RIA Weekly

In this episode, I talk with Alterity’s Brian Sweat about launching their new application, Easy Analytics for Inventory, on the Intuit Partner Platform. We talk about IPP, Flex, cloud, analytics, and the ready to go QuickBooks customer-base of 4.5 million users – a pretty exciting setup for one development team’s story on using RIAs and the cloud.

You can download this episode directly directly and it’ll also show up in the RIA Weekly feed for iTunes and other podcatchers. Or, just use the controls below to listen to it right here:

If you’re interested in an overview of IPP, I suggest last week’s episode with Intuit architect Jeff Collins.

Show Notes

• Brian tells us about Alterity and their new offering, Easy Analytics for Inventory.
• I ask Brian to tell us how they started getting interested in using IPP and what the process was like getting up and running.
• We talk about the use of Flex and how the IPP platform integrates with the UI layer.
• Brian tells us about the ability to more closely track what users are actually using and doing (or not using) in a cloud-based application.
• While customers don’t seem to be overly concerned with cloud vs. desktop deployment – at this early point – Brain says the support and even install time on cloud seems much smaller than for traditional deployment options.
• This leads to a discussion about the other effects a PaaS, like IPP, has on a company’s structure: namely not having to worry so much about billing and other back-office concerns. And then there’s all the marketing Intuit does for them in the App Center.

Full Transcript

Michael Coté: Hello everybody! We’ve got a special edition today where we are going to be talking with a guest about, it’s a nice mixture of — you could say it’s sort of a Rich Internet Application style thing — but it’s also a mixture of using cloud technologies and SaaS things and it’s all over the map with some interesting new technologies. Then it’s also one of my favorites, sort of, I don’t know, you could call it a dark horse topic here where it’s kind of this question of what do you do with, if you have already got an existing piece of software that’s doing well and it’s kind of successful and what do you do with all of these exciting new shiny objects if you are not doing a brand-new greenfield thing.

But with that, I am, as always, one of the hosts, Michael Coté available at www.peopleoverprocess.com and we are joined by the guest for this episode. Why don’t you introduce yourself?

Brian Sweat: Okay. This is Brian Sweat and I work with Alterity and we are a QuickBooks Add-on developer. We have been building software that works with other accounting software packages such as QuickBooks for the last ten years.

I started in 2001, shortly thereafter, Intuit announced that QuickBooks was going to open its doors and they publish the first SDK for small business QuickBooks users. So we jumped on that opportunity and developed one of the first — actually, we were in the first group of gold-certified developers for QuickBooks and that was in 2003.

Of course, that was 100% desktop used a COM-based API for synchronizing to QuickBooks data. They were using XML. They developed their own format of XML, but the technology has changed quite a bit and today we are launching a new application called Easy Analytics for Inventory and what we are doing is building on our experience with QuickBooks and dealing in the inventory segment of QuickBooks, and building the tool that is cloud-based that runs on Intuit’s new cloud platform called the Intuit Partner Platform.

I know in the past, Coté, you talked to a couple of people from Intuit: Alex Barnett and Jeff Collins on what the Intuit Partner Platform is, but of course, I can offer some experience on what it’s like to develop on another platform such as–

Michael Coté: Yeah, that’s exactly why it was — that’s why I have been looking forward to talking with you about it, because it is, like you are saying, we have talked about IPP as far as like an architectural overview and kind of the potentials for it and everything and I think you guys, being someone who has been in the sort of QuickBooks area for a while and doing your inventory management, warehouse management, things like that. I mean you guys have an interesting application to see if all of the great promises of the cloud essentially and then more narrowly, if IPP actually kind of play out for you.

So I mean that said, can you kind of walk us through? Like you said, you were having an existing desktop sort of classic Windows kind of application, if you will, and you guys are now transitioning to doing more, even more cloud-based stuff. So when did you start to go through that process and what was the thinking that was occurring as you were doing that?

Brian Sweat: I am not sure there is a lot of thinking, but for the most part –

Michael Coté: That’s always the best type of software development. It’s just magic. Never mind what happens behind the curtain, right?

Brian Sweat: Right. Again, our first product was ACCTivate! which is our desktop application that we built and it’s a pretty large scale application for small businesses. So it’s, like I said, thick client, Visual Basic, started in VB6 and we have migrated most of the code to VB.NET and definitely Microsoft SQL Server background, peer-to-peer client/server type technology, all on local servers of the small businesses.

Michael Coté: I mean just to interject; it’s used for doing inventory. Could you tell us a little bit about what businesses use it for?

Brian Sweat: Right. What we found is QuickBooks does a really great job and unquestionably, [is] the leading small business accounting package which it’s really good as for a checkbook for a payroll, customers and vendors. But where it kind of breaks down is the advanced inventory features. They have an item list in inventory; they don’t have the ability to track like serial numbers or track multiple locations, like if you have BINs or do any barcoding. If you want to have a scanner that syncs with your QuickBooks data, that’s really not just possible with QuickBooks. So that’s where we came in with an add-on to QuickBooks and that’s what kind of our history.

What’s changing though now is you know with IPP shift, rather than selling and distributing our own add-ons that work with QuickBooks, Intuit wants to deliver the apps for us.

So they have taken on the Apple App Store approach and they are calling it the Workplace App Center and it’s actually within QuickBooks. You know there is a button right on your main toolbar called the App Center and it takes you to a catalogue that has all these applications that seamlessly integrate with QuickBooks, where you don’t have to figure out, it doesn’t work with QuickBooks and all that and they are all certified by Intuit.

So there is a lot of advantages for us that we don’t have to deal with and it’s a lot easier to deploy and to support and to sell because Intuit is taking over a lot of that. And with ACCTivate! we did 99% of the work we had to do on our own.

Michael Coté: And so I mean essentially, as other vendors will say, you’re getting access to this “channel” that Intuit has; you’re able to essentially advertise the services that you have, the application that you have to whoever might be using QuickBooks or looking in the App Center. And it sounds like from — where it sounds like, as part of that, that means you need to run on IPP, right?

So what was sort of like the first stage of getting — like so you decide you want to have access to be in the App Center and get listed here and then so, what was the process you had to go through to kind of get in the IPP to get up and running in there?

Brian Sweat: Yeah, yeah, that’s been a lot of fun actually. Again, before we just — this is something new, we started out; attended some of the presentations when they’ve initially announced it. So I’ve actually been around it since it’s started and one of the things is that the platform they build, they’re definitely committed to this, the Adobe stack.

So we are writing in Adobe Flex and they have actually provided like a Flex plug-in that exposed their APIs, like getting the list of customers, getting the invoices, getting purchase orders. So all of the business transactions are already available using their API.

Now that’s huge. It’s a big, big part of it, because before when we started out and a lot of other developers that have worked with QuickBooks data, when you pull the data out of QuickBooks off to desktop, you have to move it somewhere else to work with it, because their old SDK just didn’t really work well for web-based application. So you ended up synchronizing the data into MySQL or Microsoft SQL Server, and then your app would run off that database.

With the Intuit Partner Platform, since our app runs in their cloud and has direct access to the QuickBooks data that’s also synchronized to their cloud, we don’t have to rewrite all of that data storage, query language, all of that goes away. So you just say, give me a list of invoices, and they’re right there. We don’t have to worry about getting the invoices out of QuickBooks.

Michael Coté: Yeah, and it sounds like compared to the past API stuff you would have sort of in the desktop area, but there is both a – that’s about just access to the data that you would have normally from whatever data source you’re using. But also, it sounds like that there is additional or better sort of data modeling that’s going on that’s sort of more native to the SDK and the toolkits that you’re using. So, you know, from what you’re saying, it sounds like you have to spend a lot of time getting the data in a format that’s actually usable in your application.

Brian Sweat: Yeah, that’s right. It’s fast enough to get it straight out of their format and work with it. Yeah, you don’t have to worry about that. So that is very significant for us, because we didn’t have to worry about all of that; the data storage, where we’re going to store the temporary data, what are we going to do with it, and how do we secure that, because they handle the authentication, the security of the data, they take care of all that for us.

So we are really just as — what Adobe Flex is really good for is building a front end for an existing data service. And they are providing the data service, it’s called Intuit Data Services and our application sits on top of that and just exposes that and provide some nice, again, it’s called Easy Analytics for Inventory.

So what we’re doing is we’re analyzing their sales data out of QuickBooks in a really nice presentation, with nice charts and histograms and we got some new charts plan that aren’t out yet, like pie charts and they gives you that nice drill down and just capabilities that you don’t find in QuickBooks.

