that you can take advantage of and tools available through some of the ISVs. Open Access is appealing since it allows the ISVs to really focus on what we're talking about, and that's creating an integrated Web 2.0 experience yet maintaining that native RPG.

MC Press: You've mentioned the Rational tools a couple times and Open Access for RPG. Is there an aspect of this that you feel the IBM i community may not be fully aware of yet? I know some of the ISVs have come up with handlers that are working pretty well, and they're just getting some of the related tooling out the door. Is there something that you would like to emphasize that you think people should take a look at?

Rowe: We help ensure that the operating system has the support required to help our ISVs and other IBM users produce a modern UI. We do help out the Rational team, and certainly HATS is a great opportunity that people may not yet have explored. That's still an option for many of our customers. From the view of the operating system, one of the things we've tried to provide for our RPG customers is Integrated Web Services support, which allows our RPG family to start the process slowly, where they can take advantage of calling a Web service, or even creating their own Web services, to be externalized for others. That support already is in the OS, and it has been there since V5R4! So that is another area where customers can start this process without having to jump in hook, line, and sinker. They can start at a more gradual rate until they become familiar with and comfortable in these technologies.

MC Press: What do Integrated Web Services do for someone? What sorts of features could someone expect to find if they explored this feature set?

Rowe: There are two aspects to Integrated Web Services support. There is the ability to take an RPG program or service program and very quickly and easily externalize that as a Web service. Basically, you can call from the Web an RPG program to perform a function for you. It also applies to COBOL as well as RPG. The other side of the Web services coin is, say I have an RPG or COBOL program, and I need to access information or I need to submit information to some function out in the Web that has been externalized as a Web service. How do I go about doing that? The Integrated Web Services support provides the tooling that will create an RPG or C subprogram that you call if you're doing an ILE C or an ILE RPG program call from your native RPG, which magically takes care of calling that Web service and doing that whole interaction or transaction. You don't have to deal with SOAP and XML and the HTTP requests.

MC Press: Do you think these features are being overlooked by some people who may launch off into some other technology they believe is attractive?

Rowe: Yes, I'm sure they are. I just had a call this morning from a customer working on moving their UI to a Microsoft solution, but they want to connect into the IBM i computer; they want to keep some of their data on the i and need to be able to talk back and forth. They heard about Web services and asked how they could do that. I was able to help explain that to them, and they said that it sounded like it could be a good solution for them. That's one way of doing Web services; you can also do Web services through Rational tooling and run that on a WebSphere service.

MC Press: You mentioned HATS too, and some people think it's been around for quite awhile now, but are there some updates and improvements to HATS that should cause people to perhaps take another look at it?

Rowe: For one thing, HATS supports mobile devices. They've put updates in it to support the iPhone and iPod, and they're updating it to be able to communicate with other mobile devices. Again, HATS does fit a very real need; it allows you to get there very quickly. Sure, it has its limitations, but what solution doesn't?

MC Press: What sorts of limitations are there that people should be aware of?

Rowe: When you're using HATS, you're still running through the 5250 data stream so there could be some performance, or other data transfer, issues because of that particular approach.

MC Press: As far as support for mobile devices, the usual suspects are iPhone, Android phone, and now the iPad and the Android tablets. Where do we think HATS is?

Rowe: At the moment, I believe they support the iPhone and iPod Touch [and several PDAs].

MC Press: What sort of things do you see emerging from the IBM i to further the integration with the Web that perhaps we haven't heard of yet?

Rowe: Certainly from an education perspective, we want to continue to spread the word about support for Integrated Web Services that already is in the OS. We also have the container already in IBM i that allows for many—whether it's PHP- or AIX-based—applications that can pretty much pop over and run without modification on IBM i. That's something that we need to get the word out about and provide information to a lot of people that they can very quickly adapt some of these technologies on IBM i. Certainly Open Access for RPG is another area we're going to continue to focus on and ensure that the necessary infrastructure is available, both for our ISVs and customers, so they can continue to modernize their existing RPG applications.

MC Press: Are you working on a handler yourself?

Rowe: From an OS perspective? No we're not working on a handler. We're working with the ISVs. The ISVs have handlers and so does IBM Lab Services, and we have worked with them to help where we need to.

MC Press: You say Lab Services has a handler?

Rowe: Yes, Lab Services also has some handlers that allow you to do a variety of things, and they certainly can support you with a handler and can also help you learn how to use it and make it work for you.

