 please a warm welcome for Bernard Golden. Thanks. We have a clicker here, so you jump into it, so you already heard about me, so you don't need to see me brag. So the state of the computing industry, I'd like to sort of start with where we stand today in the state of the computing industry. So October 27th, Amazon announced its numbers, $3.2 billion for the quarter, 56% growth, October 20th, Microsoft. Microsoft doesn't break out its Azure numbers versus all of its other stuff like Office 365, but estimated it's a very funny timer that's here, it looks like a bingo ticket. And by the way, I'm going to speak and then we hope we'll have some time for questions. I think that's what we discussed last night, we'll do questions at the end of the presentation. But the estimate from the analysis is maybe $500 million to $1 billion per quarter, 120% growth, and Microsoft just announced its numbers yesterday, maybe, last week, and they're still running at about 100% growth. Amazon is going to announce its numbers later this week and they'll be very interested to see what their growth rate is, but by comparison, VMware, which is a stalwart in the industry, $1.7 billion at about 6% growth, and they set up skyrockets at 6%. They say we're doing great. So this gives you a sense of sort of the growth rate. This is, if you extrapolate these growth rates until 2020, Amazon and Microsoft will be running at about $100 billion a year. So a huge force in the industry, a huge platform. I got exposed to photography about 10 years ago, I was writing virtualization for dummies, I put a chapter in about the future of virtualization, went to a meetup in this shabby, disused classroom at a community center and saw a guy from Amazon get up and talk about a new service they called Amazon Web Services. I said that is going to be a profound transformation in IT, and we're now seeing it really by really digging in the industry. Now this is quite interesting. Application portfolio deployment plans because there's lots of conversation about how much impact will cloud computing have for IT, how fast will it go and all this. J.P. Morgan did a survey of 200 large enterprise CIOs. Average IT spent, now that varied. Some of them spent $40 million a year, some of them spent several billion, but it averaged 600 million. So I would say this is large enterprises. Dramatic growth. So they're at about 16% application portfolio into cloud computing in 2016, and they said they'd be at 41% by 2020. Is that sort of a line with everybody here? Anybody sort of go, yeah, that looks like us? It's okay for you to put your hands up. And the interesting question is what about post 2020? They just surveyed through 2020. So does that mean it goes to 41% and then that's it? I've heard people say that there's going to be some amount of growth for cloud providers, then it'll plateau, or will it continue to grow? My perspective is it will probably continue to grow. So public cloud computing, I call the future of IT. I believe that's where the majority of computing will be done. And certainly new applications will be put into that space, and I believe there will be a huge pressure to IT organizations to take legacy current data center applications and do something with them because they represent a huge amount of spend and typically aren't seen as delivering huge amounts of value. So old computing, it took months to availability because you would often say I've got an idea for a project, I need to put it into the budget plan, it's going to take several months, several quarters, static, once you put it in, it's sort of stable, if you say I need some more, they go okay, go back to the budget plan, hard to scale, requires an upfront capital investment, and infrastructure is interesting. I was just talking to the guy next to me, we were talking about that, and his company says we've concluded that buying a server is a capital conflict with all the rest of the stuff we do, it's a capital intensive company, where do we want to spend our capital? Servers aren't that so relevant for us, but what we can spend our money on on other things quite is, so very interesting. It's very infrastructure focused, you know, we have a whole group that's called the infrastructure group, that's what they do, they buy stuff and so forth. Cloud computing, minutes to availability, it's dynamic, so your application becomes more popular, no problem, you just ask them, say send some more. That's trivial to scale, if you're doing things correctly it is. Usage-based cost, which is both a benefit and a challenge, because it can be sort of like it's great, you only spend it for what you use, but if you don't pay attention, it's kind of like parking a taxi at the bottom of your driveway, and just having it sit there all the time, in case you might someday want to take a taxi somewhere. But it's very application focused, and crucially that typically flows through a business unit that says we need something to help run our business. So let's talk about some innovation examples. Capital One, very interesting company, I started working with them when I was with ActiveState, they were using our product, and they liked our product because it was container based, it was cloud foundry based, you could do portable applications on-premise, off-premise, they said we're going to build our own private cloud because we're a financial institution, we're going to do that, we said no problem, we can run private, they had a group that was using us in the public cloud, they had their lab, Innovation Lab is doing it. Over