Michael Coté: Yeah, because QuickBooks is still, for as mature as a product it is, I mean it’s still very accounting and billing and sort of focused — and it does seem like it has — there’s plenty of room in the QuickBooks community whether it’s inventory management or tougher analytics on top of it which is — It is nice to see when there’s older successful applications that they kind of figure out how to transition doing more open way of running things, so people can layer on top of it, and it seems to make a nice market for you guys as well.

Brian Sweat: Oh! It’s huge, yeah. The term that Intuit is using, is they call it “Connected Services.” So they have four-and-a-half million companies in the US using QuickBooks. So what we are trying to do is just provide as they call it Connected Services, to give them a little more than what QuickBooks has. QuickBooks, as you said, is very mature product, focused on data entry on recording bills, invoices, transactions. But they don’t put a lot of the — they don’t focus on niche markets. So we are doing something that’s specific to people that have inventory and need this advanced reporting and advanced analysis tools.

But if it doesn’t make sense for all four-and-a-half million of their customers, it’s a lot harder to put that into the QuickBooks products. So that’s why they turned us and look for small add-ons that handle one problem and do it really well.

Michael Coté: Yeah, so getting to the technology part, getting back to that a little bit, like you mentioned several times that and of course, part of IPP is using Flex on top of it. And like you were saying, earlier, traditionally, you started out with VB and then you had Windows Desktop stuff, and so forth and so on. How is it been using Flex in this environment; like how is that sort of added to the portfolio architecture planning that you guys do?

Brian Sweat: Yeah, that’s a very, very good transition; it’s been a lot of fun. I have got four developers here that we are all working in .NET that’s pretty much all they do everyday. And we took one person off earlier this year, in the fall to start working on this application in Flex. I’ll tell you, as a product manager, I looked at that and said, okay, well, I hope he likes it, because I worked with Flex and it’s something that I did the sample, I ran through a lot of sample code and actually wrote some small applications that run on the Intuit Partner Platform just in test. I like it; it’s really easy to use, of course, anybody, who has worked with it, can say that it’s a pretty good product. Now, it came along a long way since Flex 2, the initial old version which I never dealt with.

So the transition from — if you have dealt with Java or other languages before, I think it’s a nice platform to work on. And the guy that has been programming out here, he is Austin, is really happy with it and excited to get back and split his time between .NET and Flex and hopefully, we’ll be able to give him a full-time job in Flex and never look back. But –

Michael Coté: Yeah. Since you have been using Flex more, I mean do you start thinking of using it for other products that aren’t necessarily tied to IPP or like what?

Brian Sweat: Yeah, yeah.

Michael Coté: Is it starting to spread around or something or what? It’s interesting to think, like, a .NET shop with a Flex front-end, with Silverlight and other things running around there; there’s other options out there as well.

Brian Sweat: That’s exactly right and we have been talking about that a lot in terms of ease of Flex and what they call the native Intuit Partner Platform apps, which are built on Flex. Does that make sense for all applications that we have got planned? If you are building something from scratch and you are starting on a small app that’s going to run on Intuit and integrate with QuickBooks, one of the things that — I think Flex provides a really nice, consistent user interface. The sense I get from them is that they’d like a lot of the apps to kind of have the same theme or may be like look and feel. So whether or not –

Michael Coté: That’s always a good goal to have for sure.

Brian Sweat: Yeah, I think if you look at Force.com or one of these other platforms, you want the applications to have kind of a consistent user interface. So they are allowing other as they call it federated platform, so you can write a Microsoft Azure application, which I guess is still the kind of new announcement they just made. I don’t know a lot about that, I have kind of been reading up on it over the last months, a couple of months, but again, I had no experience with it. But if you are comfortable in that, then you can build an app on Azure and still have it integrate with the QuickBooks data.

But I think my personal opinion is I do like the idea of having kind of a consistent look and feel. A great example of the nice Flex charting that I like to look at is Intuit recently acquired Mint.com like six months ago or so.

Michael Coté: Sure. My wife is paying bills today and I am sure she has been logging into that all everyday with a distressed look on her face, that’s — all the time.

Brian Sweat: Absolutely. Yeah. So I think it’s a good example of even Intuit is like, the Mint.com approach which uses like great Flex charting. An application like that, you could have some of the code written in .NET, and they are adding a new, like, a way to have what they call server-side business logic.

So rather than everything run in the actual Flash object on the browser, they are going to add some abilities to do some backend processing which is not going to be Flex, it’s Java.

But as for the front-end user interface for these apps, Flex seems to be the big winner here. But we will see and we will re-evaluate if something else comes up.

Michael Coté: So how long is the Easy Analytics for Inventory been out now?

Brian Sweat: Oh yeah, pretty short. I would say, we started with Flex like actual on Easy Analytics just in November. So we built the application very quickly and launched it right around the New Year. I think it was like January 5th or 6th. So it’s only been a couple of weeks.

Michael Coté: It’s fresh?

Brian Sweat: Yeah, it’s still new. We are trying to get a lot of feedback from these QuickBooks users. And the great part is we are watching a lot of — there is always this concern of well okay, all of their QuickBooks data is on their computer, or their server in a closet and their data has to be synchronized out to Intuit’s cloud, as I call it Intuit Data Services.

And so far, I haven’t had any problems with that. I mean, people are pretty open to sharing their data with online applications such as ours. So I guess it surprisingly has been pretty easy to deal with and we have gotten a lot of sign-ups pretty quickly without much marketing efforts from our point of view.

Intuit has done a great job in that though. They are helping out with marketing quite a bit.

Michael Coté: Yeah. So I wonder in the previous sort of IPP discussion, when I was talking with Jeff Collins, who is architect over at Intuit, I was kind of fantasizing about this kind of real-time or near-real-time access to, I guess, the analytics to kind of show you what your users are doing that’s possible in a cloud-based versus a desktop thing. And I wonder if – it’s only been out so you are saying a month. So there is not sort of like the giant pool of data to prove this out.

But I wonder if that’s something that you guys have started thinking about or noticing or something that you can actually get a sense for what your users are doing in your application more.

Brian Sweat: Yeah. Coincidently, analytics is in our product name but they have also — we have the ability to track any actual event within the application with callbacks from Flex. So they have actually set up kind of a system for that. Right now, it’s using Google Analytics, so you can set it to fire off an event when someone runs a specific function.

A lot of people talk about this and it’s an interesting problem that you have to think about when you are going with Platform as a Service like Intuit is the apps don’t run on our servers. So we don’t have full control over logging exactly what users are doing. So Intuit has provided us with some basic information like that, which is cool.

So when I sit down with my coworkers here and we look at a feature like — we have a couple of different features that we put out like right now like in the beta to kind of get a feel of what people like, and I can actually look at that and say, nobody uses feature X. It’s like not even being looked at and it really helps us shape the future of the app which on a desktop product, we don’t have a lot of data like that.

I have all the exception reports from ACCTivate!, our desktop product. So when people have a bug, I can tell how many problems people have but I don’t –

Michael Coté: Yeah, you can detect suffering but not pleasure essentially right. That seems to be the issue here.

Brian Sweat: Yeah, exactly, exactly. And that’s one thing that we are hoping that we put this app out there very quickly just to get started and I am not one for spending a year in development and then launch and find out we got it wrong. It’s raw; we are going to get a lot of feedback on things that work well, things that don’t work.

And yeah, it seems pretty easy to adjust. The way Intuit’s done, it doesn’t seem like a big problem to kind of go with that approach.

Michael Coté: Yeah and like exactly the scenario we’re talking about was the thing that Jeff and I were joking about as you know, you can have the meeting where you could spend like an hour debating if someone actually uses something or if you should do it. But if you actually had some even relatively well, they would not to be perfect kind of just like tracking of what was happening, hopefully. I guess you would find other things to argue for about endlessly, but at least you still wouldn’t have that one.

Brian Sweat: Yeah, yeah.

Michael Coté: The validity of a feature or where you should put your effort. I mean it seems like it’s an interesting product management sort of input that you get that previously was very tedious to get that. Anyway, so it would be pretty interesting because you have that pans out for you guys.

Brian Sweat: Yeah, definitely and that’s a big focus for us. So we are trying to figure out what to do next.

Michael Coté: So, another thing I am always curious about is, especially like I was getting at the beginning when you are transitioning, when you have a non-cloud sort of application and you are moving into experimenting or moving into cloud stuff is, are you finding that for your customers, like are there any benefits that they are kind of noticing to this delivery model or is it more — to put it in the positive way, is it more seamless that it really doesn’t matter either way, they are just interested in the functionality?

