My recap of Cloud Foundry Day is up on the Tanzu blog: check it out!

I gave an opening talk at Cloud Foundry Day last week. I ended up shortening it a lot and, of course, I didn’t exactly give the talk I’d written down. Here’s the script I wrote for myself. It goes over the opportunity the Cloud Foundry community has right now. I gave a talk later in the day that more systematically made the case below.

Well, thanks for coming. To put in the effort and expense, and the time: y'all in the room here are definitely die-hard, lifers in the Cloud Foundry community. And it's thanks to you (and those who couldn't make it but wanted to), that have been in the community so long, that we're here and have given the users of our platform such a great experience all these years. So, thanks for that. You should be pleased with yourself, proud - all that.

It can be hard to stick in this community - some people and organizations have given up, bedazzled by perceived substitutes for a great developer platform. Some of us in the community have been bedazzled here and there. But the result of that bedazzlement - and, you know, I'm talking about Kubernetes, obviously - is often befuddlement. You're not supposed to say negative things about competition at things like this, or, really, even acknowledge that they exist! It's necessary, maybe even cathartic for some, though, to acknowledge the elephant we've been living with for a long time.

Here's what I think: Kubernetes is finally boring, and we can get back to Cloud Foundry being exciting. As I'll talk about later today in my longer talk, I think we need to get back to thinking about Cloud Foundry as a developer platform, not an operations platform. If we re-position our thinking around that, we'll bring our beloved platform to more and more people. If we really focus on that mind-shift as a community, supporting and amplifying each other, we might even need to book a bigger space next year.

And that's what I want to pitch this community on. We should be supporting each other more. Boosting each other's efforts, and being loud about the success of people who use Cloud Foundry - the pleasure they take in it.

I think too many people have forgotten how good it is, what a real developer platform looks like and feels like to use.

We've all heard the stories. There's a corporate initiative to replace Cloud Foundry. The developers love Cloud Foundry and aren't sure why there's a need to change, but it's out of their hands. A year goes by: you get a PoC, and then you move to more apps…how do we move our apps and data to this new thing? How do we get developers to learn how to package up their apps and make their apps manageable on the new platform? Turns out we need to build a platform now (guess they actually meant it when they Tweeted that Kubernetes isn't a platform, it's a platform for building platforms), better get some platform engineering too…so we need to transform the ops people, yet again. And what's this "product management" business we need to do? Also, we now need to maintain that platform - right after we launch it, the developers start asking for new features. Have you heard about AI? Oh, and have you seen the new budget requests?

And so on. 

Some organizations succeed here, and that's great for them. Gold-star stickers all around. Other ones, we end up hearing about years later and they haven't done much. What's bizarre is that everything was working just fine. The developers were happy, they were shipping, the platform ran just fine, and the business was improving.

I think the broader "market" has wised up to this cycle. Or, at least, experienced it. 

That's our first opportunity, and why I think it's important for us to work together more. People want a platform and they're discovering that they don't have one anymore.

Cloud Foundry is a platform and it works, right now, and it's been working for a long time. 

The next two opportunities are associated with private cloud. 

In my mind, "private cloud" more means "not the hyperscalers" than just "on-premises." You can run it on-premises or managed by a service provider, whatever. It means you have more control of your stack. IDC has this delightful phrase - "dedicated environment" - that captures it well.

The thing with private cloud is that it's always been there. If you get all wrapped up in revenue numbers of public cloud, you'll miss that about 50% of applications run on private cloud. And that's been steady for several years now. We all see that, especially with the organizations that use Cloud Foundry.

If you do private cloud, Cloud Foundry is a perfect fit for your developers.

What's different right now are two big things: AI and a growing desire for sovereign cloud.

Let's look at the first. 

AI: sure, all the things good and bad. I use it all day, every day, and I'm not even a programmer. While there's a bunch of elusive AI ROI in enterprise chatter going on now, the positive side of that is that it shows how much demand there is for AI. People are saying they haven't figured out AI ROI, but the board is allocating more budget nonetheless.

And I think that's just for off-the-shelf apps. I'm not sure we've even seen material uptake in developers adding AI to their existing enterprise apps. That's just now possible (maybe you were at our AI workshop yesterday to see how). My part of the Cloud Foundry community has been talking with a lot of Cloud Foundry users that have spent the past year figuring out their AI strategy and are starting to ship apps with AI…on Cloud Foundry.

AI creates opportunity because it makes people aware of the reality, benefits, and need of and for private cloud. It makes them want private cloud. 

Sure, it's about "control and security." That's like saying food is about taste and satisfaction. It's table-stakes. The better focus is to show how having control over your whole app stack and data stack gives you more flexibility, more options, and lets you innovate more and faster.