six months, as we began working with them, they said you know what, we're not going to do, that private cloud thing isn't going to work, it's too hard, cost too much, all that, and they are all in on AWS now. So one of their innovation examples is they've taken it, they've written some skills for the Amazon Echo, and you can now say two-year Echo or Alexa, what is my account balance, and it'll tell you, or you can say pay my credit card, and they're very aggressive on this, they're really trying to develop innovation and become next generation bank, and so they're doing a very interesting kind of thing. By the way, we have an Echo, and I chose to call, you can call it Echo or Alexa or something, I said I'm going to call it Echo because I don't want to anthropomorphize this thing, I don't want to believe I'm talking to someone, and so we use Echo. This is another one, and Huddle, this is not the group wear company that you might have heard of or might even use, this is a very interesting, it was a startup about four or five years ago, I saw them speak at the first Amazon re-invent, and they were a high school football application. If you're an ambitious high school football player, you need to have your reel, just like if you're an actor, you need to have your reel. Always a challenge because some assistant coach has an old line video camera for a phone, and then where you store the videos, and how do you make them available to your college that you want to apply to. So Huddle started off saying, we're going to be a cloud-based application for high school videos, and we're going to enable that. Very successful, started to grow, by the time I met them, they were used by something like 12,000 to 15,000 high schools in America, they were used by 80 to 120 Division I football programs, they were used by 12 pro teams. They have now branched out, this is an extension of their thing, they now have sideline analysis of plays, they can talk to the guy, I guess it's always a guy, up in the press booth, very sophisticated real-time analysis, they can sort of track this stuff, all on Amazon. They started on Amazon, they completely take advantage of Amazon, and they really couldn't build their business, because as you might imagine, football is a very spiky kind of use. Very heavy on Sundays, or the weekends, depending if it's college or high school or pro, very spiky during the course of the year, they use Amazon to manage that. Very, very interesting company. So what does cloud computing mean for IT tools and processes? We had a very lively discussion about this last night, we had a speaker's dinner, and this is a big challenge. So to quote the great sage, Warren Buffett, only when the tide goes out, you discover who's been swimming naked. Which means when something changes, it's only when you find people who are exposed in a way that they weren't. And the tech translation of that is, when you remove one bottleneck, you expose the next bottleneck, and that's a very well-established technical principle. To elaborate, now that infrastructure is fast and cheap, application development and employment speed is now the primary IT bottleneck, and the primary IT business, the business differentiator. This, I believe, will be the big challenge for IT organizations over the next five years. How do we get our application process to run at the speed that we need it to, now that infrastructure is no longer a problem? And we're being expected to develop and deliver applications like that echo. So, how do we might look at and say, well, that's a startup, they do this kind of quirky area, that's not really us. Capital one, that's a mainstream financial institution. And that's what mainstream companies are doing. They have to be able to move at that kind of speed, at this new sort of startup speed and so forth. And where that's gonna play out in IT is application development and employment speed. So I wanna talk about addressing application bottlenecks. One of the biggest challenges for IT is how do we enable the business side to move as fast as it needs to in this new economy, in this new society, in this new era of customer expectations. So you need to be able to accelerate innovation. There has to be closer collaboration. There has to be really a need. The model of, there's the business people and they develop a set of specs and they send it over to us. And then we figure it out and then we deliver it. That world is dead. You have to have really intense integration. In Silicon Valley, the way that often works is you take the product manager and you plop him or her down in the middle of the engineers. So it's a constant interaction on an ongoing basis. Frequent release, continuous integration. So in other words, we have to deliver stuff into the business use really quickly and we have to be able to collaborate with business. Because IT is no longer supporting the business, IT is the business. So Agile, you've got to increase throughput. So automated operations, I was talking about that this morning. How do you automate it? Now that infrastructure is automatable, how do you accomplish that? How do you achieve that? How do you build tools that do that? Continuous deployment and distributed monitoring and management, which I was having conversations just before I walked in here, about distributed monitoring and management. The challenge I think for a lot of IT organizations is the tools and processes they built were appropriate for a time when infrastructure was that old model. When it took quarters to get it in. When you installed