Brian Sweat: Yeah, I think that’s an interesting question. For the most part I think, I most people are just interested in the functionality and I don’t think they really — from what I have seen so far, I don’t think people pay or at least, our users aren’t paying a lot of attention as where the code is or whether it’s a desktop app or web app.

But what you can’t ignore is the amount of like the kind of tech-support that we deal with on just getting our software installed and set up and for a desktop, client server, 15-20 different workstations, and all the IT type support that we have had to do for ACCTivate! over the years, a lot of that’s gone with the Intuit Partner platform.

So, there are cloud out, of course, it takes a little longer to get it set up just right from a development point of view and building the product, but once you deploy it, it’s actually pretty easy. I mean, it’s surprisingly easy how that works.

And you know, we don’t have — we are not going to have to have a lot of the same types of tech support, learning all about these different Microsoft SQL Server issues and that kind of thing, because there Intuit is taking over a lot of that for us.

Michael Coté: Yeah, it’s interesting you say that, I mean it is kind of — it’s interesting that you could have to have that shift of well, it maybe more effort of development. And you know maybe, there’s a little bit more effort upfront but then if you multiply it across all the support, you might have to have to actually do on-premise stuff, then you probably do end up, it probably does end up becoming a benefit that it’s easier just to do a web-based or a cloud-based or whatever you want to call that type of thing, that’s an interesting motivation there.

I mean, it’s the old all the software development methodology, optimizers are always talking about. The longer you wait to fix a bug, the more expensive it is.

So, in theory, if you were spending more time in development and kind of that gray area between development and deploying the cloud, then hopefully, it saves money for you guys but then also, it saves time and money for your customers, because they are not complaining about something when they are installing it.

Brian Sweat: Yeah, right. We are definitely a product company. We want to spend every hour of our day, building great software, not consulting and installing things and dealing with Windows bugs and other stuff.

Michael Coté: Yeah, I mean, if you guys have barcoders to read through, you’ve had the hardware junk to deal with too. So there is a whole other — I mean, I don’t [know] if the cloud is going to save you from barcode or barcode readers, but it’s good to have as much spare capacity for that stuff as you can, right?

Brian Sweat: Right. Yeah, that’s an interesting conversation but yeah, but we actually do spend a lot of time supporting on these like Symbol Barcode Scanners and Intermec like the rugged scanners that you see in warehouses and that’s really hard to do for — we are a company of 20 people in Texas and we have got clients all over the country and all over the world now with our ACCTivate! software and it’s really hard to tell them, okay, now scan the barcode and what does it say on the screen, because it’s really hard –

Michael Coté: Yeah, it’s almost like you need to send them like a little webcam setup, so you could just have some camera tracking them, right? So, much that that concept is probably cheaper.

Brian Sweat: A lot of people take digital cameras and just take pictures of what the scanner looks like and then we are like, wait a minute, that’s not even the scanner you told me you had. They said, they call it a Symbol but it’s really by some other off-brand and we have problems.

Michael Coté: It’s like crayon and colored pencils and [Kleenex and] tissues and everything. That makes sense and that’s interesting.

We were talking about support a little bit, and one of the things when we were talking a little while ago that you are mentioning that also the effective using IPP like a cloud-based thing is that, there is sort of backend support you need on your end like the billing and taking care of that changes. So I am curious like, I mean this is the total like non-technology, non-delivery side of it, but like what are those other ways that’s affected you guys to have sort of a cloud-based IPP thing?

Brian Sweat: Well, we are just getting started on the commercial side. So we really don’t have a lot of billing for the Easy Analytics product yet, and we don’t have any billing yet. But the dream is that Intuit is going to be actually handling all of the account management customers’ billing, the recurring fees, they handle all of that for us.

So, it’s good not just for us, it’s also good for the actual QuickBooks users because they have one account within Intuit and they pay for their payroll services, their QuickBooks merchant, like if they have a credit card processing account, and in our app, it’s all going to be on one bill. I am not sure about other apps, but like if you have five or six applications, you pay Intuit but you don’t pay my company, Alterity.

So the problem we have right now is that we spend an enormous amount of time every month invoicing customers’ collections. We have some customers pay by credit card, we have bank drafts, we also have people that mail checks. A lot of that collections and accounts receivable stuff gets off loaded back to Intuit. And of course, they take a cut, they take their fee is, I think, 20% until we scale up and sell a lot of accounts, which is a reasonable overhead for us because we just don’t want to do it. Back to my comment earlier, we want to be building software, not managing accounts receivable all day.

So that’s a pretty good deal, because it’s more than just like, if you go to Amazon EC2 and all that stuff, you still have to manage the commercial side of billing for your app. But with Platform as a Service approach, it’s pretty nice because they are actually doing that for us. So we don’t have to work.

Michael Coté: I tend to kind of like balk at double digit like takes that various people have whether it’s like 80/20 or 70/30 or whatever but, like, to your point, it is true that if you are running a business and you think well, let’s compare that 30% to having, like, people on staff to deal with this instead, or paying someone else to deal with this or the time we spend for it, and it does seem like, when you realize that that’s money you are going to be spending anyways more than likely. It does make it a lot more attractive that maybe you can at least — that you can get the benefits of having things centralized and maybe even optimize it a little bit. So it does that.

That was one of the more interesting things when we were talking that kind of [before the recording] is that there is — I always like to extend the old leaky abstraction idea to everything, and there is like a positive leaky abstraction to something like IPP where hopefully it changes your sort of business process for the better, and streamlines it a bit.

Brian Sweat: Right, I think it does, it’s just really clear to us that that’s a good business model. Yeah, I am sure, you are right, you balk at first that 20% cut, but I think Intuit is definitely making out for that by not only doing the administrative overhead by processing payments and stuff, but they also are actually marketing the app. Remember, they are listing our application on their website.

Michael Coté: Yeah, I guess that’s true. That percentage that they take is not only back office handling but it’s also marketing budget that you guys –

Brian Sweat: And that’s significant. That part is very hard to measure because they are putting our app, like straight out of QuickBooks. There is an App Center button now and it takes you right to their website that has all the apps, and the reviews and you can buy it right there with just a few clicks. So we think it’s a good setup and we are happy with it, a lot easier.

Michael Coté: Yeah, well I mean, I think that’s a pretty good overview of the experience you guys have had and I mean unless there is something I am leaving out, I mean the thing that I am always curious about is what kind of your future plans are with IPP here. Like you said, you just came out with just the Easy Analytics for Inventory like one piece that’s running on IPP. Are there other applications you are thinking of adding or converting over or what do you think you are doing now?

Brian Sweat: Yeah, definitely. Actually, we have got a handful of applications that we’ve got planned. We definitely like the kind of Apple model where you have a lot of small applications.

So we are not trying to build another ACCTivate! which is our desktop app; we are not porting that to the Intuit Partner Platform. What we are looking to do is find other small areas that we can build a nice small app that’s done really well that can help hundreds of thousands of the QuickBooks millions.

Yeah, definitely, we are starting off with inventory because that’s what we know, because we have been dealing with it for years, but the next couple of apps we have, we are going to stray away from that a little bit and just figure out what works, and just go from there.

Michael Coté: Oh, that’s interesting. In the same way that there is like an app economy as people like to talk about it in iPhone-land, I guess you guys are hoping that there will be a similar sort of app economy for Intuit stuff at least when it comes to IPP.

You can actually come up with new products and sort of — it’s interesting to think about in the same way that app is short for application, a lot of applications that get developed, they have the luxury, the way that they’re delivered, they have a luxury, they could be a little smaller then you would have to make a full-fledge application. You always encounter these applications, you always. Every now and then I encounter these applications, it’s kind of like, well that’s more of a feature than an application. But I think I have seen this kind of shift in whatever sort of app economy you have, where really apps that are just a small bundle of features, actually do really well.

You could have a full-fledged application as well, but it’s an interesting opportunity that wasn’t always previously available on the past, because there was kind of like this bucket you had to fill up, and if you only wanted to fill the bucket of functionality up so much, then you have to put in more effort to do it. But anyhow, that’s the free-associating for the episode.

Brian Sweat: And that’s all we want. We want to fill the bucket up.

Michael Coté: Absolutely! Well great! Well, I really appreciate you spending the time to talk about this. I mean like I’ve said several times, I think it’s — I spend a lot of time and I know like listeners of this; they spend a lot of time looking at potentially exciting new application. So it’s good to actually hear from people who are trying to, as it were, fill that bucket with new exciting things.

Brian Sweat: Yeah, I really enjoyed it, thanks.

Disclosure: Intuit is a client and sponsored this episode.