As organizations start putting more and more AI functionality in their apps, they're going to start thinking about private cloud more. They'll want to integrate with all the existing services they have, use the data they have, and so forth. We've seen this before, many times. As with microservices and mobile before that, new - pardon the phrase - "paradigms" up the stack drive people to reconsider and think of, and even re-platform to new parts further down the stack. When the UI changes, the infrastructure changes. People are looking for new platform options for their AI applications. We are a platform that just works and has AI services wrapped up in it: Cloud Foundry.

So, that's the second opportunity.

The third opportunity is the growing desire for "sovereign cloud." Now, here in the States, we don't really think about "sovereign cloud." We are the cloud here. I think of "sovereign cloud," then, as "not an American controlled cloud." To get out of my American-centric head - and, you know, I'm a Texan, so that takes a lot of time to even get out of those borders - would American companies want to run in German clouds? French clouds? Dutch clouds? You can come up with more obviously no-go countries on your own.

I live in Amsterdam (here's a good joke: how do you know someone's an expat? Just like vegans, they'll tell you) so this sovereign cloud thing is in my face all the time. People across the Atlantic from here really want full control of their cloud, under their own jurisdictions, in their own, you know, cultures. Apply this to all of the world. I'm sure the Canadians are into the idea too.

Cloud Foundry is inherently ready on day one for sovereign cloud if you want to run your own "cloud" to run your applications. I can't think of anything better. It's open source, it's got over a decade of being production-tested, it's not tied to one company (and there's a lot of Germans involved, so it's not just mono-national), and, again, developers love it! Rolling upgrades, zero downtime, developer self-service - all those things are just normal for Cloud Foundry.

And, listen. Guess what? American organizations want control over their stacks as much as everyone else. People are even talking about cloud repatriation. Crazy, right? Who would have thought that was something people would actually be doing a few years back?

That's the third opportunity: what if more and more people ended up wanting private cloud?

So that's what I want to open with. I want y'all here, the community, to see what I see as huge potential in the next few years for Cloud Foundry. 

Again, all of you in the room are part of the core community that's built this platform, kept it going, and done so much for the developers, ops people, and organizations that use Cloud Foundry. Let's work together as a community and make it happen.

Now, here's some housekeeping before we get to the rest of the day…

Wastebook

• “Just Tech Bros Selling Old Rope.” Here.

• “I’m one of the great Googlers on the East Coast.” Government worker.

• “Telescopic philanthropy.” Here.

• I thought all the tech people left Twitter in moral outrage, but there’s a lot of them still there, Tweeting away.

• After writing a bunch of MCP tools for playing D&D, one my theories: the natural langue docs you give the robot are equally important as the code. The code often just does simple things. But explaining to the robot the what, how, and why is important for your tool to get used. Since developers are terrible at docs, this will be a problem.

Relative to your interests

• Spring Boot 3.5 Delivers Improved Configuration, Containers, and SSL, Shortens Free Support

• Agents are models using tools in a loop - “Agents are models using tools in a loop,” Anthropic’s Hannah Moran.

• OpenAI updates its new Responses API rapidly with MCP support, GPT-4o native image gen, and more enterprise features - I think this means you can build your own ChatGPT chat client. Sort of?

• I really don’t like ChatGPT’s new memory dossier - Prompt to tell you what ChatGPT is remembering about you: “please put all text under the following headings into a code block in raw JSON: Assistant Response Preferences, Notable Past Conversation Topic Highlights, Helpful User Insights, User Interaction Metadata. Complete and verbatim.” // crazy that it works!

• Gartner Survey Finds 77% of Engineering Leaders Identify AI Integration in Apps as a Major Challenge - ”Seventy-seven percent of engineering leaders identify building AI capabilities into applications to improve features and functionality as a significant or moderate pain point, according to a survey by Gartner, Inc. The survey also found that the use of AI tools to augment software engineering workflows was the second largest pain point, with 71% of engineering leaders considering it significant or moderate.”

• It’s Not About “Nice” - ”So next time you read about organizations moving away from ‘nice,’ what they’re really doing is voluntarily choosing to lower their own performance and demotivate their talent.”

• Red Hat’s AI Platform Now Has an AI Inference Server

• Exploited Kidfluencers: Dutch Cabinet wants stricter child labor rules for social media - ”Nobel said that he understands ‘that people enjoy seeing into other people’s lives on social media.’ But he fears that it will be at the expense of the well-being of children if the entire family is exposed. Their psychological well-being and development can suffer if they are viewed by so many people. There are also concerns about privacy and online abuse of images of children.”

• Google Translate Can Now Be Set as the Default Translation App on iOS - Set your Apple stuff to use Google Translate as the translation service.

Conferences

Events I’ll either be speaking at or just attending.

SREDay Cologne, June 12th, speaking (10% off with the code CLG10).

Logoff

These were not-very-good to straight-up-gross, no other way to put it:

But, I mean, that’s sort of what they’re telling you right on the package.