MC Press: Where does the migration of RPG Open Access go from here? Once it's been released and announced, are there further enhancements that can be made?

Rowe: There have been some discussions between some of the ISVs about the possibility of creating some kind of a standard to pass metadata back and forth. And that is something we're looking at. At this point, however, we don't have anything specific; it's merely an option.

MC Press: When you say metadata back and forth—metadata for what?

Rowe: So that from your RPG program, you can control the look and feel of the screen.

MC Press: Sort of an adjustment dial?

Rowe: When you look at HTML, you have your HTML basic anchor tag, or your text tag, and then there are all of the different things that go with that, whether it's background color, font and size, location, justification—all of that.

MC Press: So would you say there are a set of questions that an IBM i system administrator should ask himself or herself before proceeding forward with any kind of modernization strategy? Are there certain steps or key decision points that you have to negotiate before you can move forward?

Rowe: Well, certainly they need to assess where they are at and where they really need to end up. Those are two really important things to get clear. The other thing that you should factor in is your skill set. What are the skills that you have in place, and what are the skills you want to use in order to get where you want to end up? There are many different ways to go about this whole process, whether it's a complete rewrite or a particular road you want to head down. That's the biggest thing to understand.

MC Press: Are there pitfalls IT managers and developers should, or can, avoid? What is the worst that someone could do to botch things up?

Rowe: Adhering to the view that you want to be at point X, but you're only willing to spend enough to get to point A. Needless to say, points A and X are a long way apart. Then you get to point A and you realize that you're far from point X. However, you did get what you paid for. Some people have certainly done that. I'll pull out the HATS example again. It's fabulous tool for getting up quickly. You can get a Web UI up and running in a couple of days in simplified instances. But if you try to use that to run a couple million transactions a day, well, you might have an issue. You got what you paid for in that case. And sometimes that's one of the pitfalls: you fail to do the proper planning, and you don't properly understand where you want to go, and thus you don't put in the proper investment.

MC Press: In other words, don't start with the cost first. Start with the need first.

Rowe: Obviously, the cost has to factor in, but then you have to adjust your need sometimes to fit the cost.

MC Press: Where do you see the industry in five years?

Rowe: That is a crystal ball question for sure.

MC Press: We've started certain initiatives today—six months ago—and there are little seeds being planted today.…

Rowe: The underlying theme of what's going to be happening in five years is—the one thing that will not have changed is that—in five years, there will still be businesses. And those businesses still will need to have platforms on which to run their businesses. That's what IBM i does. And we continue to provide the infrastructure that businesses need to run their business. It will continue to be a very agile, mobile platform.

MC Press: Do you think the i platform is going to be put out of business by the cloud, or is it going to adapt to be the repository for the cloud?

Rowe: Absolutely a repository. We have people who have been doing that for the last 10 years. So I don't see why it won't be able to continue in that role.

MC Press: Where does security play in that?

Rowe: I was just going to say the same thing. When you think about security and compliance and government regulations and those areas, the i infrastructure certainly provides a very strong platform to insure those security and compliance features.

MC Press: There is a big movement toward the cloud nevertheless, right?

Rowe: Well sure, absolutely. There have been people working in a cloud-like environment for a long time, however, whether hosting or providing other services.

MC Press: So do you see the two coexisting, or what relationship do you envision in a few years?

Rowe: There are certain companies that will never get to the cloud for whatever reason. And so you are always going to have a select set of customers who need to have their hardware on site, and they will need to be able to go and touch it—for whatever the reason may be. We can continue to fill that void very well. Why are there customers who really don't care if they never touch their hardware again? For them, the i platform can go a long way to ensure that their data is maintained in a very secure manner. It provides a good platform to run their applications.

MC Press: Do you see the cloud as shaping the evolution of the i platform in any way? I mean it already has to some extent, right?

Rowe: Yes, it already has and certainly some of the principles guiding IBM i for years are being pushed into the cloud. There also are some things that the cloud is trying to do that we're trying to react to in order to ensure that we can provide that necessary infrastructure.

***

Eamon Musallam

Product Marketing Manager

looksoftware

MC Press: Eamon, what does Web 2.0 mean to you? I take it to mean more of an interactive user experience, like social networking.