Share This

This week, it’s Silverlight 4, Azure, Google Gears “death,” Flash with Salesforce and RIM, and more. The full roster of Coté, Ryan Stewart, and Mike Downey are back for this one:

You can download this episode directly directly and it’ll also show up in the RIA Weekly feed for iTunes and other podcatchers. Or, just use the controls below to listen to it right here:

Full transcript coming soon!

Show notes:

• RichPlatform.com – Downey’s new blog
• Microsoft PDC 2009 – Azure, Silverlight 4, a little bit of SharePoint.
• We then get into what Azure brings for RIA developers.
• Silverlight 4 – much more “plumbing,” etc. Mike talks about the Managed Extensibility Framework for doing components in Silverlight. And here’s the Silverlight and WPF merging reporting from Mary Jo Foley I mention.
• COM in Silverlight 4 – for Office integration, also looking for how to do it on the Mac.
• Silverlight muti-touch app – Ryan got an HP Compaq L2105 multi-touch screen. TechCrunch’s first multi-touch application post.
• Also, Bing’s new, Silverlight-based maps. Also, see MG Siegler’s (TechCrunch) more cautionary take on this vs. HTML 5.
• Google Gears shut-down – was used for local storage, multi-threaded apps; obviously means Google thinks HTML 5 momentum looks good.
• Then we call back to SharePoint at PDC – SharePoint 2010, a lot of Silverlight integration – “web parts,” which are widgets for SharePoint.
• Silverlight Facebook SDK – we go off on a small tangent about Facebook being the new Aol in the making
• Flash and Salesforce (Ryan’s post)
• Flash and RIM – Adobe support for doing HTML stuff on Blackberry.
• Betas for AIR 2 and Flash 10.1, Ryan’s post
• We wrap up by talking about Ryan’s recent HTML 5 post.

A little more

Here’s some items we didn’t get to in this episode:

• Best & Worst practices for RIAs, a presentation.
• There’s an updated Flex Builder 3 for Linux, that doesn’t expire.
• WaveMaker 6.0 Cloud Development Platform Features Automated Multi-Tenant Support
• Winter Olympics via Silverlight
• Upcoming RIA related conferences.

Full Transcript

Michael Coté: Well, hello everybody! It’s the 4th of December, 2009. I almost said 2010 but it will be here shortly, and this is RIA Weekly Although maybe we should start calling it RIA Monthly. This is one of your podcast hosts Michael Coté available at PeopleOverprocess.com, and I am joined by two of our other co-hosts here and why don’t you start off Ryan with your introduction?

Ryan Stewart: Yeah. So I am Ryan Stewart. I am a Platform Evangelist with Adobe Systems and I blog at digitalbackcountry.com.

Michael Coté: And the other one?

Mike Downey: Yeah, this is Mike Downey. I am a Principal Evangelist at Microsoft. I work on Silverlight and I have a blog at richplatform.com.

Michael Coté: Oh, yeah. I saw it just this morning that you’d finally gotten, that started up in running there.

Mike Downey: It has been one post.

Michael Coté: That’s right. So what made you — was Rich Platform just like an available domain name or like I mean you know platform is always an interesting word. What’s going on there?

Mike Downey: Yeah. I don’t know. I was just trying to get creative and make it sound generic enough that it still applies.

Michael Coté: That makes sense.

Mike Downey: But I am going to talk about more than just Silverlight. So it’s really about the full platform on .NET and all of our Cloud service and everything like that. So I didn’t want to isolate it to just a single topic.

Michael Coté: Yeah. I mean that’s kind of a good segue to — I think most of what we will talk about is PDC was a couple of weeks ago, Microsoft’s big — what does the P stand for? Is it product or something? But basically like Developer Conference and there was a lot of Azure.

Mike Downey: I think it’s professional.

Michael Coté: Oh, professional, not for the amateurs and so there was a bunch of Azure and Silverlight announcements then that we will get to that will be interesting, but so I mean – well why don’t we jump right into that, and start — I mean since you went to that, right Mike?

Mike Downey: I did, yup.

Michael Coté: So why don’t you — I mean from the RIA perspective, I mean there was a beta of Silverlight 4.0 announced a bunch of Azure stuff. Why don’t you give us like what you think the highlights were from your standpoint?

Mike Downey: Yeah. Well, so that there is I think three big things at PDC. There was the first day was really all about Windows Azure which is our Cloud Computing Platform and had a lot of new announcements there, I think one of them was that we are going to release kind of a paid model. I think it was in February of next year. Don’t quote me on that, but it’s around that timeframe, which means it’s finally going to ship, so it’s been in a beta until then.

So the first day was all about that and all kinds of announcements around tooling. There is going to be some great Azure tools built into Visual Studio 2010 which is also in beta right now. And then the second day was really all about Silverlight 4.0 which we announced the public data for that, that’s available and I will talk a little bit about SharePoint because that was the rest of the focus on Share Point just as it relates to RIA and Silverlight.

Ryan Stewart: Yeah, that makes sense. I mean the PDC is kind of becoming an interesting conference. That let the server and tools business in Microsoft, it has all the developer tools and then all the IT stuff and Azure, like they have a lot of conferences now-a-days, that have a lot of overlap. It’s kind of interesting between like for the topics you named off between PDC, and MIX, and then TechEd, there is interesting similarity between them but –

Mike Downey: Yeah, I have the same feeling. I know we’ve got the Microsoft MIX conference as well which is really all about Silverlight, but we have a lot of developers too like PDC was huge. I mean it’s like pretty much the entire LA Convention Center, you have to walk like half the mile between each session.

Michael Coté: And so what like — let’s start with the Azure stuff. Now like we have talked about Azure and like using Cloud Computing and Platform as a Service in relation to RIAs and things like that. I mean the thing I always get to is it seems like for obviously not all the RIAs that you are writing but definitely for — if you are writing RIAs, definitely using Cloud–based things whether it’s a platform or you are building up your own infrastructure seems like a pretty nice opportunity to kind of quickly get involved in doing things.

And I guess it was always the last year so that Microsoft has been kind of like set — had a review of what Azure would be and kind of have been out there. But, now that there is like the big release out there like from the perspective of doing RIAs like what do you see as like — what’s Azure providing for people?

Mike Downey: One of the used cases that I encountered earlier this year before I joined Microsoft was I think the one that got me really interested in Azure. I was working with a group of friends and we were kind of twined with the idea of starting a very small company that would be developing a multiplayer game and as we were kind of thinking through how we would do that and starting to kind of write some of the client code, at that time, we were doing the clients all in ActionScripts and Flash.

And when we started thinking through the overall architecture we realized that in order to do this multiplayer game, we are going to have to have a rules engine that’s set on the clients and then that same basic engine would have to also live on the server and validate moves between players. And the most part it was the same code and so when we started going down the path of using — in that case I was using Flash because everybody I was working with they were all Flash people. We started to realize that once we got to the server component, which we had intended on hosting on the Cloud anyways for cost reasons, we realized we’ve got to rewrite a significant amount of code, that game engine in order to ultimately build the game.

And that’s right, when one of us kind of identified Silverlight and Azure as an alternative to where we could then write our client code in C#, .NET for Silverlight, also then build a desktop client and hopefully in the future be able to do a lot of mobile development with that same language. But then also take that same investment we’ve made in code and deploy it on Windows Azure in the Cloud and have that pretty much already done like save a huge amount of time. And I think that use case is really, really compelling. So have the same framework and language and tools for both the client and the server and be able to deploy all that in the Cloud at a low cost.

Ryan Stewart: Yeah, and I think that last part is like the — like especially over the Thanksgiving break, I found myself explaining what Cloud Computing is to like civilians if you will, like family members who were asking like, what exactly is this Cloud Computing crap I am hearing about? And all of the stuff I mean at the end of the day unless you kind of tweak the explanation that is like, well, that’s just like hosting things or something. They understand hosting a website, and it seems like that metering, or the cheapness that you are getting to is one of the core things that makes — for developers that make something like Cloud Computing much different than hosting where essentially you don’t have to — because it’s metric you are not really having to pay anything upfront, it seems like with the extra platformization stuff, so there is sort of middleware if you will, that’s something like Azure gives you. Then it’s definitely — I don’t know, it gives you a better chance and if you’ve just got like a naked Linux shell so to speak that you are paying for monthly where you’ve got to build up your own infrastructure.

Mike Downey: One of the nice things about Azure too is that we will be able to roll in other Windows server services as part of that. So like if you want to do smooth streaming video, you will be able to deploy that through your Azure instance. You can leverage all that stuff, SQL Server, and all those other types of things all built in the same platform, and you pay as you go. So you just pay for the asset that you are using.