it, it didn't change. When you didn't really scale it very often, the tools were very appropriate for that. This world where resources come and go. They're constantly joining or leaving the resource pool. The tools have to be able to support that, have to be able to monitor it, have to be able to manage it. So that looks like Agile and DevOps. Agile is sort of the development process. DevOps is the operation and deployment process. Those have to be moving on. In terms of the cloud being the location where you deploy that, infrastructure availability, I just talked about that. Scaling elasticity and resilience. The flip side of dynamic stuff is stuff breaks. The number one rule that Amazon puts forward for its applications is everything breaks all the time for its infrastructure. So you have to build applications that assume that resiliency is built at the application level, not at the infrastructure level. So you need to be able to recover from that. You need to be able to respond to changing loads and so forth. And then cost effectiveness. And this I think is an area where we're gonna have lots of pressure going forward. And again we talked about this quite a bit at dinner last night. If you're someone like Capital One and you say I'm gonna build something and this is gonna be the primary interface that my customers use. Maybe it's a mobile app, maybe it's an Echo, an Alexa skill. That becomes part of your cost of goods. So you have other kinds of cost of goods that you have to account for, you have to be able to do that with cloud cost management. And some organizations I think do a really good job of that and some don't do a very good job of it and they just hope that things are gonna work out. I recall I had this very funny conversation. I was talking with the CIO of Charter Communications, a big cable company, a big content company. And they were just getting into cloud computing and he said you know we got onto Amazon and it was fantastic. We had a developer, that developer started working and it was going really quickly. The first month it was only 400 bucks, it was absolutely fantastic. He said the second month we got the bill and it was $10,000. And he said now we're a big company, we're gonna afford $10,000 but you kinda wanna know what happened between the 400 and the 10,000. But this becomes more critical when cloud is part of your cost of goods and those applications are part of your cost of goods. All of a sudden you gotta really manage that. It's the same sort of model if you're building washers and dryers or cars or whatever it might be, you need to know how much it costs for motors or agitators or whatever the heck it is. So you can say are we selling at the right price? Do we have the right cost structure? Cloud will become that and so the ability to be cost effective and track your costs and make sure they're being used effectively and so forth. And then all of that sort of feeds back into the business in that it enables you to do things like experiment. Let's try something. Let's roll it out. Let's begin using something and see what kind of market feedback we get. Let's put trials out and so forth. I had a very interesting conversation with Qualcomm one time when I was at ActiveState, they were using our product. And the person said one of the great things about using your product and I'm not with ActiveState anymore so I'm not bragging on us or anything like that. But he said one of the great things about this and the cloud is we can have a product manager sit down with a developer, sort of sketch something out on a whiteboard, have them go up. In the past it would be, they would do that and then three months later the developer would come back and say here's what I put together. Is this what you had in mind? They said now that person goes off the next day the developer's back saying is this what you had in mind? Is this what you wanted? And they can get a much faster loop and so you can do trials, prototyping and fast follow on. And if it works well, you put the pedal to the metal and if it doesn't, you say well that was a failure, let's shut it down. So I want to talk about one of the big challenges for I think many IT organizations is, and I've been as guilty of this as anyone, is treating cloud computing in this IPS model, infrastructure, platform, software as a service. Like it's a very neat sort of set of levels and you kind of go is this infrastructure as a service or is this platform as a service or is this software? That world is dead. There is no such thing as an infrastructure as a service provider of the big main players or a platform as a service provider or a software service. They're just a big muddle, a big mush. They all provide those things. This is a screenshot of Amazon that I took about a week or two weeks ago of their services. There's some of them that you might recognize in infrastructure like EC2 or EC2 Container Service. There's some that you might regard as storage, but then there's a whole lot of other stuff that doesn't look like any of this stuff. It's not really a platform as a service, but things like Opsworks or CloudTrail. They have a whole set of code pipelines that you can use and so on and they've got a lot of things that look like applications. They've got analytics. They've got a whole new set of artificial intelligence applications that they rolled out. So really their world is much more of a mix and match services to build the exact application you want. And so the IPS world is dead and this new world is forward and the big challenge is how do you