Liferay evolving beyond portals – the app server vacuum – Quick Analysis

A new Liferay portal releases provides a good chance to ponder the state of runtimes in the Java world.

Open source portal vendor Liferay has just released a new version, building out the platform around the popular portal it’s been putting out for years. CMS Critic has a good wrap-up of the features in the new 6EE release. As with much of the Java world now, the ideas of an application server, a portal, a container, and a general platform have run together nicely for Liferay. The same is true for many others in the field since the iron-grip of the J(2)EE went arthritic awhile ago.

Think of the bucket of everything that Spring does, the opening up of the VM to run other languages, like the darlings of last year, Scala and Erlang. JEE has turned into a buffet finicky programers can pick from, not a suite of The Frameworks Ye Shall Use. On the whole, this is good as the standard, official Java world wasn’t really keeping pace with the fast evolving ideas coming from the consumer web – from the non-enterprise world.

The Motherhood and Apple-pie

Liferay seems to have done well with it’s strategy of keeping pace of “social” and the basics of portal needs. This is probably due to being open source and, as a company and project, scrappier than it’s rival products from Oracle, IBM, and JBoss. There’s a refreshing sense of simplicity in their messaging, e.g., a 215 meg runtime instead of “boxes of DVDs.” And, of course, there’s the promise of being cheaper than Brand X, often much cheaper. Their client list – with no shortage of big names – would seem testimony to the success of that strategy.

Being here at the Symposium in Anaheim, I can see that the Liferay folks and community itself (represented in something over a 100 attendees, I believe) have a nicely evolved open source warm-and-fuzzy feeling, at the same time being unashamedly open core. They’re all nicely earnest. (And, hey, it’s always nice to be in California to get one of these.)

From speaking with the Phurnace pholks before they were acquired by BMC, there’s a brisk business going on migrating from closed-source portals (IBM’s and Oracle/BEA’s) to open source ones, JBoss, Liferay, and co. CMO Paul Hinz used the phrase “end of feature” life as a jab at incumbent portal offerings that may not be End-of-Life, but are stagnant in their development. Clearly, in the portal space, Liferay is looking to pick up migrate-to-cheap portal projects. At the same time, they’re hoping to offer a harbor for Java developers, most of which are casting about for the platform to standardize on.

Write once, configure everywhere

As James pointed out recently, Java is far from “dead.”. Indeed, Java is finding much use as the basis for new middleware:

So the cool kids aren’t using Java. Or are they? One of the hottest trends in tech right now is NoSQL (If you’re a software developer get acquainted with it). Many of the hottest NoSQL technologies are written in Java.
MapReduce – one of the core technologies Google and Yahoo use for fast response times across large data sets is Java-based. A whole new industry ecosystem is growing around Hadoop, Apache’s MapReduce implementation. Ask our client Mike Olson fromCloudera if Java is dead. What about HBase? Java… Neo4J? Java. And so on. Of course we’re also seeing innovation from the new hotness – thus Erlang underpins CouchDB and RIAK. But Java is certainly core to the innovation. Lets look at RabbitMQ for example – which though written in Erlang was acquired by SpringSource as a messaging engine to underpin a Java-based programming model.

And that’s just the category of NoSQL. Sure, there’s plenty of folks who are happy to be liberated from Java – the post Rails world blew that sentiment door off the hinges.

What’s clear, though, is that Java is being used as a base language for larger runtimes and middleware, rather than a ends-to-itself, a walled community of its own. The idea of an “application server” is being eliminated, component by component. In eliminating that idea there’s a huge vacuum when it comes to the runtime developers use to house their projects.

Liferay, like so many others, is starting to move into this vacuum. The likes of VMWare/SpringSource are betting the farm on it (and looking good), while RedHat/JBoss, IBM, and Oracle/BEA/Sun are casting about to keep up. (An interesting non-Java example is MindTouch.) It’s clear that a highly component-driven runtime is needed (OSGi seems to have won out, if by sheer exhaustion in the Java Modularization War, largely forgotten): a “stackless stack,” as us RedMonks wickedly like to repeat. Open source seems required as well, if only to drive that extreme simplicity that, for whatever reason, commercial middleware is deadly allergic to. Hint: enterprises will pay cash-money for complexity, tragically, they’ll pay little, if anything, for simplicity. And once bought in, the pace of innovation isn’t always the same as that frenzied period of finding a “solution.”

PaaS and Mobile

In the wider spectrum, the ongoing evolution of Platform-as-a-Services and mobile drives both this application server vacuum and the search for a replacement. In the PaaS world, the PaaS is your runtime: Force.com, Intuit’s Partner Platform, Azure, Google App Engine, and various other apps ecosystems. In mobile, while there are several frameworks out there, it’s too early for developers to coalesce around a short list of them: rolling your own is still popular and, let’s be frank, until Android produces an app-bubble, developing in Object-C is the top concern of mobile developers. (Object-C!)

And, to call on the distinction between ISV and corporate developers from earlier today, while ISVs as always (well, since the web, at least) can sort out their own needs, the sense I get is that corporate developers are further out to sea when it comes to PaaSes, mobile, and even the search for a runtime. For vendors, this means it’s a great chance to insert your foot into the runtime door – when things are in flux is the time to cement your future shackles of success.

In addition to the above link, some excellent, in-depth coverage from Mike Johnston at CMS Critic on today’s symposium.
The official press release from Liferay.
More videos from Liferay on 6EE.
You can see a live stream from Liferay West, today and tomorrow.
Dee-Ann LeBlanc of CMS Wire covers the Liferay release and event today.
Mitchell Pronschinske writes-up the new features with some screenshots.

Disclosure: Liferay is a client and paid my travel and hotel to here. IBM, VMWare, RedHat, SalesForce, MindTouch, and IBM are clients as well. See the RedMonk client list for related folks as well.

rev="post-5256" 3 Comments

Categories: Conferences, Java, Quick Analysis.

Tags: Liferay, portals, stackless stack

By cote

— September 8, 2010 at 4:28 pm

Know your developer type

Coming up with a characterization for a developer is often handy for figuring out how to talk with them – both about technology and in marketing. As with all characteristics, you end up with something a little more useful than a paper mache mask, but it at least gives you the baseline to start thinking about how to deal with them. When I still “did something,” as I like to say of my years being a software developer, I’d come across all sorts of folks in person and through the “Internets” and made a fun game of slotting these developers into types. Very few people every fit perfectly into one slot, of course. Below are some of the types I’ve kept coming across, patterns of being, if you will.

There are many things that push a developer to one type or another. This can largely be determined by what they read, and don’t read. To some extent, demographics (age, nationality, gender, income bracket, geography, ethnicity) can be used – but I don’t know the specifics enough to generalize. Though, I do take a certain “becoming an old guy” pleasure in trying to track based on age, rather, family status. While we’d like to think that developers are motivated by being more productive, writing bug free code, and otherwise driven by “pure” engineering principals, that motivation is nuanced. Knowing their archtype(s) helps throw up guides to bump along. Some sample archtypes:

Developers Types

Snarky – believes most people are stupid and need to be ignored or coded around. You may have optimistic snarky people who simply deal with the drudgery of programming by being crass. Sarcasm is soma.
Optimistic – often smiley, works longer hours, willing to help other people. This is extremely rare, and can be confused with an arrogant-ignorant developer who thinks everything is happy-go-lucky just to get off-shored next week.
Lots of talk, little code – spends much time on the labor and process around coding, hoping that the smaller amount of time coding is done effectively by the measure twice cut once ideas. An anecdote: a NASA software developer walks into a his managers office, saying it’s great that they finally committed so much code this week. The manager has a dark look on his face, no, he says, think of all the bugs we’re about to discover.
Craftsman – interested in being better at what they do, but often lacks the tools to measure the betterment, or even get a baseline for how bad or good things are. Instead, they try new developer self-help programs (like Agile, etc.) and are always looking for the new thing that will help them be better.
Coders – spend most of their time simply coding in, mostly isolation, generating a lot of code. Managers love these types, other developers tend to bad mouth Coders behind their back as (a.) the Coders output amount “ruins the curve,” and, (b.) their code is not always liked by the group.
I just work here – just there for a good paying job, rarely works on becoming better but is often very dependable in getting work done, if only by using out of date methods.
Legacy – “this is like CICS all over again” is the kind of sentiment you get from these folks. They continually muse that 20 years ago, they were doing the same thing, except in all upper case.
Startup-dreamer – works at a “normal” company, but is always trying to start something: a company, an open source project, a service, etc.
Process-wonk – for these folks, the problem with software is the lack of discipline and group process. They may seem overly bureaucratic, but their trust is that if the group just got their act together, the code would work better. A negative version of this is Cargo Cult, a favorite derisive-term of developers.
The Debugger – whether they like it or not, this person is good at tracking down bugs. They’re often good at doing performance tuning as well.
The Baron – a developer who likes to build his own fief, an area of the overall project that no one else understands or can work on. See DBA, network admin, and “security.”
Architects – charged with being more technical than manager, they’re expected to “lead technically” and know enough about The Business to guide technical decisions. Having been successful developers, they’re often tasked with keeping that up as well.

(There are no doubt more! Have some to leave in the comments below?)

ISV vs. Corporate Developers

Of note, the distinction between an ISV developer and a corporate developer is even more high-level than the above. Many large ISV’s (Oracle, IBM, SAP, HP, Microsoft) constituency is the corporate developer, not the ISV developer.

ISV developers are even more tricky to market to (that is, get to use your software) as they expect to use very little that isn’t open source, if any at all. The reasons are many, but include: to drive down license costs per distribution, for the malleability of the code (see how many chunks of middleware Twitter has gone through in the past 2 years), and low barriers to entry to using OSS.

Also, many ISV developers are transitioning to mobile and SaaS development which traditional vendors are not necessarily well geared for. That said, the shift to centralized, at-scale deployments (SaaS) provides interesting opportunities among ISVs for these traditional ISVs, and new ones.

(Related, you might like my 2007 talk on Developer Marketing Do’s and Dont’s.)

Disclosure: IBM, SAP, and Microsoft are clients.

rev="post-5252" 11 Comments

Categories: Marketing, Programming.