Michael Coté: Yeah, that makes sense. So would you say that the Azure things are — are you sort of running stuff in Azure and/or the sort of services that you can use for, like for example with Adobe has some Cloud-based services and some of those are more sort of like interfaces and services that you are accessing versus sort of run-times and containers, even though there are some sort of container stuff especially with the collaborative stuff. But, like which side would you say Azure falls more on? Is it more like a run-time or services that you access?

Mike Downey: Well, I would say it’s more of the former, but what’s nice about it is, because it’s all the same platform you can easily access services within it. But this is more of a run-time platform on the server side, right? So it gives you the ability to write your own applications in that that are leveraging all those capabilities. So you really can get access to both.

Michael Coté: Right, that makes sense.

Ryan Stewart: And it’s a lot more like Google App Engine that it is like Amazon’s EC2 or S3 where you have kind of a dummy storage or dummy operating system that you can spin up in the Cloud and much more about building sort of things on top of that and then you guys expose services that way, right?

Mike Downey: Yeah, I think it’s kind of the combination of the two. It’s kind of like — I mean it’s Windows Server, right? So it’s most of that functionality. So if you want to leverage SQL, if you want to leverage IIS, all those things but yeah, that’s basically it.

Michael Coté: Well, so can you spend/spin up random, buy a certain amount of computing time on Windows. Can you basically just spin up random instances of Windows Server then when you subscribe to Azure?

Mike Downey: Well, I think it’s based on kind of metering, right? Where like I think you can artificially ramp it up if you are expecting traffic and things like that, but I think as I understand I am not really a Cloud Computing guy myself but a lot of that just happens automatically.

Like, all of a sudden your application demands a lot more resources, it will just automatically start creating more instances or dedicating more memory and more storage base and whatever has to happen.

Michael Coté: Got you! So like with Amazon EC2 the way I understand it at least, so you can have different images up there, so you can have an Ubuntu Linux image, you can a Linux instance with ColdFusion pre-installed or whatever you want to do and then spin those different ones up, so there are ways that you can pre-configure your Windows Server instance with like the Web Toolkit or PHP support. Then it will scale as you need to, but if you want to have separate server instances, does it allow you to do that?

Ryan Stewart: Oh! Yeah, I see what you are asking. That’s a good question. I don’t really know the answer to that.

Michael Coté: I don’t know anything about Cloud Computing really either. I was just — Azure is such a kind of a morphias thing in a good way, it’s just so all encompassing that I wasn’t sure exactly which parts of it compared to which parts of the Amazon stack.

Ryan Stewart: Yeah, I know what you mean.

Mike Downey: Yeah, from what I know of Azure, I always hesitate to answer something unless I have the text right in front of me, but I am pretty sure that’s the way it works. That’s the other thing somewhat related to RIAs, about Azure is — or Azure however you want to say it is, it’s actually kind of like, it’s sort of like the surprising new Microsoft that they support PHP and all sorts of open-source stuff things.

There was even a post by Matt Assay, who is sort of the business of open-source blogger here and there, and he had a post that’s kind of like watch out about Microsoft like taking over open-source, which was kind of like — when you actually read the post, it’s more of an headline grabbing thing of saying it’s kind of crazy that there is all this open-source stuff supported on a Microsoft platform, which is pretty interesting.

I mean there is MediaWiki on there and MySQL and essentially the LAMP stack or I guess the WAMP stamp or you might even call it the AMP stack, who knows what you call it.

Michael Coté: A WAMP Stamp.

Mike Downey: There you go. But, yeah — so with the Azure stuff. So then shifting the Silverlight business, it is basically about the Beta release of Silverlight 4, right?

Ryan Stewart: Yeah, exactly. So we disclosed all the major features and functionality and at the same time released a public beta, the run-time, which if you have been keeping track of Silverlight history, this is an even more accelerated development cycle than the last one. So it seems like with each release, it’s getting faster-and-faster as far as release time, which as a side-note seems to be a problem if you are book author, because it’s very difficult to get a book out on it. By the time you get the book published, the new version is out.

Michael Coté: Yeah, I think AIR and Flex over the last couple of years — a couple of years ago, I went to, at least I saw it on my bookshelf or I think I still have some little AIR pocket reference here and there, but it seems like people like O’Reilly have this good strategy around that where they make these pocket references as sort of like a bridge between the two.

But it is interesting. I wonder if online publishing helps with that more, because it kind of is –

Ryan Stewart: Well, I was just talking to O’Reilly yesterday about that and that’s exactly what they are looking into doing is doing more of those kinds of pocket guides and then augmenting it by online publishing, so they are getting these guides out faster.

Michael Coté: Yeah, I mean like I am staring at a Flex 2 book now which I know is not that old, right, and it’s kind of like not very useful now, and there are all sorts of book like that, but it is — anyway, that’s an odd rabbit hole to go down, but yeah, like to your point of –

Ryan Stewart: Sorry!

Michael Coté: No, no, no. I remember back at the Silverlight 3 launch when I was talking with Scott Guthrie, kind of asking him like what the general theme of Silverlight 4 was going to be and he said, I am casting this in my own words, but more or less it was like making Silverlight a real development platform if you will, as far as a real UI platform and by real, that just means comprehensive and has a bunch of stuff in it.

I have been looking over the Silverlight 4 like quick Beta list and summary of stuff and there is sort of a crap pile of normal, boring almost system level stuff like just accessing the My Documents folder and doing all sorts of little things like that which do seem to make it a pretty beefy release for a UI technology.

Then I guess this sort of connected to that is there are people like Mary Jo Foley and some registered people have been kind of hunting down quotes from people like Brad Becker over at Microsoft saying things essentially like Silverlight and WPF are eventually going to merge into the same thing and have Silverlight be the UI technology of the Microsoft world, which if you are looking at the release notes and the public quotes along, then it does certainly seem like a fair trajectory for the technology.

Mike Downey: Well, as the Silverlight event that would be pretty certain for me, but I want to comment on any kind of future technology releases, but yeah, interesting point though. So you are right, like there is a lot of pluming in this release, that kind of low level functionality that you mentioned and I think you could probably best summarize it, it’s a release, it’s very heavily focused at RIA and enterprise level applications.

So a lot of what you see in Silverlight so far has been very media-centric. That’s been really where we focused early on and trying to win the battle is really around media-centric applications and with Silverlight 4 I think you are going to see a much heavier focus move into enterprise RIAs, and really approaching it from that perspective.

So we’ve done a lot of things, in the pluming area that kind of helps make that happen. And at the same time we are also, for me and partly because of my background working — I was one of the first product managers for Adobe AIR, one of the most interesting things I have survived for is the Out of Browser features that we are adding. We started that in Silverlight 3, and in 4 we are really expanding on what you can do with the Out of Browser application. I think most notably is that you can host HTML within your Out of Browsers for light apps.

And that HTML content can post any third-party application as well, so you can do some really interesting applications by using that.

Michael Coté: So do you think something like that would – and I am just making up stuff, I forget all of the social networking legions, but as an example, you could put Facebook, connect things in or Google widgets or something like that. I guess, that’s one of the benefits you would have if you had a full HTML and JavaScript browser in Silverlight, right?

Mike Downey: Yeah, I mean that’s exactly the kind of these cases I would think of is doing that type of content, accessing social media capabilities and then having interface between Silverlight in that content. So that kind of JavaScript API between the run-time and that being incorporated in the app.

Michael Coté: Yeah, I have been sort of realizing over the past couple of weeks that the other social media people are really sort of — they are still like — I guess there are basically two implementations of it, but they’ve really finally ironed out the — I don’t know what you would call a consumer single sign on, if you will and there is really good identify management for going — it seems like the two areas of the Facebook connects stuff or there is sort of a Google account, but man, those guys like pop-up everywhere as far like managing the credentials that you have to log into things.

It would seem like getting — if that sort of like identity management to use a really boring term for it is sort of like spreading as much as it anecdotally seems to be, you definitely will be eager to get it into Silverlight and Flash and all the other RIA platforms otherwise you have the same annoying having to log into things all the time.

Mike Downey: Yeah, I totally agree. And then the other area I think that is really notable at 4 is text. We are doing a lot of workaround text, having full bidirectional text support and better layout management and rendering and a lot of really great stuff there. And that’s part of that enterprise push of course, because when you start building more applications that kind of thing becomes far more important.

So, yes, that’s kind of one of the big areas is text. We also are doing some stuff around — I just learned about a couple of days ago for a project that I am working on which is the Managed Accessibility Framework which basically provides a much better model for creating modular controls and libraries, and we are talking about within reference to the Silverlight Media Framework that we just launched at the same time, which is basically like a media player component framework that we’ve released on CodePlex. People can build their own video players and their applications. So we are looking at docking this NEF in sort of like four shifts.