assemble applications or how do you decide which are the right services? So welcome to the world of infra-platware is what I call it and that's the world that you'll be living in. You'll have to be deciding this and you'll sort of make it a bet on which provider am I gonna do that? Because every one of those services, I could put up an Azure list and it would be maybe not as big as that but also quite large and they'd be very different. They would operate differently. They'd be sliced and diced differently in terms of the services they provide. So you sort of have to choose about that. So this is the Amazon yearly feature improvements over time. Now some of these are major new services, something like Kinesis or some of those Lex or Poly which are their new AI things. Some of them are relatively small improvements or they might be a new instance type. Like for example, at the last reinvent, they announced an instance type that you can attach a graphics card capability to a regular instance. So if you have certain kinds of analytics that you wanna be able to run, machine learning runs really well on that but just wanna attach to a regular server, you can do that. That would count as a service improvement. It might just be you can use SMS but now you can use larger things. But the pace of innovation is remarkable. They hit over 1,000 feature improvements last year and you can sort of see that that's a very steeply going curve. So it's likely to be 1,500 or 2,000 this year. Every one of those is some improvement in what they offer. This is the challenge for IT organizations I believe is where do they decide that they're going to focus their attention and energies. And again, the infrastructure I think, cloud provides real benefits in that. But you have to say, what am I gonna do with around all of those services? The old application services, IT was responsible for installation, configuration, management, and upgrades. So if you said I'm gonna have a DevOps thing and I'm gonna use Jenkins and I'm gonna use this kind of, and I'm gonna use, the name has escaped me, but I'm gonna use some automated test harness and all that. It was basically, I'm gonna install all those components. They're open source, so I don't have to pay for them. That's great. But open source is a science project. You download them. You install them. You configure them. You make sure that you're up to date with them. You integrate them so they can talk to one another. Those are typically custom integrations. It's often costly. I once again heard the joke yesterday that open source is like a free puppy that doesn't cost you anything until you have the puppy and then a lot of attention. But it's often costly. And the cost may not be immediately obvious. But six months later, when you've got somebody saying, well, I've gotta upgrade this component because it can only talk to this other component at this level and I've gotta upgrade the integration. I gotta redo that. It's hard to scale because it's challenging because you gotta go okay. So we just ran out of resources on our Chef server. We need to upgrade our Chef server. You're back into that old thing of I need more resources and so forth. You do get low lock in though, which is great. You're not committed to something. You can say I'm gonna use Chef no matter where. So that's a real benefit. But crucially, this is below what I call the value line. The value line is from the business perspective, does this help my application? And if you say, well, I'm building a whole internal DevOps capability, that's a good thing. But nobody from a business unit is gonna say, oh, that helps my application. That's specific to my application. That helps me roll out some kind of analytics around my washers and dryers. They go, that's great, but how does that help me? So it's below the value line. The value line is the place where it delivers value to the specific business application as opposed to being in general capability. Cloud application service. Well, the cloud provider takes care of all that integration, your installation configuration, all that. That's all off on them. It's really, it's open sources of service in a lot of ways. And there's a lot of discussions in sort of the open source world and all that. Is that fair? Is that right? Do they really contribute? Amazon for a long time was notorious for consuming open source, but not really contributing much back. They seem to really change that tune. But it is open sources of service in many ways. So you have a choice. Do I do it myself? Do I use it from them? It's pre-integrated. They make sure it talks to one another. There are other services, which is great. It's often free. Many of the Amazon code pipeline services are free. So they're just added on. They expect you to pay for their computing resources, but they throw in other stuff. Nord's very cheap. Anybody, so anybody here working with serverless stuff like Lambda and so forth? Few people. Unbelievably trivial. I went to the first, trivial in terms of cost. I went to the first serverless conference. It was sort of almost a community company. And somebody from Nordstrom got up and they had built like a real deal production application in Lambda. And they said, yeah, we're using it, but we haven't yet broken out of the free tier on Lambda. So we run our whole application and it doesn't cost us anything for the Lambda stuff. So often free. Trivial to scale. And in particular, this Lambda, this new functions as a service stuff, it's great because you just put in your code and they figure out how