By cote

— September 8, 2010 at 10:22 am

Links for September 2nd through September 6th

Mark Hurd Joins Oracle as Co-President
Well, that's kind of a fun bucket.
Food on the Table
The State of Open Source System Automation | Linux Magazine
Pretty good, quick yet technical overview of Puppet, Chef, Cfengine, and Bcfg2.
Assistly | Help Desk Software, Customer Service Software and Customer Support
Burning question: Time to split IT?
"I think it is time to move the CIO and a few of IT staff which is business process, architecture and innovation focused to report to the CEO (or the COO), and move the infrastructure, procurement/legal, security and compliance and standards part under the corporate controller (or the CFO)."

Disclosure: see the RedMonk client list for clients mentioned.

Comments Off on Links for September 2nd through September 6th

Categories: Links.

By linkposter

— September 7, 2010 at 7:58 pm

Data integration in a cloud world with Joe Dwyer – make all #009

A cloud-based application usually has to integrate with many data sources – in this make all episode, I talk with Joe Dwyer about how his company, PropelWare architects that solution.

In this discussion we do a good job of pulling out some best practices and tips for others who’re in similar architectural circumstances.

Download the episode directly right here, subscribe to the feed in iTunes or other podcatcher to have episodes downloaded automatically, or just click play below to listen to it right here:

Show Notes

PropelWare’s solutions integrate with many accounting services, PaaS’s, SaaS’s, and other “cloud-based” services. One of their primary integrations is with Intuit’s Partner Platform, giving them access to QuickBooks.
We talk about how it fits in and what services are used, etc.
I ask about the rise in mobile usage and what it ends up looking for their customers.
I test one of my pet theories: does simple API mean interop matters less?
Is there a common models between these accounting things?
Feelings about switch to SaaS usage among customers? Are they afraid, happy – do they even notice?
Azure usage for Office Manager.
Their experience in the mobile world: what are some lessons learned.

Transcript

Michael Coté: Well, hello everybody! This is another edition of make all the, podcast about fun and interesting stuff in the software development world, and just kind of the wacky world of computers as it were. This is your host, Michael Coté, available at People Over Process as always.

We’re joined this time, again, by someone else from the Intuit sort of ecosystem or community if you will, which, it’s always a fun story to see how people are extending that big hunk of data and process out there. So with that, do you want to give an introduction to yourself?

Joe Dwyer: Yeah, this is Joe Dwyer from Propelware.

Michael Coté: So your website as your name implies is Propelware.com. You have a pretty straightforward way of putting out the different products that you have on the website there, and I personally like that they end with the cutsey-like “f-y” end phrase there. I’m always looking for fun names like that, but could you give us like a quick overview of the sort of three products that you guys are working on?

Joe Dwyer: Yeah, so the product that we have — Autofy is our main product and that stems from the combination of words, Automatic and Simplify. So what we do is we try to automate processes that small businesses use when they run QuickBooks. So, with Autofy, you can do things like automatically bring across eCommerce transactions from your shopping cart, also submitting the data entry that goes along with that, a process.

And things like Synchronize, other systems that you undoubtedly use with QuickBooks, which are CRM systems, budget management systems and other applications you may have, crated or used along the way, we make it easy to integrate that data back and forth.

Our other products, MobileBiz is QuickBooks for mobile phones. I’d say it’s a little bit more of a light-hearted application because of the mobile aspect of it, and things like that, but essentially you can enable a few of the basic workflows of QuickBooks, onto your mobile phone, for just creating invoices and things like that, and synchronize those up into our system, and then across into QuickBooks.

Michael Coté: So, I mean, by light-hearted, do you mean that every now and then, it sort of slips in a joke or something, or what’s the light-hearted aspect of it?

Joe Dwyer: The light-hearted of it is, it’s a little bit more trendy application, we don’t typically roll it out to hundreds of users at a time, but people who have – they have real small businesses within QuickBooks, they’ve got either their sole proprietor or they’ve got five or ten employees, smaller companies that want to push some of their QuickBooks data entry to the front-office, to their sales staff, and Intuit is just a lightweight application. So, that’s what we look at, as a bit more light-hearted, it’s just light-weight, it’s not the full-blown QuickBooks or anything like that.

Michael Coté: Yeah — no, I like that disclaimer. I mean, I should start using that, because it’s — like you were saying, you could say light-weight, which is descriptive and everything, but it’s a little more fun to say light-hearted.

Joe Dwyer: Yeah, exactly, it is.

Michael Coté: It kind of humanizes it.

Joe Dwyer: That’s right.

Michael Coté: I’ve been fascinated to kind of see Intuit kind of transform QuickBooks too, what I would think of call a very traditional sort of desktop-bound application or something that is — yeah, as we would call it, now, sort of cloudified or whatever. I’m curious — I mean, how long have you been working with, integrating with QuickBooks?

Joe Dwyer: Yeah. So I’m a software developer by background. I started the company in 2000. So we’ve been in business in 2000. We’ve been very, very focused on that QuickBooks market since 2002, which was, literally version 1.0 up there SDK. We’ve been working in the QuickBooks integration space since 2002 and that spans their lifetime of both their SDK based system for outside developers and also with the Intuit Partner Platform that they launched a couple of years ago.

Michael Coté: So, you’re in a kind of a — I guess, that’s about ten years, essentially. It’s an interesting position to have seen the — I don’t know you would call, but the evolution from sort of native SDK integration to vary web service, or it’s services based, whether they were web services or other thing’s integration. Can you kind of speak to like what you’ve seen in that evolution from doing sort of binary desktop things, to more service-oriented things?

Joe Dwyer: Yeah, Intuit has been a — they’ve been a great partner for us to work with in their developer relations team and the Intuit Partner Platform team and those folks, and it’s really out of that team that built that original SDK. The core of that team also now provides the Intuit Partner Platform. It’s been great to work with, and we look at them in some ways they are two different things to work with.

One is, obviously, much more desktop-centric, it’s COM-based, things like that, very much meant for data to move in and out of QuickBooks on the desktop, whereas Intuit Partner Platform, where it becomes more cloud-based, you’re still at the end of the day interacting with QuickBooks data, which lives on the desktop, what you’re doing is more through web services, and from the web, from the browser-based applications.

So they are very different, and the look and feel of them, in the way you interact with them as a developer is very different with authentication and all those sorts of things that go along with taking data like this to the web, but the people at Intuit are very much the same people, the same community and things like that, which provide those types of tools. So it’s been a nice —

Michael Coté: It sounds like it really is sort of just — programmers are always trying to reduce things down to it, it’s just a tool that does the same thing, right. It sounds like to some sense, it’s sort of the same backend, no matter what, and it’s really just different tools you’re using to access this data and workflow, whether it’s the native stuff or the IPP stuff.

Joe Dwyer: Yeah. With IPP, it is a little bit different, because they also wrapped a lot of authentication requirements onto this system and in tools that go along with that, whereas when you’re running a piece of software on the desktop, well, the end-user installs that software, when you are running it on the web, you need to make sure, the people that are logging into the system are authenticated properly, so there is a heavy dose of authentication.

Even within their web services, they’ve also tried to take it to the next level of building out a more complete system of it, more of a generic system, and things like that. So, the XML formats and the web services that you integrate with are a bit deeper, they’re just more there, more ramp up time, to get comfortable with and understand how they work and what they do and such.

Michael Coté: So, in this — I hesitate to call it new because it’s been around for a while, but relative to SDK-based ways of doing things, in kind of the new world of doing integration over like web services, like particularly looking at like your Autofy custom stuff, you guys also integrate with like Salesforce and Peachtree and other stuff.

I wonder, is there a lot of commonality with those different things, or when you go to a different web services thing, is it sort of like a brand-new thing you have to learn, or is that more reusable?

Joe Dwyer: So, within the Autofy, and what we do with our Autofy custom system, you’re right, we integrate with QuickBooks, but also with Peachtree and [Microsoft] Great Plains and Salesforce, and even PayPal and eBay and some of these guys. Essentially, what we have done is we’ve created our own set of infrastructure, where we can move data back and forth between various systems.

When you look at integrating data directly with QuickBooks to either SDK or versus IPP, they are very different, they are different in a way in which we interact with them. But the types of data that people are moving in those systems that we work with within Autofy, an invoice is still just an invoice, a customer is still just a customer. They all have similar attributes.

So that’s where the power of Autofy comes into play is, we can take and create an invoice into Peachtree or into QuickBooks, via the IPP or via the desktop, or to Great Plains, because an invoice is just an invoice at the end of the day and so our system allows us to do that pretty well.

Michael Coté: I mean, as far as kind of giving advice to other people in the situation, and I would imagine this, but correct me if I’m wrong that you guys have sort of developed your own internal model of an invoice, and you’re sort of mapping the various different chunks of data, the various model from different systems to your intermediate model and then kind of transforming it back to another one or something like that.

Joe Dwyer: Yeah, you’re dead-on. That’s exactly what we’ve done, and ultimately, then our system becomes sort of a superset of the various fields that are contained within the various accounting systems.

So that’s exactly what we’ve done, and then the other part of it that we have done also is, to dynamically be able to switch back and forth between retrieving data from one location, entering into the other location. So for example, we can have a single application that will connect into QuickBooks or Peachtree live, without having to reinstall new pieces and things like that.

Michael Coté: Is that not knowing the — I don’t know, the small and medium accounting kind of world, perfectly, I mean, are there sort of industry standards for those kind of models, or is that, in other areas of development, it is always an attempt to have a standard around various models, but when it comes to accounting and invoices and stuff, I’m not sure if there is popular standards that are used.

Joe Dwyer: There really is no — there is no clear wining standard. There are industry groups out there that have tried to push through XML format and things like that for data interchange, but oftentimes the individual accounting packages don’t really reach to those. They all have their own ways of doing things and that sort of thing. So they don’t accept data that’s in those standard formats.

Of course, there is the — it’s similar to like the EDI world where you’ve got standard integration format for transferring data to relocate on locations, but when it gets down into the financial system, they end up being employed a bit different.

Michael Coté: Yeah, I mean it sounds like — obviously, to put it one way the container for the data is standardized or the protocol and all that. But sort of semantics or the meaning of the data isn’t always as standard and interoperable and everything like that.