Musallam: Web 2.0 to me would be not only the ability to interact with Web applications but it also means a richer interface. What I mean by that is that if you look at Web sites from 15 years ago, built with the earlier versions of HTML and JavaScript, they were pretty basic. Web 2.0 is more than just sharing and interoperability, it also means a richer experience.

MC Press: And what do we mean by richer?

Musallam: There's something called RIA, or Rich Internet Application. There is a bit of turmoil today because by far the most successful RIA technology has been Adobe Flash. Flash in the basic sense is much more DDO and multimedia, but in its broader sense it allows you to have a more graphical interface and also allows more interactivity. So it's almost a gateway to running the application.

Think of a static Web page such as Wikipedia… whereas if you think about applications such as salesforce.com it's more like Outlook running on the Web and a much richer experience set. And then you get into things like Flash. The problem for the past couple of years is that the standards are in a state of flux because the Web is more about interoperability—running on any platform anywhere, running on any device. Apple has really done a thing on Flash because they said they were not supporting it on any of their devices.

Apple is the second largest company in the world now in terms of market capitalization. They've blown Microsoft and now Google away in terms of market capitalization and do have a lot of influence. Just having Apple say it's not going to support Flash on the iPhone and iPads has resulted in what I'm seeing as a big shift. I'm now seeing less and less Flash on websites. What Apple has said they would prefer to see grow is something called HTML5. And it seems that the other vendors are following. Everyone appears to be shifting toward HTML5, including Microsoft, though Microsoft has something called Silverlight. So those are some of the challenges. Standards still have to emerge and be embraced. The thing about Web 1.0 is that it really was universally embraced; everyone runs on it. The challenge for Web 2.0 is to get everyone playing nicely.

Web 2.0 technologies would also include the massive surge of mobile devices that require support as well. Web 2.0 is quite a nebulous term. I gave a presentation at COMMON, and I boiled it down to two main things that are practical for the IBM i community which are very much a part of Web 2.0 as far as I'm concerned. These are cloud computing, or running applications on the Internet, and the other thing is mobile devices: the iPads and iPhones and the Androids and BlackBerries. There is some genuine growth there and some very good return on investment.

MC Press: So you think that the defining elements of Web 2.0 are cloud computing and mobile devices?

Musallam: I would say that in terms of practical examples of Web 2.0 in relation to the IBM i customers with whom we consult—some practical examples of some of the things that genuinely generate interest and real value—are cloud computing and mobile devices. The benefits of cloud computing are that it can drastically reduce your costs because you don't have to buy extra hardware, and software is typically less expensive. Cloud-based applications typically charge a monthly user fee. So in the IBM i community, where they are used to spending tens or hundreds of thousands of dollars for an ERP system, now you are finding applications being sold for $30 per month per user. We have partners that are offering software as a service, a component of cloud computing.

The other thing is what you were talking about in terms of collaboration. If you think about a 5250 or RPG application, it typically only serves customers inside the four walls of a building. But the nice thing about the cloud is that as you start to expose your application to the cloud, you can now share documents, you can work on applications and access data from anywhere in the world. Those are some of the really exciting things about Web 2.0, and I'm focusing in particular on cloud computing, but I would say those are things that are very appealing and should be adopted by the i community.

The problem, of course, is that there still are a lot of concerns regarding security and trust and also the implied dependence on the Internet. My Internet was down all day recently after a big storm the day before, and it really crippled me. It crippled my ability to get anything done. So that is a problem with the cloud—your dependence on bandwidth and your dependence on the Internet. It isn't always reliable, and it's an unknown. So there are tradeoffs, and that's why I think it's going to take awhile for the cloud to get adopted. As far as security and trust, when you think of most IBM i people, their data is so valuable that the idea of sharing it on the Internet and giving people access to it is incredibly scary, and rightly so.

Cloud computing can get people in at a far lower cost and really take advantage of new technologies. But it's also extremely scary, and that's why I say I don't expect it to be adopted quickly by the i community. I think it will be a gradual migration over the next few years.

The other thing regarding Web 2.0 is the idea that the browser on a PC is just one of many ways to access the Web, and if you look at the statistics of how much Web traffic is being funneled to the iPads, iPhones, and Androids, it's still a small percentage, but the adoption rate is so compelling that I've heard people say that by 2014, more people will be accessing the Web from handheld devices than from traditional PCs.