Michael Coté: So it’s basically a way of defining and I guess deploying and then using components or whatever in Silverlight which is what MEF is?

Mike Downey: Yeah, and we are using code between applications. It really becomes really useful I think when you do really large scale composite applications, so you can reuse code and save on file size and things like that.

Michael Coté: Yeah, yeah. So the other thing to mention that Tim Anderson has written awful lot about is the COM support that’s in Silverlight 4, and I mean since we have you here — what exactly is COM in first place?

Ryan Stewart: So this is basically an interface that you can call some Windows specific API. So we really intended it so that you can have Silverlight apps communicate with Office.

So it’s a great way if I am scripting Office and accessing Outlook, if you want to access a calendar or something like that in your application or access Excel spreadsheet data, there was a great demo of that, a PDC actually, where you are calling out to Excel, having it launch and create a spreadsheet and then reading that back and doing application.

So, it’s essentially kind of like scripting the OS for specific Windows APIs. So our focus on that is enabling office integration and we are also in the process of looking at how we can do the same thing on the Mac.

Right now COM is a Windows-only feature and there isn’t really an equivalent on the Mac. But we are in the process of kind of evaluating how we might be able to do something like that on the Mac using something like AppleScript. But it’s not a parallel capability. So the tricky part is providing a developer API in Silverlight that is consistent between platforms, so, you don’t have to write a lot of If tabs in your code. So, it’s awesome.

Mike Downey: Yeah, I am like far from an expert on COM or even sort of Apple, like Intra or inner application communication either on Windows or Mac. But on the Mac there is like that wacky services thing and there is always AppleScript being used. But it’s never really, I don’t know, I have never gotten the sense that as a whole the Mac community is really taking advantage of that stuff as much as they could or although that said –

Ryan Stewart: Yeah. I haven’t either and I am not even sure how heavily COM is used, I mean, I think, because of the Office, that’s probably the primary used case for COM, it’s integrating with Office. Beyond that I am not really sure how much it’s used.

I would imagine that most of the used cases are going to be behind the Firewall in really Custom Enterprise applications where you have a pretty captive audience and you could have custom applications that are used within that enterprise that you want the Silverlight experience to be able to tie into.

Mike Downey: Yeah, now I mean to your point in theory, there is a lot of like custom COM code out there, to be all alliteration-friendly and if you wanted to convert your old stuff over into using Silverlight or something else that could rely on COM then it would probably be handy to like stay with that, those sort of backend things, if you will, rather than having to go rewrite the connectors for everything. But anyways like you said, we’ll have to see what’s actually out there.

Michael Coté: So what percentage of Silverlight developers you think are going to take advantage of that Mike, I am kind of curious. Because to me that was the big shot across the bow of WPF, I know you won’t comment on that. But the COM support was kind of — here we are going to provide some more data function or some more interlope with the rest of the MIX stuff platform at a level that we can’t necessarily replicate on the Mac as easily. So, I am curious as to what you think, how many percentages of Silverlight developers do you think will use the COM feature?

Mike Downey: Well, I think, it’s probably something that was driven by existing Microsoft developers who are like Coté said, we are kind of in a process of migrating applications and don’t necessarily want to recreate all of their different libraries that they have, that have existing functionality integration. And so I think, because Silverlight 4 is really heavily focusing on enterprise applications and going into more into the RIA space, this is probably a requirement that was driven by some of these big enterprise IT groups and that they want to do things like that and they want to use Silverlight to do it.

So, I think that makes sense, if that’s the type of application you are going to deploy and you want to access that type of capability, why not give developers that functionality. We are also, of course, always interested in making sure that, we maintain that kind of cross platform capability as much as it’s possible, that’s really the intent of the run-time. But just like multi-touch in Flash Player and in Silverlight the APIs aren’t there on the Mac.

So, if we want to provide multi-touch there has to be Windows-only for now until Apple provides that API under OS. So, there is going to be cases where the functionality, there just isn’t a parallel another platform, but once there is, there are some work around, then our intent is completely to support that in all cases.

Michael Coté: So, that’s actually, Ryan was mentioning before we were recording that he wanted to talk about the Silverlight multi-touch stuff, so that’s a nice little parenthetical aside there. Where did you get your hands-on there, Ryan?

Ryan Stewart: Yeah, so I picked up an HP Compaq, I guess L2105 touch monitor. So it only does two input points, it doesn’t do genuine multi-touch but a lot of the stuff is really just gesture-based anyway. So the number of points don’t really matter.

So, the only way it works right now is on Windows 7 because as Mike said, the multi-touch on the Mac is a little bit weird and so right now Silverlight 3 is the one that supports multi-touch and so we don’t have any great Flash Player 10.1 samples. But I have been really impressed with Silverlight’s multi-touch and kind of, I guess what brought it up was recently TechCrunch had a blog-post about “First multi-touch application” and it was built on Silverlight.

And so I was happy to see the people are actually starting to build stuff on Silverlight and hopefully for Flash when 10.1 is more widely available. And I think that means good things from the UI because it’s really fun to just sit here and touch and feel like your minority report with these applications.

Michael Coté: What’s the first app that the TechCrunch wrote about?

Ryan Stewart: It was some cheesy little — I mean it wasn’t spectacular. It was some cheesy little kind of like a movie browser basically. There were like nine panels on the screen and you could pinch to zoom out and see all nine or pinch to zoom in and see an individual one and then it supported gestures, so you could kind of browse around.

So, it was a terrible implementation but one of the cool things was the Bing Maps now has Silverlight support. And so Bing Maps has some multi-touch support, so that you can pinch and drag your map. You can double-click to zoom in and so if go into their 3D mode, you can basically just with your hand, walk around your city, if it’s supported in the street-view mode.

And it’s just a much more interesting and engaging way to browse content. I thought the map example was the best one in terms of just sort of immersing yourself and making gestures and your hands part of the overall experience.

Mike Downey: And you can have like the CNN war room in your living room. I mean, they have those crazy big touch panels that they are always futzing around with. At times –

Ryan Stewart: Exactly! I will be Wolf Blitzer and you are flipping around States.

Mike Downey: I am glad you mentioned that Bing thing. When you are inside the Microsoft network and you go to Bing, it automatically shows you an internal version of Bing that has features that the external version doesn’t have yet. So, I will come across something and I will stop saying, oh man! I wonder if this public or not, so I just don’t talk about it because I can’t tell which stuff has been released and which hasn’t. So I am glad you mentioned that.

Michael Coté: Yeah. I went to check those out early this weekend. The maps are — especially — and it’s just a matter of the photos taken but the satellite view is pretty cool in a sense that maybe Google maps has this but I haven’t gotten to it but it has more of a three-dimensional look to it, so that you can kind of rotate around and it’s fun to like go to maps.bing.com or you have to get into their beta thing and then, of course, you need to have Silverlight. But you can kind of rotate around your house and it’s always fun with satellite maps to play the game, what year was this picture taken in. And like at my house, my neighbor, he used to have like a rotting old, like ‘62 Chevy or something in his backyard and it got taken away a couple of years ago and you can spot the car on the photos there, so you can figure how old these photos are.

But yeah, the maps on there are pretty impressive and I would imagine, no matter what — I mean just like playing with Google Maps on your iPhone or if you are playing with whatever maps on a bigger machine, it’s pretty fun. It’s like just to fiddle around with that. It’s pretty easy to waste away a lot of time.

Like, you are saying, it would be fun if there was a little thing where you could be a little walking guy with your fingers and have that be a gesture to walk around the map. I think that would be damn cute.

Ryan Stewart: And you could make it like the actual, like move your fingers while you are walking, that would be awesome.

Michael Coté: Yeah, yeah. I don’t how would track that.

Mike Downey: Speaking of Google I heard that, Gears shut-down. I don’t know if Ryan remembers that, but when Google first announced Gears on the scene, we were talking to them about how we could message that and how Gears is, in a way, kind of competing with AIR and all that. Now, that wasn’t even that long ago and now it seems like they are already shutting it down.

Michael Coté: Yeah. That seems to be, I guess they open sourced it a while ago and in retrospective sort of obvious what they were doing because nowadays, they talk about HTML5 all the time and everything and they haven’t really mentioned Gears but I guess, we can now really think of Gears as Google’s like, I don’t know, UI/RIA sort of thing. I mean, they are definitely pull hug on. And whether it’s their own Chromium browser or more importantly, on just HTML5 stuff. But I mean, I guess –

Ryan Stewart: Well, I think, I think Gears, I think that was the –

Michael Coté: Gears is mostly used, it seemed, in practice for a local storage and doing sort of like offline access and everything, right? Well, it’s used for like multi-threading your apps too, right?