Joe Dwyer: Yeah, that’s exactly right. And also, even the rules for entering data into the various systems become quite a bit different. For example, in QuickBooks, we have a pretty large infrastructure around determining is the data that one software or product is giving us, is that compatible with what we can get into QuickBooks?

So in an invoice that’s the customer exist, meaning name, address, phone number, all that sort of stuff, well, our Autofy system will automatically enter that.

Are there any other errors that we have to deal with, are the sales tax, so all of those things had to go along with just getting that single transaction into QuickBooks.

Michael Coté: Alright, so, when you are dealing with that with IPP, what’s the kind of support it gives you for that kind of verification in checking things out? That kind of data DMZ that you need to make sure things are in the right format to work with IPP?

Joe Dwyer: So with IPP what we are doing is with Autofy we’ve built that out, deep and wide with the QuickBooks traditional SDK. We are building that similar capabilities out with Intuit Partner Platform as it’s becoming more mature and things like that.

Also, the models always become a little bit different, because instead of Autofy being installed on your desktop, it’s now installed in a browser essentially. It essentially runs in a browser.

So moving data into QuickBooks via the IPP is really neat in that, it gives us these capabilities of creating invoices, all that sort of stuff, but you still have those same rules that need to be followed.

You can’t create an invoice unless you first set up the customer information and make sure that the features are done in the right order, that they are validated and all those sorts of things that go along with a standard invoicing or accounting model.

Michael Coté: Right, that makes sense. You mentioned something else that I was curious to hear about, especially with the kind of customer base that imagine you have and that’s that — with a lot of this stuff including the Intuit stuff and other things you integrate with, it’s more running out of the browser versus running out of a desktop application. Or with your light-hearted application, running off of something you are holding in your hand if you will.

So there is this kind of – the screen changes around a lot. And I wonder if — since 2000 like, how have users felt about that changing of the UI now-a-days? Have they even noticed or are they happy about it or is it a little bit of friction for them? How does that kind of play out?

Joe Dwyer: For us we ultimately have two different types of customers. One is the end-user who is using QuickBooks and oftentimes for those folks simpler is better. So the more we can set up Autofy and MobileBiz and our other software to be one-click friendly where it can be simple to set up, simple to use, automatic, then UI becomes secondary. You don’t need the nice Web 2.0ish sort of interaction and things like that.

The other core customers that we work with are software companies where they have a product and they wanted to integrate with QuickBooks or another accounting package and whether that’s through the SDK or the IPP, they are a software company and they want their software to integrate with those systems.

For them the UI becomes more important. You want it to look pretty, you want it to look nice, but at the end of the day they are going to deliver it also to the end customer who wants a simple one-click solution.

So they are a little bit different, where we need to spend a bit more time when we look at the user interfaces, make sure they are modern, make sure they are nice, but also make sure they are simple with software companies.

Michael Coté: I mean, it’s interesting because there is always this — I feel like there is always this tension with user experience and software where — obviously, to be tautological about it, you don’t want it to be terrible. You don’t want it to be complicated, and you also don’t want the user interface to be kind of baroque. Like I always call it like the Winamp user experience thing, right? I mean, if you remember Winamp, it was —

Joe Dwyer: Yes, that’s right.

Michael Coté: It’s one of these things you could add different skins to, and you could really just go crazy with the user interface to where you could, didn’t really even know how to use it, but it was very beautiful or pretty looking.

It’s interesting in the business context. It always seems like you were saying, the thing that trumps anything else is just simplicity, like almost — versus consumer web experience, or consumer software experience where the consumer wants, you want the user to be with the software as much as possible. It’s kind of, whether they are playing a game or they are looking up on Facebook or whatever, it’s kind of like you want them to be engaged with the software, but oftentimes, it’s not the majority of the time with business software; the longer they are using the business software, the poorer the user experience is.

Joe Dwyer: That’s exactly right, and that’s also where you look at QuickBooks, has been around for 15 years or more. So, people have grown with that application, and they use it and they use it and they use it day-in and day-out, whereas, with the Intuit Partner Platform, what’s being enabled is, applications that other people are using, or the enablement of different workflows, that can help people.

In QuickBooks, they need to solve problems for the four million companies that they serve; whereas a company like ours, we can roll out a new application with the new workflows around, the data that we integrate, and make it look and feel nice for customers and feel a unique niche.

One example of that is, we’re in the process right now of rolling a beta version of a new application called Propelware Office Manager, fully runs on the Intuit Partner Platform, also runs on the Windows Azure system and SQL Azure.

So this doesn’t have anything to be installed on the desktop computer, but it allows us to enable certain workflows that can be made simple and easier for those QuickBooks users, but also give them the nice Web 2.0ish sort of — and even just the overall web experience, being able to update your software on the fly, and all those sorts of things.

Michael Coté: Let’s peek into that Azure thing a little bit, like what made you select to deploy on top of that? I think most listeners will know what Azure is, but it’s basically what you would call Microsoft’s Platform-as-a-Service. I always think of it, my simplified way of thinking of what a PaaS or a platform is, is basically the middleware part of cloud computing, and that sort of provides you a place to run your software.

Joe Dwyer: Yeah, so with our Office Manager application, and work that we’ve been doing with IPP, we had started hosting it. We were on test environments and things like that here, on our local servers, but we needed a hosting platform that was obviously off-site, we don’t want to run it in our offices. And for us, we’re a software company, we’re not a hosting company, we’re not an IT infrastructure company, anything like that, so we needed good quality, 24×7 hosting that could scale for us, as we scaled our business application.

So we’ve done some of that hosting through Amazon EC2 system, but with Azure, when Azure came about and started becoming a bit more ready for primetime, we jumped over into that, and for us the big thing is that, we can run basically the hosting services, so basically, we’re running IIS and we’re running SQL Azure, all in a single place.

At the same time, we don’t have to worry about configuring IIS, or FTP, and worry about other ports that might be open in systems. We can put all of that into a SAS 70 environment, which Microsoft provides. We can also sell that to our customers by saying, we are taking security and hosting very seriously. So it’s through the Azure system, it’s not through — it’s obviously over and above what we could ever provide as a small business.

(00:19:54)

Michael Coté: Yeah, yeah, I have some friends who are starting up a SaaS-hosted thing themselves and they were kind of struggling with that kind of balance of efficiencies of running something on your own versus sort of the premium of giving your customers the warm and comfy that everything is very secure and compliant with everything.

So as a technologist it can be kind of a frustrating thing to deal with because they are sort of non-functional requirements if you will, but I don’t know. I mean, that’s kind of, one of the promises of cloud computing is that it’s not really — you don’t have to worry about these things that previously would have been so frustrating.

Joe Dwyer: Yeah, that’s right, and along the security road customers still ask about it, is my data secure? We’re running, this is financial data that we’re interacting with, and not credit card members, but still it’s a small business financial. And there is no way to show anyone the lockbox where their data is, it is sort of this abstract thing.

So to be able to explain security in a way where you say these are the systems that we provide are on Microsoft servers, people can connect with that very quickly.

Michael Coté: And now if I recall correctly Intuit, or IPP and Azure have sort of a tooling partnership at least with each other, are you guys using that?

Joe Dwyer: We are, and what we’ve done is — so we are using the toolkits that they’ve provided to connect IPP and Azure a bit more easily. What we’ve done is we’ve taken those tools, and Intuit has open-sourced some of those.

What we have done is we have taken them and tailored them for our own purposes. We have actually put, what I would say is sort of a next version around them, and we have made them a bit more generic for our own purposes, put a bit more power behind them. And the reason for that is, because part of what we are seeing some momentum around is our Autofy system running in an online world.

So we have clients right now and we have other prospects that are coming to us and saying, we have a web-based application, we wanted to connect into the Intuit Partner Platform, can you help us? And so we’ve taken their toolkits and are now basically enabling us to make it easier and more comfortible to get other companies connected into the Intuit Partner Platform via our tools running on Azure and things like that.

So in the coming months we will have other companies also running on Azure and the IPP where we are basically doing the development for them.

Michael Coté: Oh, right, so I mean, basically you are kind of — you are not sort of a cul-de-sac of the community if you will. I mean there are people that you are working with yourself, that are kind of building on top of what you’ve been doing.

Joe Dwyer: Exactly, exactly, and that’s part of our practice, that’s part of our Autofy system where if a software company comes with and they have some sort of a client server application, we can enable Autofy to get their data connected into QuickBooks, if they are a web-based system, and they wanted to take advantage of the Intuit Partner Platform, we can do that for them so they don’t have to go through the learning curve of the security model that IPP has and the web services they have and the QuickBooks or consolidation, all those sorts of things. We can take our tool, plug them in for them and get them connected up, typically faster and less expensive. Then they might be able to do it themselves.

Michael Coté: Yeah, so the last thing I was interested in talking with you about and we made reference to the light-hearted MobileBiz several times, or at least I have, and I am curious like what — this is kind of like at this point with the huge interest in mobile, it’s kind of a rhetorical question if you will. But I mean what, in your specific customer-base, like what was pushing you to do mobile stuff essentially, like what kind of finally pushed you over the age where you thought like, alright, we are going to do this?

Joe Dwyer: For me it happened way before any of the iPhone or anything like that. For me MobileBiz was the first place where we started working in the QuickBooks integration space. When we found out that QuickBooks was opening up in API, my first reaction was great, I can finally put QuickBooks on my PDA, and so that was — the cool, neat thing for me was to be able to create invoices wherever I am, from wherever I am, expenses, track expenses from wherever I am and through that into the PDA and have it go into QuickBooks.

So that was on old school PDA, and so over time, that there is just a momentum around that, but I have sort of — a lot of that has come out of my own desire to run the business in that way, and then of course now over time to sharing from customers and what they need and what they look for and things like that.

Joe Dwyer: Our mobile application — what we often hear is, we charge $7 a month for it. So it’s not one of these free applications that you download to your device, and you’re off and running and traveling with it. It really is an application that people look at us and say, this is good value for them when they want to bring their phone with them.

Michael Coté: And have you been experiencing, for lack of a better phrase, a kind of mobile bubble if you will? Whether there is a lot of interest and people do — a lot of interest in what you are doing back in the PDA days where people really want to move what they are doing with their software into their palm so to speak onto their phone versus on their desktop?