MC Press: What about the bandwidth? It seemed the recent AT&T offer to buy T-Mobile had some bandwidth issues connected with it.

Musallam: Rightly so. At the moment, the devices are really only practical for light use. If you're going to use anything intensive, like a Web 2.0 application, you really need to be on Wi-Fi. All these devices support Wi-Fi, and under Wi-Fi, as long as you have a reasonable Internet connection, they are very usable. Once you step outside onto the cellular network, the majority of devices are using what is known as 3G, at least in the big cities. That isn't bad, but it's certainly not good enough for a fairly intensive session.

Something coming along now, however, is 4G. I was in a Verizon store this weekend, and they just launched a new phone called the Thunderbolt. It's a device running Android, and it was unbelievable how quickly the Web pages were being served up, including videos. And it was running on 4G; there was no Wi-Fi. But that's going to take two or three years before it gets really widespread adoption.

MC Press: Did you buy it?

Musallam: No, I'm an Apple fan! You know the iPhone 4 is now running on the Verizon network, and everyone is hoping that the iPhone 5 will have support for this 4G network. I can see why, now that I've seen 4G in action and how fast it is; it's really impressive. If you start having speeds like that, you could be out in a coffee shop somewhere and have your phone on, then connect your laptop and your iPad to it, and still have superfast speeds.

MC Press: And use it as sort of a router or something?

Musallam: Yes, it's called tethering or MyFi. Now tethering is a practical thing if you have these superfast speeds. So it's just a matter of time, and it's so low-cost.

MC Press: Given the landscape that you're painting, what are the implications for the System i manager?

Musallam: The implications are that the System i manager should not just be thinking that he or she just needs to graphically enable the green screen. He really should be thinking of a multi-channel world that is based on different types of users and how they are accessing the server. They should be asking how can I deliver the application to different users.

If you're an executive who spends most of your time in the office, a dashboard is the most optimal view. The question is, what is the most optimal way to access the application? If you walked into a Dell shop or Apple store, you will see larger PC devices, you'll see laptops, iPads, and you'll see handheld mobile phones. In an Apple store, you'll see big iMacs, Apple notebooks, iPads, and iPhones. If you ask, "Which one should I buy?," the answer is that, obviously, every device has its strengths and weaknesses. But each is optimal for certain activities. What it means for the i community is that you've got to think in terms of the way it's being used.

Let's say you have JD Edwards or some type of ERP application, and you're a salesperson, and you're driving around and visiting customers and taking orders. Wouldn't it be nice if you could just whip out your iPhone and ask for the address for the customer? When you get there, you pull it out again and say, "This is what you last ordered," and have direct access to the information sitting on the IBM i? So to a traveling salesperson out in the field, a handheld device like an iPhone or an iPad is very practical and what I would call "optimal." Someone like an executive or a manager might want a dashboard where they see company performance, what's selling, in a larger screen that is more interactive. So the implication for an IBM i manager is really to think of multi-channel—think of the type of user and how they're accessing the application. That will determine how you deliver the application; that should be the main thing.

 ***

Dr. William Hansen

Owner

Manta Technologies

MC Press: What we were wondering is what might be coming with Web 2.0 and how it will affect the IBM i professional.

Hansen: An article in Wired magazine last year suggested that people are getting away from the browser. They're still using the Internet to communicate but are concentrating on a handful of specific Web sites. People are going to Facebook through an app on their phone and not really caring about the rest of the Web.

I had already looked at this area from the viewpoint of technology, so I was concerned about AJAX and some of the technologies, but Tim Rowe's presentation at COMMON was talking about Facebook and Wikipedia and social networking. But his whole point of view was more about social change—that all the technologies are creating a movement toward social networking and crowd sourcing. The things that we use for reference materials have changed quite a bit; we're relying on the crowd whereas we used to rely on experts.

MC Press: The crowd-sourcing phenomenon is certainly growing. As far as what you see IBM i professionals asking for as far as Web technologies, what are you seeing now, and where do you expect that to go in six months or a year?

Hansen: I'm seeing more and more shops that are finally getting it, topics that we were writing about 10 years ago—that you need to bring your company to the Web. Many companies that went and found a consulting company and just slapped up a website for information only are now finally getting the idea that this is a business tool in a Web setting and that they need to integrate all aspects of their company.