much you need to run it. So you don't have to pay any attention to the scalability. It does have high lock-in. That is the flip side of it. But crucially, it's above the value line. You can say I'm gonna use this at the basis of the application where there's some notion of it's delivering value at the business level, maybe it's delivering revenue and so forth. So it's above the value line. So the challenge of open. So it's the extent of support for rich below value line features. So how well will this provide the kinds of capability? If you decide to do it yourself, how well does this support the kinds of capabilities you have? And crucially, how well does it compare to what you can get from some external provider who delivers it on support? Exposure to above value line services. One of the challenges, so I worked at a company called Instradius and we're a cloud management product and our core value proposition was you can design an orchestration of an application that you can migrate from Amazon to Microsoft to OpenStack and so forth. Great value proposition. People loved it and so forth. But one of the biggest challenges was they said, yeah, but you're only doing the vanilla. You're only doing the lowest common denominator. What if I want to use that instance type that has a graphics array card attached to it? You don't have support for that. So you're sort of limiting me to the most basic capabilities, not exposure to above value line services or those rich below value line features. CapEx versus OpEx, this is, as I said, I think it's gonna be a really ongoing challenge and requirement for IT organizations. Figuring out CapEx versus OpEx and cost versus business value, it'll become part of COGS. So in conclusion, and so cloud represents an increasing proportion of the IT portfolio. And the interesting question is, does your experience attract what JP Morgan found from its surveys or and where will that end beyond 2020? The IPS worldview is dead. So then you've got this rich melange of services. You've got to figure out how to construct from. How do you build really innovative applications using them? Which is the right provider to provide the right set of services? Which of the providers are you willing to commit to and get married to because there is high lock-in around those kinds of services? The value line presents the lock-in challenge, as I just mentioned. There's an ongoing tension between open and business value. Businesses, so for example, Capital One said we want to really embrace the Amazon Echo. We want to embrace Alexa. That led them down the direction of AWS and they're sort of committed to AWS and it'd be very difficult for them to say, we want to pick that up and move it. It'd be very expensive to do that. Cloud computing changes everything. This is a food fight, which is in a way the future of IT in some respects. And I do believe we are in the most exciting and yet the most challenging part of IT that any of us will have ever seen and it's going to be a challenge going forward but a time of incredible innovation. I write for CIO and one of the things you always see in CIO magazine articles is that IT wants to seat at the table. How many of you have seen that or said that? The rest of you are just being modest. Everyone's heard that. We want to seat at the table. We want to be at the decision, which is great. Except when you sit down at the table and everyone turns to you and says, so what are we going to do? And that's where IT is because now IT is the business, not supporting business. So I think we have some time for questions and I'll take them. A bit bright, but. Round of applause for Bernard, please. Thank you. I said last night, is it Bernard or Bernard? And he said, hey, I've been called worse. Either it's fine, depending on where you come from. And I said, well, I'll try and do Bernard, but I just lapsed into my natural Bernard. Really Bernard sounds more sophisticated, so I'm okay with it. There you go, there you go. So if anyone has any questions, then please. We've got one or two people wondering around collecting them, I know there's one over here. If you can gather them Maggie, that would be great. And I'll, we'll go from here. But one I had, my background, which I don't usually admit to anymore is an IP attorney. And one of the questions I had, does this pre-configured open source collection of things take away my challenges of open source licensing, the complexities of that? I've never really thought about that, but I guess you'd hope that Amazon and Microsoft and Google have really good IP attorneys and they've sort of, sort of through that. I mean, I think it probably doesn't theoretically, practically, I think you'd have a pretty strong shield. That's what I think, if you're buying it as a service, it's less a case of, okay, I'll mix this with this. Somebody comes to complain and you say, hey, talk to Amazon. Yeah, talk to the guys with the deep pockets. Okay, we've got questions floating in. There's no mention of Google Cloud. Any comment on that? Well, Google is one of the big three or four IBM also would perhaps fall into that category as well. Google does not announce its numbers. They just roll them into their overall general main numbers, which is some, mostly advertising. So it's very difficult to figure out how much that would be and I really want to focus on the financial side of things, so I think many people underestimate or under appreciate the trends around revenues. But Google is finally looking like they're very serious about the cloud business and