Joe Dwyer: Yeah, certainly over the last couple of years, especially a couple of years ago the world really changed where they don’t want it on their PDA, they want it on their phone. People are like carrying around one device. So certainly there was a mind-shift change there, but also just the volume. The number of people that we interact with, with that product has changed. Lots of people want it.

So we look at having to manage the various platforms of iPhone, Android, Windows Mobile, things like that. We are working as the new Windows Phone 7 for a launch of that platform, so that our MobileBiz system will be ready when it is launched as well.

So there is definitely a lot of momentum in that mobile space, but for us we very much — we are not a video game or anything like that or a 3D application.

Michael Coté: And I wonder in doing both PDA and sort of post-iPhone era mobile development: in the same way that we were kind of talking about the user experience with business software, sort of like make it as simple as possible so people don’t have to actually use it very much.

I wonder if there are best practices like that you discovered for business mobile applications like what are the ways you make the user happy with a software, when they are not like playing a game or whatever?

Joe Dwyer: Exactly, well, the funny thing is if you go back and you look at our original version of MobileBiz, the way that it worked was you could create an invoice on your PDA, but in order for you get that into QuickBooks you had to take and plug your PDA into your PC where QuickBooks is running. And then it will move the data back and forth. There was no such thing really as a data plan or as WiFi on my PDA, that stuff just wasn’t there.

So for us that transition happened — at some point we had to say, okay, we need to make this disconnect. We need to allow people to synchronize their data when they are not plugging it into their computer. It needs to do through the clouds or for the web, that sort of thing.

So for us, and I think for our users that really is a neat experience in and of itself where they can be completely disconnected, look at some of these transactions, look up a customer, create an invoice for them, but then send it back to home-based even when they are out and about. They don’t need to be connected to QuickBooks in order to get that data back into QuickBooks.

So that’s been a big thing, neither user experience, aspects of the application. Of course, it’s a small form factor and things like that, so those things are always important.

Michael Coté: I think you are right there. When I am using productivity apps to use another word for this kind of app or whatever, the thing that I will notice that’s missing is that it’s not really connected to the Internet, or it’s not really connected wirelessly if you will.

When you are doing that in the mobile world, I would imagine in most cases that means you guys have to host a server or something somewhere. How do you take care of that end point that you are sinking through, I guess?

Joe Dwyer: Yup, and so for us that essentially what we have is we have the mobile application. So the app lids on the mobile device we run a SQL Lite database on the device when it comes to do the data synchronization, you’re right, we send that up to our web-based platform. So that is now running on Azure.

So essentially what we have on Azure is a web service that we can connect from our mobile application into our web application and then from there we can sink the data into QuickBooks via the Intuit Partner Platform.

So it’s been a great — it’s been a nice standard approach way where we don’t have to jump through hoops to make all of that data move back and forth using our Autofy system and our expert piece in synchronizing data, we are able to pull that off pretty smoothly and without a million different requirements that are all duck-taped together but rather a pretty nice architecture overall.

Michael Coté: Yeah, I mean you guys — it seems like you guys have basically built a data hub, if you will, with many nodes, and endpoints, and pipes. As with the perfect idealized architecture of that, like once you sort of establish the baseline for really you just hook up other nodes to it and it doesn’t necessarily matter if they are on the cloud or a desktop or a phone or whatever. It just matters if there is a network connection that they have.

Joe Dwyer: Yup, you got it dead-on, and that’s whether it’s a mobile device or it’s someone’s one-off access database they created or another software system that someone has built, that’s for Autofy and where all of our tools fit in really well to make those guys connected a little bit easier.

Michael Coté: Well, great! I mean that’s — it’s always nice to like encounter someone who is doing kind of fun boxes and arrows stuff like that. The full data integration hub stuff and kind of just lightly pick apart the various way things that you guys are doing. So I appreciate you going over all that stuff, that was fun.

Joe Dwyer: Yeah, no problem. I appreciate! Thank you very much, Michael!

Disclosure: this episode is sponsored by Intuit, who is a client.

rev="post-5248" 1 Comment

Categories: Cloud, make all, Programming.

By cote

— September 7, 2010 at 1:51 pm

VMworld 2010 – IT Management & Cloud Podcast #79

John and I give our take on VMWare’s cloud and application development announcements at VMware. We also discuss private cloud and what the announcements mean for other vendors, if anything.

Download the episode directly right here, subscribe to the feed in iTunes or other podcatcher to have episodes downloaded automatically, or just click play below to listen to it right here:

I’ll have a write-up sometime soon. In the meantime, we say plenty in the recording.

Transcript

As usual with these un-sponsored episodes, I haven’t spent time to clean up the transcript. If you see us saying something crazy, check the original audio first. There are time-codes where there were transcription problems.

Michael Coté: So here we are at San Francisco, at the Moscone Center, VMworld. It’s the IT Management Cloud Podcast, Episode 70 something, probably the late 70s, I don’t know. We are starting to collect social security, getting some healthcare from the government, it’s great. You know in our 70s.

John M. Willis: Yes, I guess nostalgically.

Michael Coté: As always, this is Michael Coté, available at peopleoverprocess.com and you are?

John M. Willis: Live, in-person, John M. Willis at johnmwillis.com.

Michael Coté: You can hear the background of the city here. We have got the Apple Event going across the way, which I understand there is iPad Supernano or something. I don’t really know what’s going on there.

John M. Willis: It’s like a Cray-based process or something like that.

Michael Coté: Yeah. So a lot of the discussion here has been about cloud stuff, and so I sort of think as a cloud vendor, there is sort of this first phase of cloud adoption that you go through, and you have this epiphany of like what a cloud is and why you would use it and all this technology stuff, and do you think — and then you go out and you start telling people about it and you are like, oh, we got this stuff figured out! We know what this cloud business is! It’s like when people figured out open source. Do you think VMware has moved beyond that first stage?

John M. Willis: Yes, unquestionably. I mean, we can be cynical about VMware in some ways. But I think this year, I mean, I have got to be clearly honest, maybe I am in the conference hype of, when you are at a conference, you think —

Michael Coté: Sure. Demo Glow they call it.

John M. Willis: Yeah, right, okay, Demo Blow, good.

Michael Coté: Glow, but Demo Blow is good.

John M. Willis: Oh, that’s what I am saying, yeah.

Michael Coté: That’s kind of what happened in the 70s, in the 80s.

John M. Willis: Not that I would have any clue of what that would be all about, but anyway, it’s a good story – I mean the vCloud, the vCloud API 1.0, which is what they announced, and the vCloud Director, those two things, I think they have up the — I have been saying this, like there is this like — there’s two ends of the cloud; there is that commodity open source grassroots app, that’s coming out of — like started with Amazon, Eucalyptus, now OpenStack, and they are kind of moving kind of up. And then you had the kind of enterprise-ish guys like slowly, like stealing some of the fanfare, we are a cloud but you are really not. But I think that VMware has moved the bar closer to the middle. Does that make sense, from the top-down.

Michael Coté: Yeah, yeah.

John M. Willis: And so I have to give them their due, that I think they have moved the bar closer to where the guys on the low-end are moving the bar.

Michael Coté: Sort of the middle.

John M. Willis: The middle, right, yeah, which is — but the middle is the kind of where they both meet and we are solving both, the kind of web operations and the enterprise problems with the same solution.

Michael Coté: Yeah. I mean, to that end of things, I have been curious about is all the like sinewy, glue stuff everywhere. There has been the phrase around like policy-driven automation, and there is also self-service things, and like all that stuff that kind of like — like you are saying, on the infrastructure, on the middle end, they are sort of like, all the virtualization stuff that kind of makes this happen, and then on the high-end, there’s sort of like all the process that it takes to like execute all of this.

So like a ticket gets opened, I need to provision something, right? And then there is everything that happens after that ticket is opened. And there is always a lot of speaking to what that stuff is, and I don’t really know quite — maybe I am making too big of a deal about it.

John M. Willis: I think that it’s funny, because I was just thinking while you were talking about that, it’s like, what are the common themes that you hear from both worlds? So if you go the Rackspace, you hear them talk about OpenStack or you hear Ubuntu talk about cloud or even Red Hat, and then what are the common things that you hear between like the VMware and like the HPs and the IBMs.

So today, we were talking before this, I had got to go to HP Boot Camp, and so one of the common things, of course, what everybody says is like self-service, pay as you go.

Michael Coté: Metering.

John M. Willis: Right, metering, compute power on demand, I mean you are starting to hear the self-service, frictionless IT from both ends. I think when I say the DevOps movement is — part of that is to try to get in a place where operations provides self-service operations for developers. You are even starting to hear hints of that from both sides.

But the thing that I am hearing that’s positive from VMware, and to a certain extent HP and I think IBM is like starting to integrate some core enterprise concepts, and they are not saying it, alright, they are being real careful to bury that in like the sub-bullet of the sub-bullet

Michael Coté: No. And in fact, in the financial analyst, industry analyst meeting yesterday, there was a couple of throwing around the phrases like “post-ITIL” and things like that.

John M. Willis: But what they are saying, which I think is really important, which is really a process that is core, is “service catalog.” So that seems to be a theme that the enterprise is moving and putting on the table, right, okay, guys, it’s our time to ante up some. We stole a bunch of your stuff, we are saying your speak, like you guys kind of spoke self-service first. You guys spoke pay as you go first. We got kind of closer to that, and in some cases we are there.

Michael Coté: Right, right, right.

John M. Willis: But we are going to throw something on the table now, and it’s called catalog, and I think that’s one of these things, like both sides — I say there is this convergence, where the WebOps meets enterprise, and I think that that’s the part that the enterprise offers, because that’s solid stuff in the cloud. You build a catalog and now your delivery becomes — it’s like, you saw the keynote yesterday, right?

Michael Coté: Yeah, yeah, yeah.

John M. Willis: Yeah. Like they talked about like your VDC clouds, your gold, your platinum, your silver, all driven by service catalog. So the user has no clue, other than I want a service, and policy-driven automation has decided, what cloud and what level of service you are going to get based on who you are in the service catalog and what you are asking for.