It's not just one-way information; you want customers to be able to place orders and you want your salespeople out in the field to be able to look up customer information and integrate it with the internal database. Imagine business partners logging in, checking orders, and paying invoices. Increasingly, more and more is happening online. Fortune 500 companies have been doing this for 15 years, but now it's becoming common for small and medium businesses to do it as well. Where the i comes in is that now the tools are mature enough to let them do that. Or rather, perhaps the tools have been there all along. One of the courses we teach is how to connect your RPG applications to the Web. That technology really hasn't changed that much in 10 years, but people are finally realizing that it's there and that other companies are using it to do something.

MC Press: They're being gently pushed into it because other companies are doing it, but do they really think that it's a safe thing to do? Are they comfortable with the move, or are they just being forced into it?

Hansen: The change may be that there is a level of comfort knowing that other people are doing it so it can't be all bad. They may not have been comfortable on the leading edge.

MC Press: But at the same time, we're hearing stories about systems being hacked and data being stolen from some fairly major organizations out there. So are they just throwing up their hands and saying, que sera sera?

Hansen: At Manta we used to process credit cards directly, but with PCI standards, the costs got quite high. So now, we contract with Google, and they take care of that. Perhaps that's what's happening with the whole movement to the cloud. There are levels of detail that people don't want to get into. They can have a UI from a larger company to handle those things. Going to the cloud and relying on somebody else's expertise is something that I can definitely see accelerating.

MC Press: Back to the demands for training in this area, what specifically are people asking for most now?

Hansen: I'm seeing a bump in HTML and I have been teaching classes at COMMON in HTML. The people that I'm seeing in those seminars are people who have been RPG programmers. Their companies used to farm out Web work; now they're saying, "It's your job," so they're bringing it in-house and telling the programmer that they're going to start playing with applications that access the database. In some cases, the students say they have been trying to get their bosses to look at the opportunity for years. In other cases, they've been dragged in after being quite happy doing RPG programming, but management decided the company needs to do this, and if it's not possible, then they are just going to get rid of the 400.

It's not exactly clear what the driver is—whether it's basic competition that their competitors are starting to allow their customers to order things online or if there are new MBAs now in supervisory positions who say it isn't as scary as you think it is. In any case, they are approaching it as brand new, even though it's the same thing that we've been teaching for years. I've updated my materials, but they don't see that; they're coming into it fresh.

MC Press: What do these updates consist of? What you were teaching 10 years ago compared to what you are teaching today, even though the request is the same, what they're being served up is a little bit different?

Hansen: In most courses, it's basic HTML and JavaScript and basic Web development. When I first wrote the course, I had to cover Internet Explorer that had 95 percent of the market, while Netscape had 5 percent. That is a major change, since now we have to deal with Safari and Chrome and Firefox and other things. Internet Explorer had been the driver because there were half a dozen technologies that were only in Internet Explorer. Now at least 90 percent of those technologies are provided by everybody. Nowadays, it's more likely that Safari, Chrome, and Firefox support a feature, but Internet Explorer does not; it's lagging. Now, you approach it by getting a Web page or function working in the other browsers and you may have to do something different for Internet Explorer because IE is handling it differently from everybody else. It's backwards compared to what we had been used to 10 years ago. One of the things about HTML5 is that it determines what device or node is running. It was driven by the iPhone and larger devices.

MC Press: Do you see the mobile devices as driving a certain amount of functionality now of Web applications?

Hansen: Definitely. I definitely see more and more business applications going that way. There are some things going on, and one you brought up earlier--security. How many of these security breaches were due to someone leaving a laptop in an airport and it had the entire western region on the hard drive, or whatever? Companies are realizing that it's hard to protect themselves from the employees, so what they're doing is rewriting the application to run within a browser.

MC Press: So if the data doesn't reside on the device, should the device get lost or misused, it's OK?

Hansen: Right. A browser with encrypted communications. People like tablets because they're just that much smaller and they don't carry all the weight. Being able to carry a tablet certainly is a better way to handle that. More and more applications are going that way, where the back-room processing still stays the same, it still runs on the i, and it doesn't matter where the i is anymore, a partition that is running in the cloud somewhere, and the end user has a tablet and finds it very easy to get and update information.

MC Press: But with the app downloads, it seems that having some kind of software on these client devices is still quite popular. Is that a substitute for a browser-based way to access the application, or does it just make it better?