really spending a lot and investing a lot into it. They are not there in comparison to the richness of Azure, I should say, they're not there in the richness or the adoption and it remains to be seen how fast they'll catch up or if they'll catch up. But I would definitely put them there and I wouldn't count them out by any means. And they do have a fortune of having found the greatest business model in history and they just basically open the doors every day and money floods in. So they have no problem investing at the level they need to. Right, the one to watch. You shared the Buffett quote on application development being the primary bottleneck for IT. When that bottleneck is relieved, what is next down the line? You know, I'm just solving the next bottleneck, I'm not. Well, I believe that the next bottleneck will reside in the how do you more closely integrate business and IT because that's the way companies are gonna be doing business in every dimension of their stuff. And again, at dinner we had a very lively discussion about the role of product managers. And that's sort of the interface of business to IT. I said, so where do you find product managers who really understand what you can do with IT in terms of a product? The example I used was this withing scale. How many people have heard of a withing scale? But it's this IT enabled scale that basically stand on it. It tells you your weight, it tells you your body fat composition. It has an app that basically you can track it over time. But it'll also share it on social media and you sort of go, well, why would you wanna do that? You know, who? But Penn Gillette, the magician, you know, was a big tall fat guy. His doctor said, if you don't lose weight, you're gonna be dead. And he sort of said, well, that's not a very good thing. So he bought a withing scale and every day it was sending out his weight and he had this reinforcement stuff of his social group that would go, you go girl, you know, that was great. He waited and paying in there. He lost 100 some pounds, which is fantastic. He's there and you can go see him. But how do you develop product managers who have that kind of vision who can understand what can be done with IT and so forth? And the answer was sort of like, yeah, that's a really good question. So I think that kind of a long-winded statement about where the next bottleneck is. Okay. He's one here that can you tell me how sure you've had before. But is very, how can end user organizations measure the value of using cloud in business terms? In other words, how would you pitch it to us? How would you kind of sell it to the people? Well, I mean, I think most people work from examples and stories. So I wouldn't abstract it as a theoretical thing. I would either come in and say, look what we built in our skunk works that show off what can be done. Or perhaps just as effective, look at what competitor X did and is now running it, has a new offering that is $50 million here. So for example, I've got to believe Omron, which is a Japanese corporation that does a lot of home medical type things. They have scales they have. I used to have, but with these scale now, I used to have an Omron scale. They do blood pressure and all that. You've got to believe that they are under a ton of pressure to how can you come up with something as functional as with things. And by the way, the withing stuff is beautifully designed. I mean, it's sort of, if I use Apple as the benchmark, it's at that level. So you've got to believe those. The Omron CEO, if they come in and say, we've got a way that we can go about that, we've got to use this thing called the cloud. It's going to be like, well, why are you asking me? Get on it. Because my bonus depends on us. It's more by comparison than warning out them. Would your thesis that IPS is dead, not conflict with the aim of agility, i.e. flexibility without lock-in and freedom of algorithms? Well, for sure, as I said, it conflicts with around the issue of lock-in. This new world is much more difficult, much more challenging to be portable. So I think that's the case. There was another half of that question that I asked. Oh, does it conflict with the aim of agility? Flexibility without lock-in and freedom of algorithms? Well, yeah, the flexibility, it definitely conflicts with that. Whoever put in about the freedom of algorithms, if you could track me down, I'd like to understand kind of what you meant by that more. It's pretty intriguing. But for sure, algorithms are... Maybe that's the next bottle, I don't know. Okay, this is a question headed data privacy. How does all this cloud stuff work under the rise of constraints of geolocation of personal data storage, the conflicts inherent in overlapping requirements, and the need for legal certainty around these at very high precision? Well, the answer is it doesn't solve it at all. It presents additional ones beyond even if you had your own data center. No question about that. But I wouldn't say that even if you have your own data center, you're still subject to things like conflicting requirements or requirements that are extended and so forth. The practical result of this, and this is becoming a more important issue, and I believe with the actions of our new administration, it's going to become even more important going forward for many, many countries. The practical is that the cop providers are going to start locating more... They're going to put more readings out there. They're going to be more country specific, so we've seen that in Europe for sure. Amazon and Microsoft