Michael Coté: Yeah. I mean, I guess that is the aspect I am kind of most interested in is the automation aspect of everything, because that seems to be — in the past, when we tried to have Run Book Automation in catalogs, it seems like it all broke down because you couldn’t automate it. And so the question is like, why is it going to work this time? And you know obviously that —

John M. Willis: For clouds my friend. Why are you here? Yeah, right. It’s the technology. IBM talked about autonomics 15 years ago, right? And it was a great idea 15 years ago, but the technology to deliver that concept was like still 15 years away.

Michael Coté: And in fact, I went to go look up “policy-driven automation” on Google – and was I telling you this [before recording] – and there is actually a patent that IBM people filed in 2004, 2006 for that. But it’s funny, you read it, because it’s like — it starts out great and you are like, is a patent office guy really going to read this? It says, “in an autonomic system,” and you are like “oh, I am sure the patent guy is like, ‘is that a word?'”

John M. Willis: Yeah, yeah, yeah, it’s like, I can’t find it, I will just leave that one. That’s probably the guy like looking, is anybody looking, I will just do it. I am not going to ask.

But yeah, I think so. I think that, again, going back to the theme that, like the enterprise is anteing up and VMware is like putting some stuff on the table in reality. I think when they talk about kind of vCloud, I mean, if you look at the architecture of the vCloud, it talks about organizations and virtual data centers with allocation from a service catalog. You know what I mean?

Michael Coté: Yeah.

John M. Willis: And so —

Michael Coté: Which is all like extremely service delivery, ITIL stuff.

John M. Willis: Right, right, that’s right.

Michael Coté: I mean, that’s pretty much it, right?

John M. Willis: That’s the stuff that needs to happen to WebOps, right?

Michael Coté: Yeah.

John M. Willis: For everybody — WebOps are enterprise. Like what’s Zynga, right? They probably got on like 100,000 servers, right? Are they an enterprise?

Michael Coté: Yeah, yeah.

John M. Willis: Probably so, right?

Michael Coté: Yeah, I would be curious to see how they run their — like the process they use to run all of their stuff, because it’s probably —

John M. Willis: Well, I am sure it’s so grassroots, because it grows so fast, but the point is they are at a point where they have to like — even if they are still using commodity stuff like open source, cheap as I can get, to run fast, build fast, they still have to start or already are paying attention to enterprise-ish policies on stuff.

Michael Coté: Right, right, right.

John M. Willis: Because you can’t run an upscale operational with things not like change management and incident management and all the things that are in a service catalog.

Michael Coté: And speaking of automation and going back to the keynote a little bit, when I was flying out here from Austin, I was starting to think that — like I am doing some work with some automation people at CA now and I am going to — a little advertisement, I am going to do a little webinar.

John M. Willis: Is that right, they have automation people over there?

Michael Coté: Yeah, yeah, but there will be a little webinar sometime with them. But I was thinking like, present company excluded automation in general as kind of like not too exciting, unless you are building it from the ground-up and doing something interesting. I mean, at the end of the day, you are still automating configuration and provisioning and everything.

So I was thinking like, what would the future functionality of automation be? And like they had this project horizon they were mentioning, that was basically like, if I remember, it was like doing SSO between your desktop and SaaS-base stuff you are using. And it does seem — at some point —

John M. Willis: Don’t call VDI on me here.

Michael Coté: No, no, don’t worry. I was kind of thinking like, it seems like part of what automation will need to do is basically configure the SaaS services they are using, in the same way that like you would automate servers and automate other stuff, but at some point like there will have to be some chef recipe to provision a Salesforce account or something like that.

John M. Willis: So my favorite phrase this week has been, from your lips to God’s ear, because I have heard that a couple of times this week from a few people. Yeah. I mean, so — let me say –yeah, I think there is a thing, the vFabric stuff, that’s really important there. I mean, let’s circle back for the purpose of our listeners, let’s review what is significant about vCloud. I would like to hear your opinion. I have kind of opinions on why vCloud is — if you look at the Paul Maritz’s presentation, he defined three things that are significant emerging technologies or things that VMware has to face, right?

(00:10:15)

Michael Coté: Right.

John M. Willis: And he talked about infrastructure.

Michael Coté: Wasn’t it Twitter, Facebook, and Virtualization?

John M. Willis: Yeah.

Michael Coté: No, that’s the —

John M. Willis: No, that’s right. No, he talked about like infrastructure, we have got to deal that infrastructure is like changing.

Michael Coté: And then there are legacy apps and new apps or there are so many —

John M. Willis: He actually did a good job. They nail it down to three different things. He said, there was infrastructure, there was these emergent frameworks, and then there is the whole, your VDI stuff, which is the emerging technologies and the end user. So I don’t give a rip about that last one. You can do a podcast with somebody else. Maybe someday I will care about it.

Michael Coté: But it’s funny. I mean, being parenthetical, VDI is one of those issues where like everyone, but vendors, and a few people are just like, it’s just —

John M. Willis: I mean, we can spend the next hour like both agreeing why in the hell everybody hasn’t just adopted without — don’t — Pasco, take your $200.

Michael Coté: It’s like the electric cars.

John M. Willis: Yeah, really. At the infrastructure level — so I think you can break down the discussion if you followed those three themes that they were trying to build is, the infrastructure level, the answer, yes, it was vCloud and vCloud Director. At the framework, the emerging framework, it was vFabric. So I think from the vCloud perspective, my perspective, and I definitely want to hear your perspective, is that, what they have done significantly, vCloud, is the — they have broken down cloud to three different components; compute, storage, and network. So that’s apple pie, right?

Michael Coté: And they threw in security there, which is —

John M. Willis: And security, the edge case.

Michael Coté: Which I thought was great, because the way they illustrated security in the Burger Stack was they drew a red line around that.

John M. Willis: Right, right, which is good. And security is important too. But one of the things that I think they did — so like compute for everybody is commodity, whether you are talking about Amazon, whether you are talking about IBM, or anybody who is trying to do a cloud today, the compute is not the differentiator.

So what VMware to me has done is, which is significant, is they have upped the ante on the network as a first class citizen, and they have upped the ante on storage as a first class citizen, although some clouds treat it better than others.

But like those two, as part of their cloud story, are now treated all kind of at the same level. I think to me that’s the first significant thing about vCloud is that, like they put an enterprise perspective on it. Like, God, how could you like ignore the importance of the network?

Michael Coté: I mean, part of what you are saying is, they are making hardware insignificant?

John M. Willis: I think we are all trying to do that, but if you, like for example, don’t really address network as a first class citizen like you do as compute power, then you are missing something huge. I mean, that’s obvious, but not obvious to all, right?

Michael Coté: Yeah. I mean, I think the network guys still kind of have their kingdom going on, but there are kind of cracks starting to form in it. The only kingdom people will be left will be the DBAs running around.

John M. Willis: Yeah, but they will figure out those guys too, how to nail them. Yeah. I mean, so the whole vApp thing is about those boundaries of — I guess that’s more security, but the boundaries of the network that you can isolate by vApp, you can isolate by VDC, Virtual Data Center, or you can — boundary by the whole —

Michael Coté: That’s the way like Cisco and other people wangle their way into this whole thing is allowing to — as you guys and other DevOps people would say, to treat infrastructure as code essentially, to kind of make networking gear as malleable as software is, which currently it’s not.

John M. Willis: And treat the whole infrastructure, and this would be the whole Chris Hoff and why DevOps is broken argument, is that we don’t like talk about the network, we always talk about like kind of the OS infrastructure’s code. But the truth is, we don’t really see a viable way to automate all of it, but like what vCloud does is like kind of gives us a way, because — again, upping the ante, making network and security and storage, and compute, all for –.

I think one last point, then I want to hear your opinion, is the idea of the vCloud and the providers including that together. I think vCloud Express was an immature —

Michael Coté: Yeah, yeah. That was actually something I was going to say is, is I have noticed in the past six months that service providers or MSPs are suddenly like, this is yet another thing that was like boring and no one cared about, suddenly they are a hot commodity for vendors.

(00:14:59)

John M. Willis: They are a hot commodity, right? And that’s a perfect way to like try to consume this cloud story. In other words, find yourself a TeraMach, a BlueLock and like –

Michael Coté: Or Verizon.

John M. Willis: Get Verizon, and get them to buy in to the vCloud API.

Michael Coté: Yeah. Well, I mean, what I would say to be a bit 15:14 is, it’s a legitimate way to sell private cloud, because you are basically selling the tools to run a public cloud, where it’s like — I mean — long time listeners and yourself will know I waffle walk back and forth on this, but I am at the — whatever the opposite of an apex is, where I am thinking like private cloud is just a bunch of crap, not that it’s a crap, but it’s a distraction, right?

John M. Willis: Right.

Michael Coté: And it is like, it’s just frustrating to like think of, there is so many legitimate reasons to like have your own infrastructure and do all this stuff, and yet it’s kind of like, well, but then you have to manage it all.

John M. Willis: Right, right, that’s the whole like why a private cloud doesn’t exist story. But the truth of the matter is, it’s like everything else, there is never a one answer for all those things. Like these questions are never binary. You are like, public cloud, alright, it’s all over, we will send it, there will be no more discussions. I mean, truth of the matter is, like customers have to have the capability to control certain parts of the infrastructure in the ways that they find unique.

Michael Coté: Sure. Absolutely!

John M. Willis: And that’s why it would be interesting to say that, like everything is going to be a public cloud, I think is somewhat —

Michael Coté: Yeah. And like the theoretical advice, and I stress the theoretical that I was typing out the other day, is that it seems like if you are an ISV who is becoming a SaaS or a SaaS, then obviously you run your own cloud, maybe not obviously, but it’s much more legit for you to run — to have a private cloud that you are using it. You know what I mean? You run your cloud, whether you are a Salesforce or whatever.

John M. Willis: That is a cloud. I mean, there are people that would argue against that, but I don’t, because I say SaaS is a cloud by definition. It’s just such an isolated silo of the cloud.

Michael Coté: I mean, I am sure there are some exceptions, but you would be insane if you were doing a SaaS that wasn’t run in a cloud like nowadays.