Hansen: It could make it better. What I'm reminded of is that there was a question: Is it better to write an application for a specific device? But then, you have to modify it for each depending on what device you have. Either you buy 10,000 iPads for your company or you're coming up with five different versions of the device software. Do you make it generic so that all you really need is a browser? There are pros and cons to each, and there will always be customers for whom you can't control the hardware, so a generic browser is perhaps better for them.

What people are really using more now is AJAX. Google is really the first big user, but the techniques are not that difficult. All the pieces really pre-date the guy who first called it AJAX. If you do this, this, and this, you get something really cool—talking to the server under the covers with only a part of the page changing dynamically.

MC Press: Would you call AJAX a technology or a collection of techniques?

Hansen: I call it a collection of techniques in my course on how to do it. What is interesting for me is we were doing it in our course software a good five years before the paper came out telling everybody it was AJAX. But it was more difficult to do it at the time. You had to have Internet Explorer, which was the only browser that dynamically changed the screen, and you couldn't talk to the server directly; you had to have a Java applet to have a conversation under the covers with the browser.  We did a lot of clever stuff, for instance, when someone opened the course, while they were looking at the first screen, it was pulling down pictures and text in the background. Our motivation to do it was because our customers had dial-up connections.

One of things about HTML5 that is coming is that you should be able to use a browser and not need a plug-in such as an audio player or a video player. Each browser will have that built in. Currently, it's a fight between Apple and Adobe. What you do in Flash, you should be able to do in the browser because the plug-in can create security issues.

MC Press: If it's in the browser, then it becomes standardized and there is no option for doing your own thing?

Hansen: There are some places where you want to do your own thing and others where there is no reason for everybody to do the same thing. We all don't need our separate way to play audio files. Now, I don't think there is a single audio format that works in all browsers. For our online training courses, right now, if we're writing for an earlier version of Internet Explorer, we don't have any problems; it's always going to have Windows media player, which supports certain standard formats. But as soon as we start looking at having it run on the iPhone, we find that Apple doesn't like the format. If Microsoft likes it, then Apple hates it.

MC Press: That does become a problem for developers. And it seems that it is becoming a more difficult problem than an easier problem. I guess the popularity of the iPhone gives Steve Jobs a lot of clout, but Adobe has quite a few followers too.

Hansen: If I were going to make a prediction, I would say that Apple is going to win this battle. Certainly, they already have with the next generation of standards in HTML5. They're basically saying no, we're not going to require Flash as a part of any browser; you should be able to play audio and video in the browser itself. I think the success of the iPhone really made an argument. From the end user's point of view, they don't care about Flash. The people who like Flash are the people who are developing the content. The users don't want their experience any more complicated than necessary.

MC Press: So if you were to sum up where you think this whole thing is headed, how would you put it? Or are there different roads being followed by different armies?

Hansen: I think different companies are farther down the road, and the roads are all heading increasingly toward what we would call mobile computing—people doing things on their tablets and their smartphones. Companies are optimizing their resources through things like virtualization and cloud computing. I definitely buy into IBM's whole vision there. In general, computing is becoming more of a user-driven experience. A family member of mine hates having to install drivers on the PC, but they love their iPad because they don't have to worry about software; it updates itself automatically. They don't have to worry about all the nasty details.

A teacher once noted that people are forced to change their habits in order to make technology work. I think we're over that hump now. It's now the technologist's job to make the device intuitive.

From the line manager's point of view, they've always wanted something developed quickly, whereas the IT manager would want to make sure that the solution was supported properly, and the IT manager also had responsibility for security. It was a battle between the two groups. I think we're seeing more and more the direction the industry is taking. With the tools that are out there now, it's becoming possible to maybe finally reach a consensus where the IT manager is implementing policy for corporate security but doing it in a way like giving an iPad to each of the sales reps, which keeps them from violating security policies.

People are realizing there are tools out there for rapid application development, whereas it used to be if you wanted a modern application, you had to request it and it might get into next year's budget. Now, more and more, you're seeing responses like, "Let me build you a quick one this morning, and you go play with it and then we'll tweak it." And that really is a result of where the technology has gone. There are tools that now allow you to do things like that quickly. It's not perfect, but it satisfies 80 percent of the users' needs and represents a better starting point than 10 years ago when it was a two-year project.

as/400, os/400, iseries, system i, i5/os, ibm i, power systems, 6.1, 7.1, V7, V6R1