were both originally in Ireland. They are now opening up in Germany. They've got it in England and so forth, and Microsoft even did something where they sort of outsourced something around to T-Mobile, so it was owned by a local company rather than by an American company. We'll see a lot more of that sort of thing, and this is going to come. So cloud computing makes it more complex, but it's not an either or where it's like, well, if I did it myself, it would all be no problems, and if it's a cloud guys, it's a real problem. It's a continuum. What warnings would you have for large organizations that are going to build a private cloud or stepping stone to public cloud? Friends don't let friends build private clouds. So there's a couple of things. One, I was very involved with OpenStack, so it's an open source, which is great, but I think it's fair to say that that has been very challenging to make it consumable by organizations. And running cloud environments, I'm going to talk real cloud environments, not just we have a way that you can fill out a web form and send something off as a trouble ticket to somebody who will do that. I mean, real cloud, they're complicated to run. They're really hard to run, and they're super hard to run at decent scale. And I think the biggest challenge is how good are you at being able to do that and how much value does that provide versus what you might do from a public provider? And I think that's going to be a challenge going forward. Just got a few minutes left, and many questions. And they're very lengthy questions. So given the number of cloud providers who support open standards and prevent lock-in, why would an organization commit to a vendor that promotes vendor lock-in? Why would they commit to a vendor that promotes vendor lock-in? I don't know who the provider would be that promotes open. That would be an interesting question. But generally speaking, the big three, the big four, their pace of innovation is unbelievably faster than everybody else in the industry. So it's sort of if you want to be taking advantage of those kinds of services that I was talking about, you've got to go with one of those. The other ones, you know, I mean, some cloud providers are at the sort of the level that I call, they vend virtual machines, which is fine. But then you're responsible for all the value added into that. And the question is where do you want to be, you know, providing value versus doing below-the-value line investment? Okay. Let's see. When organizations use LONDA functions, are they getting more provider-dependent? Should we need an open standard for all integrations as a code and that the serverless deployments can be provided agnostic? So the question, so the answer is yes, you are getting locked in. Amazon's Lambda works differently than Azure functions, which now I believe gone into production, are different than what Google's done, are different than blue mixes version of that. IBM did release an open source version of a function of service. It was just released in February, I think. You know, you're trading off openness versus functionality. And in particular, if you use the open source version of that, again, you're back to the free puppy, which you have to decide, is that where I want to be? Is that where I want to make my investment? You know, you're trading off openness versus speed or efficiency. You've got to decide where you sit on that continuum. Okay. With componentization and mixing and matching, is the end user customer really ready for this? What about their own enterprise architecture? How do they avoid locking? Well, I think we talked about locking quite a bit, so I won't go after that. But yeah, I think this is a big challenge because many enterprise architectures have been designed with, you know, sort of more of an IPS type view of a very structured set of components and these new services. And by the way, it's a challenge for everyone, even the providers. Simply learn who I partner with, you know, does certification training. One of the challenges is how quickly can Amazon, and I'm just going to use that as an example, roll out a certification around Lambda? They're probably a year or two years away from that and yet people are adopting it right now. So it's a big challenge for everyone. There is no easy answer. I mean, we're at a time of an incredible pace of innovation. You've got to decide where you want to sit with that. A couple of questions I didn't get to, but we have one minute. I'm going to do this one. IT is still typically recording into the CFO and driven by cost efficiency, is the statement. The business has already moved to exploit the cloud. Is it too late for IT to catch up? This is sort of the shadow IT slash rogue IT. And I was speaking one time and somebody in the audience brought up and I said something about cloud and all this. He said, you know, that sounds like shadow IT, an IT audience. I said, well, you know, one person's shadow IT is another person's getting my job done. And I don't think it's too late because I think, frankly, business organizations, you talk about the peer BU's, I think they're not going to, this is very challenging. I mean, you could sort of say the same thing. If you're Toyota and you're somebody who's in marketing cars, there's a whole lot of knowledge around manufacturing cars that that person can't have, that they have to rely on somebody to rely on that. So I don't, I think these are the glory days for IT, but it's a challenging, very changing world. But now, thank you very much. Alright, thanks.