And the other thing is like, but then if you are like an enterprise or a buyer, your first assumption should be like, why can’t — so you are consuming two types of IT, well, at least two that I am going to care about here; one of them is off-the-shelf applications that you customize the crap out of, like your ERP system and stuff like that. And those are things that you should be pushing vendors to run on a public cloud somewhere, to run somewhere else, if possible. And it’s not all possible at the moment.

Then there is stuff that you actually — from the ground-up, customer applications that you might build, and those seem a little bit more, depending on what they are, they seem a little bit more legit that you would run them as a private cloud. And again, this is like strategic stuff you are pushing people to do, that you can do right now. But it does seem like, if you are just taking off-the-shelf applications, even like Exchange and you are running it on a private cloud, I don’t quite understand why you don’t just run it somewhere else.

John M. Willis: Again, then you can get into, like 17:47 is one of those weird animals where like you feel more comfortable about all of the email that goes on in your organization. Is that proprietary too? But I agree with you, I mean, if it’s – clearly, it could be isolated as a commodity service, like Salesforce.com, that’s a classic CRM, right?

Michael Coté: I guess the way to boil it down to is, if you are developing and deploying your own software, whether you are an ISV, a SaaS, or an enterprise, then maybe you can run it on a private cloud, but if you are just running IT in general, it seems like the push should be towards running it on a public cloud.

John M. Willis: If the application makes sense, like if you can — like the checkbox is, CRM in our organization, it’s okay to be out there for any of these reasons, if Exchange can be out there. So I think those are all — I think that — where was I going? It goes back to the catalog, right? So what makes the catalog so important is that, like the people who like — I always say, that operation is — Ops guys, we are like the groundskeeper of the baseball field, right? So play with me on this, play.

Michael Coté: They do like the mowing and the patterns, how do they do that?

John M. Willis: Yeah, they do some cool stuff, their grass looks pretty good in most cases. But here is the thing, like they said — like they make sure the chalk goes from home base to first base. They make sure that it’s the right distance. They don’t let anybody but the ball players walk on the grass. When it rains, they are the guys that kind of raise the hand and they have got to put the stuff up. So they set the kind of boundaries and then the players go off and like do what they have to do, but like, guys don’t run from first to third, right, because theoretically the ground — keep the lines correctly.

Michael Coté: There is policy that they follow along.

John M. Willis: There is a catalog. So my point is that, I think Ops guys are about like defining the catalog of services, making it frictionless, that they don’t have to come to me, they go to the catalog, and then they get what they need and when they need it.

Michael Coté: I mean, the 19:43 way I always put it, it’s like, IT people don’t actually want to talk to end users, so that’s kind of the goal.

(00:19:50)

John M. Willis: Yeah, right. And end users don’t want to talk to IT guys. But the kind of last meta point is that, like the guarantee is that you are going to get what you want, but I am still going to be the guy that kind of controls the way the field is laid out, because even though you think you are getting, you might go to a public cloud for this service, but the policy infrastructure says, that’s okay to go there. Well, you might go to some private, behind the firewall, can’t get even close to, and you are not even part of that decision.

Michael Coté: Yeah. As was said in the keynote yesterday, IT is left holding the bag no matter right. I mean, that’s the other interesting —

John M. Willis: The bag is ours, right?

Michael Coté: That’s the other interesting like longer cultural term thing is that, at a certain point there is only so much optimization you can do if IT is left holding the bag, because they are not going to let you be free wheeling and run from first to third. So if your business it makes more sense to run from first to third, you have got to figure out breaking that IT head —

John M. Willis: Or the assumption is that, you are in a business that your field is first to second to third to home, and so the baseball players figure out like phenomenally creative ways to get that — to score runs.

Michael Coté: Wormholes.

John M. Willis: Yeah, right, that was the answer. Okay. So yeah, I mean, that’s what I think vCloud is and then there is the whole vFabric, which I think is interesting. I don’t know —

Michael Coté: So you wanted like the typical like market context take or something like that, right?

John M. Willis: Well, it’s your analysis. So what I would ask you then, this is — so you spend a lot more time with IBM than other vendor. So what does vCloud mean from an industry standpoint, like you are going to talk to IBM?

Michael Coté: Yeah. IBM is an easy one. I mean, the people — let’s see, I will see if I can remember this list, the people that you would care about would be IBM, Dell, HP, Rackspace, just because they are into this, Amazon, and Microsoft, and there’s various little startups here and there and stuff like that, but I mean those are the big ones.

And I think with IBM, it’s pretty easy. IBM is — they are healthily schizophrenic on this stuff. On the one hand, they release like the zEnterprise, which is a mainframe, that’s sort of like against all this hardware doesn’t matter stuff we were just talking about. But the point of the zEnterprise or the zEnterprise —

John M. Willis: We are going to do it whether you like it or not.

Michael Coté: Yeah. Was like, there are actual cards and silicon and hardware in this machine that you use, that you can only get from this, whether it’s like encryption stuff or whatever. So it’s an increasingly narrowing sort of field that they sell it to.

John M. Willis: It seems like a really bad choice, but they are IBM.

Michael Coté: Well, I think there is a bunch of like financial people and insurance folks who are just like, that’s the platform they want. It’s kind of like people back in Texas who drive the big Diesel F, whatever trucks.

John M. Willis: That’s what they want.

Michael Coté: That’s what they want, and they need that pulling, and maybe they got those crazy crumb pipes coming out of them, and the little nuts hanging off of the tail end. So you got the —

John M. Willis: Little nuts.

Michael Coté: Yeah. You have never seen the — it’s like little crown nuts in a bag.

John M. Willis: Oh no.

Michael Coté: So you take like two little balls and put it in a rubber sack and so — sorry for the masculine humor there. But anyways. But the rest of IBM is pretty hip to all this stuff. I mean, the Software Group, which interestingly enough the Systems Group are now a part of, I would suggest probably under. I mean, Steve Mills runs Systems, which, he is the Software guy, so at least at the moment, Systems is kind of subservient to Software, which is an interesting thing for IBM. I don’t know if that will last or whatever. But anyway.

John M. Willis: Do they pick up the vCloud thing and run, or do they just — are they going to try to solve this with hardware?

Michael Coté: I think they will do both. I mean, what I tell people about IBM, I am sure I have said it here, is IBM’s business model is this, first, give us the money and then we will give you whatever you want. If we do it, it doesn’t matter, if it’s from someone else, who cares. You Give Us the Money Lebowski and we will give you the IT. Like what exactly you want, doesn’t matter, it’s just, you get the money.

So in that sense, if someone was an IBM customer and they wanted to get vCenter, IBM would gladly sell it to them, and they will try to wangle in Blades and WebSphere and DB2 and all sorts of stuff, but it’s not going to matter, because IBM is the primary point person. So I think, like I said, I think that’s a healthy schizophrenia, that allows IBM to survive through all sorts of things.

And then on the other end of that is someone like Microsoft, who are like, they are really locked in to only what they have. I mean, they have partnering things and stuff, but Microsoft doesn’t really have hardware, so they are kind of caught in that trap.

But the thing that Microsoft has, just like VMware, is ubiquity. I mean, it’s sort of — like with vCenter, if you are someone considering cloud stuff and you probably already have a lot of VMware, so it’s kind of like just use that. Like why try out something new from one of these wacky little startups or someone else when you could just keep using this stuff. And people love VMware. So Microsoft has that going on.

(00:25:06)

I think the biggest problem for Microsoft is getting people to extend .NET development into non-Microsoft worlds, right? So you have got like the vFabric and vCenter, and like it’s all Java and all this, but there is not really a lot of like .NET development we are talking about. And part of the issue is that, all the cloud stuff, like while there is vFabric, there is also AppFabric, which is a Microsoft thing that isn’t really the same thing as vFabric, but it’s all like the same.

Anyways, it’s a classic problem of Microsoft of, they like to isolate themselves, and then they are usually not quick enough to realize they need to work with other things. But I don’t know, we will see.

On the other hand, they are also — the good thing about VMware and Microsoft is, they are both — I mean, I don’t know, I have this secret theory, don’t tell anyone this. In the same way that Larry Ellison has a — what’s the politically correct way of putting it? He is very enthusiastic about being the IBM of the 1960s, I think VMware wants to be the Microsoft of the 1990s. I mean, they both sell license software, and they are both sort of infrastructure layer stuff, they are both Operating System. And their whole thing is moving licenses for things, and getting as many instances of their stuff deployed as possible.

John M. Willis: Oh yeah!

Michael Coté: The closer that VMware works with Intel and AMD and all those people, the better it is for them. The same way that — we used to have a phrase called Wintel, and it’s a similar sort of thing. In that sense, I mean, they are kind of unlikely, what’s that old word, they are simpatico, isn’t that what people used to say?

So then there’s people like Dell and HP who — I mean, one of those people, you get their software act together. That’s the concise way of putting it.

John M. Willis: Yeah, right. Well, I mean, Dell needs a lot of pieces, HP just needs a software piece, but there is none left. I was trying to think this morning, so I went to an HP thing, and I told you, I asked that question in IT Boot Camp, Oracle has the developers, VMware has the developers with Spring, IBM clearly is like always on the developers with like — so what does HP have? And the answer is nothing.

Their story is the kind of IBM, well, give it — I mean, I think you could simply say they are the same thing, give us the money and we’ll do whatever you want us to do. But I wonder, even this morning, like who would they buy, like is there even — I guess we could save that to last predictions.

Michael Coté: I mean, they could pull like a CA, like CA bought 3Tera and Nimsoft and stuff like that. So conceivably, someone like HP could buy Eucalyptus or they could do something even more wacky and become a significant contributor to OpenStack, like something like that. I mean, there is all sorts of stuff —

John M. Willis: But even then, I am still worrying about the — like the Oracle has the WebLogic.

Michael Coté: Oh yeah.

John M. Willis: IBM has the — where do you get that like app?

Michael Coté: Well, I mean, HP has this with EDS, but that’s the part missing from all these people, including VMware, is the services angle, right? I mean, that’s the thing IBM has.

John M. Willis: Yeah, right. So the vFabric thing, so I think the vFabric is the second really important thing. How much time have we got?

Michael Coté: I just have a few minutes.