Ryan Stewart: Yeah, exactly the 29:21, and so I guess what I thought was really interesting about this announcement was that, I found this in a couple of things. So, one, I think it means that momentum for HTML5 is at a point where Google felt comfortable kind of backing off a little bit on that. And Google Gears is a plug-in and just going with HTML5 is the way to drive innovation.

They also have Chrome so they can cheat a little bit because they can discontinue Google Gears but then they have this Chrome plug-in for IE. But I think it does, I think it’s a big day for the HTML5 people, because I think it does mean that sort of like Apple has done completely focused on making HTML better inside the browser that Google is now going to go in that direction, not worry about a plug-in.

Michael Coté: Yeah. Now I think that’s spot on. There is even — in doing my little RIA consulting here and there, there are some quotes from Google people where — I forgot the exact wording, but they basically say, we are putting all of our money on HTML5. So I think Google is clearly in that camp.
             
Oh yeah, we were also going to talk about SharePoint briefly in reference to PDC. So other than it being a Microsoft technology, and probably — I forgot how many; there was some billion odd dollar of revenue nowadays that SharePoint has, which is more indicative of how much it’s used, but what was interesting to you about SharePoint at PDC.

Mike Downey: Well, so SharePoint is like a behemoth of a content management system. I think it goes way beyond that and does document management and a bunch of other things, and of course, it integrates into Office, as you would expect. And apparently, it’s doing really, really well. I guess it’s one of our fastest growing businesses.

But the SharePoint 2010 was announced and I think it’s shipping really soon, if not, just recently started shipping. But the thing at PDC that I thought was interesting and relevant is that they have done a lot of integration with Silverlight. So now, there are all kinds of built-in Silverlight, what they call, web parts that are essentially like widgets for SharePoint. And then there are also APIs for building Silverlight widgets in SharePoint.

You can also like consume web services in Silverlight. That’s an easy way to do all of that. So basically, you can use SharePoint as a back-end for a Silverlight side, and it’s really easy to do that kind of integration and then offload content management over to SharePoint for that Silverlight application.

Michael Coté: Oh yeah, yeah. Well, that’s interesting to see it’s spreading to — I always think of SharePoint as like one of — it’s kind of Microsoft’s current kudzu, like it just spreads without even trying, and it certainly is everywhere. So it’s a nice plant to attach yourself to.

And I think the last Silverlight thing that I at least had to mention, it transitions nicely into. I guess, the more the Flash portion of the news we are going to go over is there — this is a lot like, I think only a few days or a week after we recorded last time, but there was a Silverlight SDK for Facebook out there now, which I think I talked with Brian Goldfarb a while ago about this. And he was showing me some demos of things where it’s basically, there is a huge App Ecosystem out in Facebook and there are essentially wrappers around doing the whacky Facebook stuff so that you can implement things in Silverlight, which I suppose brings you lots of videos and the other neat visualizations and things like that, that you could do in Facebook.

Having fiddled around with Facebook apps a little earlier this week, when I was setting up a RedMonk fan page, man, that stuff is tedious to do like apps on Facebook. It’s really weird. It’s strange even to like find where you go to do things, if that makes sense. It’s just not as intuitive as I would think it would be.

Mike Downey: Yeah. So I mean, hopefully with these SDKs, like there has been a Flash one for a while now, another one is for Silverlight, hoping that, that really makes things a lot easier. Of course, it’s also going to end up with a lot more applications out there in Facebook.

Michael Coté: Yeah. I mean there is definitely a — I think Facebook is an interesting platform like other platforms. So there are Salesforce or Intuit or these other people, who have — they have a bunch of data and users who are on there all the time. I don’t know, I think there is a lot more runway, so to speak, for doing something with all of that, beyond the kind of simplistic apps that are out there. But then — so speaking of — sorry, what were you going to say?

Mike Downey: Well, it’s starting to remind me of the early days of the Internet with like CompuServe and AOL, where you are kind of in this captive environment, where everything is specific to Facebook, and it’s like it’s backing into that, which is kind of interesting.

Michael Coté: Yeah. I mean that’s definitely, I always think of Facebook as the new AOL, in the sense that it’s like — if you really want it to and I have put that in like dripping irony, you can pretty much just stay in Facebook for your whole Internet experience. I mean, you have got emailing and IM and you have got apps and everything, in the same way that you could stay in AOL.

And then the other thing that’s weird is like, it’s really difficult — well, I think it’s really difficult to kind of get Facebook to work with the rest of the web. Just figuring out how to hook up an RSS feed so you are like cross-posting stuff, requires like way too much work, and obviously, it’s in Facebook’s interest to make you stay in Facebook.

But as it sort of increases in size, it would kind of be terrible for them if it did become that kind of wild garden like AOL was, where it starts to rot after a little while just because there is not real stuff happening in there.

But, yeah, speaking of API, there was also like Adobe had a big partnership with Salesforce around — basically, maybe you can fill in the details I get wrong Ryan, but it seemed like, essentially, there is a version of Flash Builder, if I remember, that you get that sort of has a Salesforce SDK in it, in the same way that you can build apps on top of Facebook, or analogously, I should say. You can start to use Flash to build UIs on top of Salesforce and things like that.

Ryan Stewart: Yeah. So basically Salesforce has their own Eclipse-based IDE. And so we did some work with them to incorporate their stuff into Flash Builder, so you can get one unified IDE that includes a lot of the productivity enhancements that Salesforce has; makes it easy to connect to their APIs, has some built-in Flex components that you can, I think, drag and drop on to the design view to get started very quickly.

But yeah I think it’s — we have been pretty closely with Salesforce for a long time, and I think they are just as excited as we are about what that means. So this was just another example of kind of coming together at least at a tooling level, and making it available for developers in one kind of easy-to-use package, so that you can start integrating Salesforce and Flex applications pretty quickly.

Michael Coté: And did you get — I think you wrote a post about this, but you did get a sense of like the kind of apps people are building or thinking of building with Flash?

Ryan Stewart: So I have been spending a ton of time with it. James Ward has been doing a lot of the lead on the Salesforce side. But I mean, any kind of CMS or any kind of Customer Management System that you are building and you want a little bit better UI. I think one of the cool examples is kind of that offline-online context. So if you are a sales guy, and you want to have access to all your contacts and all your leads, you are offline, you can use AIR pretty quickly to do that, and then synchronize your Salesforce content, as soon as you get back an Internet connection, you just push it back to the Cloud.

So we are trying to make workflows like that a little bit easier, because I think that’s one, the online-offline case, is kind of a killer for that semi-disconnected salesman on the road.

Michael Coté: Oh yeah! That makes sense. Yeah, I mean, I think, I have even mentioned this several times, but that’s been a curious sort of use case for RIAs that have offline ability that I have heard from people is, whether it’s people basically who are out in the field and they need some application for like blogging different data that you are doing; whether they are sales people or whether they are maintenance crews or people like that who — they might be using like a little handheld thing or a laptop, but they basically want the same application that they can just go like, plug into their network and have it synch everything up, which is something you could do with a traditional gooey for sure. But I would imagine, whether it’s Flash or Silverlight or whatever, that kind of use case is usually built in from the beginning to use, so maybe it’s a little easier to use those kinds of platforms for.

Ryan Stewart: Yeah, exactly.

Michael Coté: And so you have these –

Mike Downey: When we first launched AIR, I used a lot to go around talking to the press and analysts all the time, saying imagine a use case where Salesforce builds an app that runs offline, enables sales people yaddi yadda, and it’s great to see that it is finally an actual application, because that to me, that’s the best most compelling use of an offline RIA, and it was always vaporware when we talked about it.

Ryan Stewart: Well, now it’s possible. You can grab the toolkit, Mike; you can build your dream back then.

Mike Downey: That’s right. So then the other sort of Flash announcement, another partnership you had was with the Research in Motion (RIM) people which is basically using Flash to do BlackBerry stuff.

Ryan Stewart: Yup, yup. So yeah, basically they joined on to the Open Screen Project, so they are interested in using the Adobe tool chain to help their developers create content. So we announced the Open Screen Project that will support Flash and Max, and this latest announcement was kind of, we are going to help work the tooling, we are going to provide some BlackBerry profiles in Device Central. It would just make it easier for our both HTML developers and designers and Flash developers and designers to create content for those devices.

Michael Coté: Right. So I mean, do you think it’s fair to say that you can write BlackBerry apps in Flash now?