John M. Willis: Alright, so this will finish up on the vFabric. So I think it’s an important thing to talk about is, so I think that’s another thing that I am excited about what — they are putting together this idea of you being able to kind of componentized, like abstracting the framework or the app layer as cloud componentry.

Michael Coté: Absolutely!

John M. Willis: So again, like this is the first time I have seen where I think vCloud and vCloud Director are real enough for me to say, yes, what’s that, you past that taste test or —

Michael Coté: It’s the cloud founder stuff, which is basically a way of packaging up Java-based applications.

John M. Willis: But really doing it right, because they have the whole RabbitMQ thing going now, and so they have really kind of positioned themselves to kind of build your own paths, and I think that’s —

Michael Coté: People have kind of joked that — I have been sitting through a lot of presentations, and we do as an analyst, and I think the best slickest thing I have seen so far is what you are talking about, is the — like having Rod Johnson, the SpringSource guy, go up there and explain — he doesn’t quite put it this dramatically, but almost, it’s basically like, there is a new way you develop applications and this is it. It’s all about queues and no SQL stuff and like whatever.

John M. Willis: 29:39.

Michael Coté: Exactly! That’s the phrase you use all the time, and that is like the — that’s been the most well-presented thing that they have done, because it really does — it kind of gets to that missing component stuff that I was talking about. I mean, I was talking about the Ops side of missing stuff. But the missing component is going to developers and saying like, you are a cloud developer, you develop things in this fashion.

(00:29:59)

John M. Willis: Right.

Michael Coté: I mean, they have stuff like Jim Stone, so they have got the queuing, and the Rabbit, which is a queuing system, and they have got a whole bunch of interesting stuff that kind of fills that out.

John M. Willis: Yeah. And like raising the bar, like so here’s the thing that’s important, and we can wrap up on this, is that, they have made a lot of acquisitions, but they are framing them well. So HP has made like a gazillion acquisitions, and I have still not seen the clarity in how they put it together.

Even to be critical about IBM, I mean, you have got your cloud burst, but lot of acquisitions — and I am not seeing the like, oh, I get it, and I think vFabric is like the first example of, like we do Infrastructure-as-a-Service, vCloud, we are now doing Platform-as-a-Service, but in the right way you can build your own and you have that flexibility.

It’s almost like private pass. Remember somebody asked us about a year ago on, like is there such a thing as a private pass, and I think that vFabric, it might actually fall into that category.

Michael Coté: Yeah, yeah. No, it gets back to like Microsoft and VMware being interesting, whatever, friendly pseudo people, is that — I hesitate to make such a sweeping statement, but I am pretty sure they are the only virtualization cloud people who also have a huge developer community. I mean, you could technically put Oracle in there, but I don’t think Oracle really — I don’t think that’s the thing.

John M. Willis: Those are those nasty DBA guys you keep talking about.

Michael Coté: Yeah, yeah. Whereas like IBM, they don’t really do software development framework anymore, I mean, Rational is all about processing and managing software development. And then HP doesn’t do software development. So it’s interesting the VMware and Microsoft are the ones who can bring that kind of — from Ops people to developer sort of thing. But on that, I have got —

John M. Willis: Then there is a Red Hat — I know you have got to go — we can’t do that now, but that’s just something — we should get a Red Hat guy on and really have them give us their perspective, because that’s an interesting developer, could be a cloud source as well.

Michael Coté: I think Red Hat is kind of like a dark horse in the sense that, they need to eat the right meal at some point and then they will get the power.

John M. Willis: I totally agree. I totally agree, no doubt.

Michael Coté: I don’t exactly know what that metaphor means. There is some sort of shift they have to go through. They will kind of get back to it. But anyways, that’s the report from VMworld 2010. We will see everyone next time.

John M. Willis: Sunny San Francisco.

Michael Coté: It’s always nice weather here.

Disclosure: numerous people mentioned are clients, including VMware.

rev="post-5244" 1 Comment

Categories: IT Management Podcast, Systems Management.

By cote

— September 2, 2010 at 11:05 am

Links for August 30th through September 1st

I’ve been at VMWorld this week. It really is the JavaOne of the operations world – nuts! Also, as you see, I did some reading up on Eucalyptus – look for a video going over their new release soon.

The Links

Alcatel-Lucent Acquires OpenPlug, a Cross-Platform Mobile Software Development Tool Provider — PARIS, September 1, 2010 /PRNewswire-FirstCall/ —
Apple event, September 1: The Boing Boing Liveblog – Boing Boing
"230,000 IOS activations per day"
ongoing by Tim Bray · A Story of O
Great mixture of sentiment in the comments.
Eucalyptus Anchors the Latest Cloud Software Stack
Eucalyptus Builds Scalability Into Private Clouds
MomentumSI Describes Complete Solution for “The Perfect Private / Hybrid Cloud” Technology Platform | Business Wire
Red Hat targets interoperability with cloud updates
Apple Said to Prepare 99-Cent TV Shows to Fend Off Netflix, Hulu
Consider the "what if Apple really got into TV" rumors and scenarios.
New entry into cloud computing
On vCloud competing with OpenStack – “With the popularity of VMware and maturity of the company, I'm sure it'd be on the private-cloud shortlist,” said Michael Coté, an analyst with RedMonk, in an online chat.
Puppet Camp: Learn More About Open Source Data Center Automation | Puppet Labs
Probably a good place to see what all the hoopla about next generation automation is.
Belgacom ICT Commercial Data Center Hosting Services Selects AccelOps Integrated Data Center Monitoring Solution – MarketWatch
"Belgacom ICT employs the AccelOps' platform to monitor, alert, assess and report its data center hosted services that support over 200 commercial enterprise-class customers"
VMware Outlines Vision for Next Decade of Information Technology
Combining Java (SpringSource) and virtualization (VMWare) to do private clouds. Requires a "new platform," of course – and the big question (on VMWare's path to become the next Microsoft, or IBM) is around parts of the stack VMWare doesn't own: an operating system, networking gear, servers, storage, services – "To ensure that the new generation of applications takes full advantage of virtualized and cloud infrastructure, a new cloud application platform is required. The platform combines an application framework that enhances developer productivity and ensures application portability with comprehensive integrated platform services that speed and optimize deployment. Applications built according to this new model will be portable across heterogeneous cloud environments and, when running on VMware-enabled clouds, will be able to intelligently share information with underlying infrastructure to maximize application performance, quality of service and infrastructure utilization."
VMware Plans New 'Cloud' Push – WSJ.com
Curtain Raiser: At VMworld, Cloud Will be Center Stage
Retargeting Ads Follow Surfers to Other Sites – NYTimes.com
More from the Better Junk Mail department.
Intel to acquire wireless chipmaker for $1.4bn
HP punts turnkey Matrix cloud • The Register
Hardware, VMWare, and services bundled together.
Citrix buys virtualization vendor VMLogix | The Industry Standard – InfoWorld

Disclosure: see the RedMonk client list for clients mentioned.

Comments Off on Links for August 30th through September 1st

Categories: Links.

By linkposter

— September 1, 2010 at 5:36 pm

OpenPlug – Another Alcatel-Lucent Buy

Aclatel-Lucent is buying OpenPlug, which is a development suite that aims to make multi-platform development in the mobile space easier. You start out developing in the OpenPlug IDE, ELIPS Studio (“based on Adobe’s Flex Builder”), and then your application is code-gened, cross-compile, and all that to different platforms.

Addressing Mobile Platform Fragmentation

Here’s the quote I gave for the press release:

“While demand for mobile apps is peaking, so is the fragmentation of mobile development platforms and tools,” said Michael Cote, analyst for Redmonk. “This makes it very expensive to target all devices and can lead to putting all your eggs in one, often tightly controlled basket, like Apple’s, or doing a least-common denominator, mobile web app that can’t benefit from full native device access. A tool that can target all platforms is hugely needed in this space, and OpenPlug should work well with the total mobile app development portfolio Alcatel-Lucent is putting together.”

Alcatel-Lucent has been building up a suit of tools and services to help telcos enter the mobile app space, and this purchase is trying to make the actual development of those apps in the highly fragmented market easier. Us tech-fiends may obsess over the iPhone and (increasingly) Android, but globally there’s many platforms, esp. when you throw in the “dumb phones,” that is, regular cellphones and “feature phones.”

Even among Android, recent release like the Dell Streak with its older version of the mobile OS have shown that application writers need to target fragmentation within a platform. It’s like the days before “write once, run anywhere” of the Java and .Net hegemony…and the days after with the PHP, Rails, and web application framework explosion.

Rising Interest in Mobile

“Mobile” is just about the only thing developers talk with me about now-a-days. On the non-nerd side, businesses are starting to realize that just as the tooled up for web sites over the past 10 years, mobile is increasingly a place their customers expect to do business. So there’s anecdotal evidence of a huge surge in interest.

What’s interesting about Alcatel-Lucent getting into this business (along with others – Motorola recently bought mobile tool vendor 280 North, IBM is an interesting example, recently having show-cased their mobile portfolio) is that they’re a known, trusted vendor to telcos. Getting large companies, like carriers, to change the core business – selling apps instead of just voice and data – is a Herculean task, and it helps to be friendly with the change-agents.

(As a side note, since I’m at VMworld this week: you see a similar thing happening, but with much more hunger, with service providers and hosters. Many of them are realizing – as Rackspace did awhile ago – that they’d do well to change over to providing cloud computing, giving them something more than the “stupid computing” [to ape “stupid networks”] to sell at ever reducing prices.)

Press release.
Coverage from Phil Harvey over at Light Reading Mobile.
My previous write-up of Alcatel-Lucent buying ProgrammableWeb, another items in this mobile app and services strategy.

Disclosure: Alcatel-Lucent is a client, as are Dell and IBM.

rev="post-5235" 1 Comment

Categories: Companies, Development Tools, Programming.

Tags: Alcatel-Lucent, m&a, mobile

By cote

— September 1, 2010 at 9:41 am

On HTML5 killing Flash

There’s an HTML5 killing Flash scab that keeps getting pulled off, every 1-2 months or so. In the most recent round, as I often do, I answered several question from a reporter (Howard Win in Computerworld) on the topic. The resulting story was pretty comprehensive and interesting.

Here’s the questions (from Howard), and my full answers, with some added comments here and there:

Q: What kind of challenges does any alternative to Flash face in the market now?

A: The main thing is gaining the ubiquity that Flash has. Flash is everywhere (desktop wise), trusted [well, except by several technical elite who see Flash as an annoyance], and works. Next up are tool support and video delivery support, which are not too difficult as folks like YouTube, Brightcove and others who have quickly tooled to alternatives show. The other difficulty are the features and DRM that Flash has built into it. These are things that could be built into any technology, it’s just that Flash has had time to work out the kinks. There was actually a great post from a YouTube staffer this morning going over all the reasons that they still need Flash, which applies to most everyone.

The interesting note there is that YouTube didn’t really express that
they “love” Flash [or “heart” it, a common techno-hipster way phrasing], just that they need it for these features. That’s a subtle, but important thing to notice.

Q: How much of a factor does online advertising play in the adoption of an alternative, since Flash is the entrenched technology most used by the online advertising community? As a counter view, there has been recent discussion, reporting, speculation that online porn could be a major factor in helping to accelerate the usage of HTML 5.

A: Those things play a role, for sure. If content people want to distribute (ads) and content people want to see (kittens, babies, kids singing, [pirated] videos, porn, etc.) is not available in some video format, that format will be less popular. However, I don’t think any love is lost from “the public” by making advertisers’ lives harder. Companies like Adobe, Microsoft, and Google
who want to suck in money from advertisers want to cater to them, but if those oh-so-popular videos of babies watching cats eating sandwiches and free porn are only available in a particular format [Flash or otherwise], people will stick with that format. Look at Real as an example: compared to Flash, there’s not much available in Real that people seem to care about. When’s the last time you played something in Real Player?

The point is, advertisers and other content producers will follow consumer demand or they’ll [have to] create such desirable products that people will use whatever video format they have to. No one is going to download some video player so they can see an ad. They’ll do it for cats, kids, and porn and as Netflix has shown, they’ll do it for movies and TV (Netflix requires Silverlight and has however many people using their on-demand stuff).

Q: In hindsight, what were the key factors that led to Flash’s success? Basically, what did Adobe do right?

Making the player free and embedded in the browser was a big deal. Other video players like Real and Windows Media didn’t have this “embeddedness” to them. Also, the frameworks for creating interactive animations let people build games and those darned interactive ads which motivated content producers and game players.

Q: Can Adobe’s success and strategy be copied by another company to champion another technology (such as what Apple and Google are doing with HTML 5), or have circumstances and times changed? In this regard, what do you think of Microsoft’s approach with Silverlight, up to now?

The problem facing Adobe now is from two ends of the spectrum: luxury computing and Wal-mart computing, Apple and Google. Apple used it’s design, advertising, and retail prowess to do the unthinkable: breath life into the dead space of consumer smart phones and then translate that into a vocal threat for Flash’s dominance on the web and mobile. Google, on the other hand, can use its piles of cash from selling better junk mail (online ads) to fund almost anything without having to worry about revenue generation.

The problem for any threat to Flash is adoption: giving people a reason to use the technology. If Silverlight is used as the “number 1” way to do DRM friendly video (as with Netflix) across all the global media companies (think how many cable and satellite operators there are, how many movie houses, TV interests, etc. – all of them wanting a way to deliver video on the web and make money at it), then it will be able to get Flash-like ubiquity.

The task is extremely difficult, taking many years and lots of money [or incentives] to grease the skids.

An important thing to keep in mind is that both Adobe and Microsoft have application development desires with Flash and Silverlight. While much of the focus is on video, both companies are hoping to use each platform as a new platform for programmers. Gaining video ubiquity helps with that effort, but the two are not exactly the same concerns.

Q: So what is your opinion of HTML 5, the Web formatting standard many are hoping will replace Flash — especially for delivering video online?

It’s impossible to predict which if HTML 5 video will displace Flash at this point. The YouTube post from today [back when I responded to the original inquiry] points out all sorts of missing things that HTML 5 video needs to catch up on. Also, I’m never really sure why HTML 5 and Flash video can’t co-exist, why it’s a zero-sum game. There are many people who despise Flash, but I’m not sure they’d love the alternative right out of the gate. Adobe has spent a lot of time optimizing Flash and I’d wager it’d take some time get HTML 5 video as awesome. The open source world hasn’t blown everyone out of the water with their video work thus far (hence the ongoing existence of Flash, Silverlight, etc.). Unless there’s some non-technical thing preventing it (patents, dastardly tactics, lack of available talent, etc.), it seems like video is simply difficult and takes time. Maybe if Google and Apple dumped a lot of cash and time into it, then perhaps whatever has prevented an open source video alternative from flourishing would be solved – Google’s VP8 patent freeing up is a good example [and Adobe’s announcement of support for it shows a willingness to go along with whatever encoding schemes come around], while Apple seems happy to hug patent-trolls). It’d be great if video could be commoditized and “free” in all sense of the word.

Q:. Ultimately, what do you think Flash’s, and HTML 5’s, futures will be within the next few years?

I think the best bet is for various video patent holders to try to cash in. Most people forget the GIF patent stink in the 90’s, resulting in an unthinkably fast creation of the PNG standard. With the threat and FUD of patent litigation, all sorts of things become possible and all the various parties involved get their acts together. Oddly, I think Adobe would come out OK in this, where-as folks like Apple would have some weird soul searching to do. Someone like Google has enough cash and will to bankroll patent disentanglements and, ultimately, the piles of cash they have are a good resource for evolving open video.

The other issue to keep in mind are the “Hollywood” interests. They saw what an open format like MP3 did to their music buddies and are not interested in that kind of disruption. People who own movies and TV are going to want as much DRM as possible, and new video formats that don’t satisfy those requirements are going to be tough to spread. Sure, there’s piracy, but once Hollywood gets it’s act together and figures out a Netflix/Hulu model, I think most people will pay. [Update: check out the Apple $0.99 cent an episode rumors out recently.] Most people already pay $80+ a month for cable which a bunch of crap no one wants to watch, so there’s a tolerance for a subscription budget, esp. if it’s for shows you actually want to watch.

Disclosure: Adobe and Microsoft are clients.

rev="post-5229" 2 Comments

Categories: RIA.

By cote

— August 30, 2010 at 10:29 am

Links for August 25th through August 27th

Rumored SAP-Red Hat Merger Seen as a Long Shot
In which I and fellow analyst-types dismiss the idea.
Buckwheat crisis hurts Russian pride
The author says that buckwheat crop failures in Russia are causing much distress, esp. since it means importing from China.
The Importance of Context: Why Enterprise 2.0 Still Fails to Deliver Value
Providing a data fabric, sort of like if mashups and SOA had a really productivity centric kid that was looking to use all this 2.0 stuff to get stuff done.
Strata 2011 – O’Reilly Conferences, February 01 – 03, 2011, Santa Clara, CA
A new Big Data/analytics conference from O’Reilly. I bet all the Big Data and NoSQL people will go wile for it.
Netflix Adds iPhone, iPod Touch Apps
Really? Watching Netflix on my iPhone?
ACUG – Sponsorship details – Austin Cloud Users Group | Google Groups
A pretty nice deal, esp. if they start recording a publishing the sessions.
Salesforce.com Positioned as a Leader in the Magic Quadrant for Sales Force Automation — SAN FRANCISCO, Aug. 25 /PRNewswire-FirstCall/ —
Hey, it’s nice to be in the upper corner.
Marten Mickos defends honor of Ubuntu’s Koala food • The Register
On the topic of “scuttlebutt” that Eucalyptus doesn’t scale well.
Amazon: New Kindles Selling At Record Rates
Now that’s tablet pricing for you!

Disclosure: see the RedMonk client list for clients mentioned.

Comments Off on Links for August 25th through August 27th

Categories: Links.

By linkposter

— August 27, 2010 at 4:59 pm

IT Management and Cloud Podcast, Episode 78

As mentioned earlier, I gave an introductory, overview talk on cloud computing earlier this week to the Austin Cloud User Group. I managed to make a recording of the talk, and in place of John and I figuring out when to record a real episode, I bring you, dear listeners, that talk. For long time listeners, it may be worth skipping: it brings together several points from past talks on this topic, but it is the first time I’m put together “everything” into one talk. As such, it stands a good overview of what cloud computing from the angle of “why should I care?” and, more importantly, “what can I do with it?”

Download the episode directly right here, subscribe to the feed in iTunes or other podcatcher to have episodes downloaded automatically, or just click play below to listen to it right here:

You can check out the slides in my previous post on the topic.

The Austin Cloud User’s Group is starting off nicely. I’d love to see people come in and talk about their first-hand experience using the cloud, as well as more technological overviews. The next talk will be about security in the cloud, and then some Rackspace OpenStack folks will come in to talk about OpenStack. It’s also a great, cheap sponsorship opportunity – just buy pizza and drinks and you get sponsorship which means a 5 minute speaking slot at the beginning of the meeting. It looks like they’ll be recording and publishing these meeting on the web so they should get some nice reach and permanency for those who can’t show up to the meeting.

Disclosure: Rackspace is a client, as are several people mentioned in the talk here. See the RedMonk client list.

Comments Off on IT Management and Cloud Podcast, Episode 78

Categories: Cloud, IT Management Podcast, presentations, Systems Management.

Tags: ACUG

By cote

— August 26, 2010 at 1:54 pm

Next » « Previous

Blogs

RedMonk

Liferay evolving beyond portals – the app server vacuum – Quick Analysis

The Motherhood and Apple-pie

Write once, configure everywhere

PaaS and Mobile

More

Know your developer type

Developers Types

ISV vs. Corporate Developers

Links for September 2nd through September 6th

Data integration in a cloud world with Joe Dwyer – make all #009

Show Notes

Transcript

VMworld 2010 – IT Management & Cloud Podcast #79

Transcript

Links for August 30th through September 1st

The Links

OpenPlug – Another Alcatel-Lucent Buy

Addressing Mobile Platform Fragmentation

Rising Interest in Mobile

More

On HTML5 killing Flash

Links for August 25th through August 27th

IT Management and Cloud Podcast, Episode 78

About Coté's People Over Process

Recent Posts

Recent Comments

Archives

Categories

Meta