Ryan Stewart: So I don’t know, that’s not a fair statement, because I am not sure what’s — because AIR is going to be our mobile application platform. So you would use AIR to create applications, and we had to do something very different with the iPhone because Apple doesn’t play nicely with their partners, in some cases. I don’t know exactly what’s going on with BlackBerry, but as of this announcement, there is nothing else new to report other than we are working on them with AIR support.

Michael Coté: Okay.

Ryan Stewart: Yeah. So then that’s how you build full BlackBerry application. This is more for kind of web content or whatever HTML Shell, BlackBerry has that lets you to create HTML applications.

Michael Coté: Oh, so this is kind of like, it kind of like complements the -– the other thing that MAX was like basically a way to develop iPhone apps in Flash Builder and the Adobe tool chain. And then so this is a way to develop BlackBerry apps in the Adobe tool chain essentially.

Ryan Stewart: No, no, no, so this is not that, at least, not at any kind of native level. I mean it will be kind of whatever — if BlackBerry has an HTML wrapper like you can build sort of native apps with HTML. You can do that but there is no Flash to compiling the BlackBerry at all.

Michael Coté: Okay. So it’s just HTML, not native support. That makes sense.

Ryan Stewart: Yup, exactly. And then we, of course, have Flash player in the browser that they are going to support when 10.1 shifts.

Michael Coté: Right, right, right, that makes sense. So I think the last like little news item was, I mean basically, there is beta for AIR 2 and Flash 10.1 out, and you had a wrap up of that, Ryan. But what are like the highlights going on at the moment with those two betas?

Ryan Stewart: So the biggest highlight in my mind is the CPU usage. So because we are doing a lot of mobile work for Flash player 10.1 because that will be the runtime that is in the browser on those smartphones, we completely overhauled the virtual machine and optimized it and went through and cleaned up a lot of things. So if you download both for AIR and Flash player, if you download the betas, you should see a lot less CPU usage for your applications. So already, I have gotten a lot of feedback from the AIR 2, the beta is there, that people are seeing the huge improvements in what used to be a CPU hog a little bit for AIR. So I am exciting about that, it’s kind of the biggie at least.

And then some of the other things that I am excited about are multi-touch support, we’ve also got a lot more native functionality in AIR 2. So if you want to create native applications, you can bundle them as native installers and then access basically any code that you want through whatever bridge you have to do. So it will allow you to take your AIR and Flash development skills and start to experiment with native applications.

Michael Coté: Okay. So it seems like both Silverlight and Flash will have lots of native access coming up in the next year or so, over the next releases. That should be like an interesting expansion of what people do with things, because there will be a — I mean it seem like there would be a lot more desktop-centric things happening or at least integrating with everyone’s desktop instead of just being web-bound.

Ryan Stewart: Yeah. I think –

Mike Downey: Trying to make the better use of the operating system too.

Ryan Stewart: Exactly, yeah.

Michael Coté: Yeah. I mean there is already like native graphic accelerator sort of integrations here and there but it is always kind of funny to hear about hardware and OS integration, where it’s like now it finally works with webcams, or it finally works with mics and USBs which are, I guess, things you kind of take for granted when you are using traditional UIs. Those kind of things would come in pretty handy, pretty quickly.

So was there anything else you guys wanted to go over?

Ryan Stewart: No, not on my side. Anything you want to talk about? Any cool RedMonk client’s things happenings in the RIA world?

Michael Coté: No, no, it’s the winter slowdown. Once that’s come, everything starts to become very slow. One thing I was going to — you had a post, Ryan that was basically HTML5 can’t happen without Flash. I don’t think you said, can’t happen.

Ryan Stewart: I think in the title, I did say, can’t happen.

Michael Coté: Okay, yeah. But I always like those kinds of posts because the content is always interesting but it usually generates some interesting commentary around it. And I wonder like there was a moderate amount of people leaving comments. What do you think the cinnamon has been for that?

Ryan Stewart: I think it sort of depends on who it is. I think if you talk to rational people, I think everyone understands the value that Flash has brought to the web in general. I mean things like video, things like 3D, things like multi-touch that we are adding now and Silverlight has added. So I think people understand that Flash did good for the web in general and has kind of been able to lead, like several reasonable people will agree that Flash has been able to lead some of those things. I think a lot of people just want HTML to be around for a lot of different reasons. It’s kind of a baseline; it’s supported on the iPhone as well as every platform. Standards are always better when you can have them than having a proprietary platform. I think people sort of universally agree with that, but there are drawbacks to that and when you are going through the standards process, you just won’t be able to be as innovative as someone who doesn’t have to do that.

And so I think what my basic point was is there are a lot of bad uses for Flash but one of the good uses for Flash and Silverlight will always be to create cutting-edge content and to be able to experiment so that the web can see what works, what’s important and then the standard’s bodies can kind of coalesce around those things and those experiments that our communities create and then make them a part of the entire web.

And I think that’s worked pretty well in the past four years. HTML5 has been a long time coming but I think we will still see that. Flash and Silverlight can innovate and add things and experiment and push the envelope and then HTML will take the best of those experimentations and codify them into a standard that everyone has. So I think that solution works.

Michael Coté: Yeah. I mean this is like the time of year where everyone is all excited about doing predictions for the next year and it definitely seems like things are well poised to — I mean I say, hopefully, because it was a fun time but hopefully we will have that same. Next year will be the start of like the sort of like hot and heavy like browser sort of wars, if you will, or whatever that modern equivalent of that is where essentially, there is a whole lot of people innovating outside of the standards body just because they want to go for speed and coming up with all sorts of new things. But another more important aspect is finally winnowing down to whatever the de facto standards are and it does seem like — we’ve mentioned a few times here and there and even in this episode but I mean HTML5 is interesting. There is a lot of people who are putting a lot of big brains behind implementing stuff for it. So that should be fun next year.

Ryan Stewart: Yeah, exactly. It will be a good year. I think it will be great year for RIAs.

Mike Downey: Steven Sinofsky, who runs the Windows group at PDC, he showed off some of the features of Internet Explorer 9 and they are doing a lot of really interesting innovative stuff around HTML5 and also, I mean I think what you are going to see and it’s going to happen a lot I think is a competition to create the fastest JavaScript VM out there, and that’s what’s (Inaudible) is trying to do.

Michael Coté: Yeah. Now, I think you are spot on with that because every time I get into discussion with developers about JavaScript, it is at that point where it’s all about benchmarking. Like I mean I guess, there’s not really that much more evolution that’s going on with JavaScript so people get obsessed with the different VMs out there. Like what’s the one Google has, it has some funny name to it like VC or I forget what it is but that’s the one that people talk about quite frequently is being like super-fast and easy to use. And it’s interesting to see what people do with that sort of proliferation of JavaScript engines out there. When they start using it more as a core runtime versus like sort of Ajax thing to do rotating crap on the web. So that should be exciting. I don’t know, for lack of a better phrase. I think someone was using the phrase crappy web apps the other day and I thought that was fantastically phrased. But anyways, I think with that, we actually covered a crap pile of stuff to use the crap word again there.

Ryan Stewart: Crap pile, well it’s almost like a month’s worth of stuff as if we haven’t done this for a long time.

Michael Coté: No, that’s crazy. It is weekly, what are you talking about?

Ryan Stewart: Exactly! We are always punctual.

Michael Coté: I think there is only a little — over at RedMonk, we found a pretty good transcription service. So I will have this episode transcribed, just to see what it ends up looking like. And if anyone out there in the listener world actually likes that, it will be great if you’d leave a comment or send me an email so I can figure out to keep doing that or not. So it would probably take a day or two to get a transcript, but obviously –.

Ryan Stewart: You should let Mike and I transcribe each other’s words so that we can put words in each other’s mouth on the text. That would be fine.

Mike Downey: Yeah, exactly.

Michael Cote’: That would be great. And obviously, if you’ve listened this far, you don’t need a transcript but it will be out there anyways. So with that, we’ll see everyone next time.

Disclosure: Microsoft and Adobe are clients. See the RedMonk client list for other clients mentioned.

Share This

I’m always searching for more advanced uses of desktop-bound RIAs, like AIR, Silverlight Out-of-Browser, and JavaFX’s out of browser options. Thus far, most examples have really just been web pages on the desktop that, at best, have off-line modes or are closed-up video players. It was nice, then, to find a use of AIR that took moved beyond these uses in Konductor’s AIR client. As Derek Zarbrook walks us through – first in an interview and then a demo – the AIR client hooks up to the Konductor hosting system to give non-designers an easy, but controlled way to update their web sites.

Interview

Demo

(As a side note, this is the first time we’ve used VideoPress to host videos. What do you think?)

Disclosure: Adobe is a client.

Share This