Does kuberbetes make application development and delivery better? - 2024 Watch List

I don’t like annual predictions. Thankfully, no one’s asked me to do a big post about them this year, but usually I’m asked to do something formal. I always end up just predicting the same thing - predictions turn a lot more into hopes and dreams. I do like flagging things to pay attention to, and especially in a sort of “will this finally happen, or will we start re-calibrating our expectations from the apex of inflated expectations.”1

I’ve got three things to watch in 2024, here’s the first one (predictable for many of you, dear readers, no doubt).

Does kuberbetes make application development and delivery better?

For years we've read that kuberbetes is powerful but complex. We've also heard that application developers are not actually supposed to use it directly, let alone build all the app platforms and tools they need to make kuberbetes easy to use.

So, who is kuberbetes for, and are those people getting value out of it versus just using a standard VMs or more exotic things like serverless?

Survey results have been sort of good, people get some benefits. If you look at estimates for how many apps run in containers to get a sense for how much kuberbetes usage there is, you see something like 15% to 20% globally. If kuberbetes is going to be universal, it needs to expand to 50%, maybe even 70% of workloads.

Is that what the community wants for it, or is kuberbetes meant to be further down in the stack than something that application developers ever touch?

I try to follow the original intention of the Kubernetes Krew, which is, developers were never really meant to be exposed to it:

Well, I don't know how many of you have built Kubernetes-based apps. But one of the key pieces of feedback that we get is that it's powerful. But it can be a little inscrutable for folks who haven't grown up with a distributed systems background. The initial experience, that 'wall of yaml,' as we like to say, when you configure your first application can be a little bit daunting. And, I'm sorry about that. We never really intended folks to interact directly with that subsystem. It's more or less developed a life of its own over time. Craig McLuckie, SpringOne 2021.

If that’s true, when it comes to application the developers, I think the most valuable thing the CNCF (as an aggregate of the kubernetes kommunity) can do is define the application architectures, patterns, and practices application developers and architects need to use to be successful with kubernetes. And, you know, that might be done enough, in which case it’s just the endless slog of developer relations and thought-leadership to teach everyone and keep it rolling and updated.2

I haven’t programmed in a long time, so my historic examples are old, but the one I always think of are all those J2EE patterns and books. There was an underlying, infrastructure layer for Java (trying to re-write the CORBA death-star to get a better distributed application model), but if an application developer just made up their own way to use that underlying infrastructure, it all tended towards unique variation. Once every application follows its own design and practices, you have to spend a lot of time learning how to manage each one, and you typically lose any organization-wide benefits you were hoping to achieve: hopefully I don’t need to make the case for standardization?

Similarly, if we don’t start getting application developers (and architects) to use standardized design and patterns for their apps that are running on Kubernetes, we’ll get a bunch of unique garbage, er, sorry, variation. I feel like this happened with OpenStack: the conversation never evolved a new branch from operations into application development.

Good luck to the kommunity in 2024!

Got Java Apps? Stay on-top of security patches, upgrades, and out of support apps

We have a new tool at Tanzu,3 the Spring Health Assessment. You can use it for one project for free by uploading your dependency info from maven. It’ll create a report for telling you the, well, health of your app (see report example below). At a larger scale, we’ll work with you to do this at enterprise scale, you know, across hundreds and thousands of apps.

There’s all the usual stuff of wanting to keep your Spring apps secure and updated, but also you can figure out when your older version of Spring (and other Java components, I think) are no longer supported by the community or otherwise.

Report preview

I think there’s a lot of use for this tool ahead because (1) recent changes in JVM licensing costs, (2) Spring Framework components rolling out of community support, ands, if I remember and understand correctly, (3) there’s going to be more frequent releases of the JVM from now on. And, of course, there’s the never ending slog of modernizing applications.

There’s thousands and thousands - and even more thousands! - of Spring apps out there in the world and keeping on-top of all of them is a chore. Check out our tool to start automating all that.

For even more, my team-mate DaShaun has a demo and discussion of the Spring Health Assessment, fresh from the streams.

Anyhow, look, if you’re using Spring, you should check it out. And if you want to apply it at a larger scale, you should email me, ‘cause the team is interested in how it goes for you.

Relative to your interests

  • Austin Tech Scene - Tracking events in Austin.

  • Coding at Google – text/plain - File under “probably should read this.”

  • The Best Return-to-Office Policies Aren’t One-Size-Fits-All - Good summary of the return to office thing. Basically, workers hate it and say it’s a waste of their time, but management is freaked out that workers are slacking off. “Oftentimes, it’s more of a preference for leaders in organizations. And when they don’t have visibility into employees’ work lives, those concerns certainly come about. And again, our research doesn’t suggest that there’s a fire burning and we’ve got to bring folks back into the office. But it certainly has been the way that many organizations have responded to some of these concerns about productivity and collaboration in some of those areas.” What you have here is a classic anti-pattern of work: the people who are not doing the work are determining how their work is done. There’s a principle from lean that says the people closest to the work should define how the work is done - they are experts in it, after all, and will spot ways to improve it day to day. That applies here as well. What management needs to do is establish the strategy, context, and principles, even metrics, that those employees follow (I don’t know, insert all that military, small teams Make Your Bed Everyday productivity cult stuff from the 2010’s), and then let the workers define how they work. This is about more than morale, it’s about productivity. Anyhow: like any worker, management needs something to do, and the easiest thing to do is to inspect workers. It’s at lost harder to build a system where workers are autonomous, but a lot better than just being in a bunch of meetings no one wants to be in and that everyone except the boss thinks - knows - is a waste of time for them. Speaking of metrics: I’ve yet to see any ongoing metric-driven model that tracks any of this in office; remote work stuff. That’d very something helpful! Even more damning, in that interview, you can see in the research that management has to come up with - invent - reasons to make being in the office worth it to employees. Applying that lean principle (the people who do the work define and optimize how the work is done): if the workers think that being in the office is the best way to do the work, they should go into the office. Otherwise, you have to budget for starting up hot-dog roasts every Thursday, or whatever, to make being in the office worth it.

  • A Simple Hack to Help You Communicate More Effectively - “Break down your message into three parts: What? So what? Now what?” - I like to do these in reverse order to get people’s attention, but I’m also impatient.

  • Silicon Valley runs on Futurity - Story (or “vibes”) drives a huge part of tech company valuation.

  • Rebuilding Trust and Breaking Free From Trust Proxies and The Swirl - “1. People underestimate how powerful it can be to take responsibility for past behaviors. In many cases, that is all people wanted to hear. They wanted to hear that leaders could accept responsibility for being part of the swirl. 2. Don’t underestimate the ritualistic shedding of past wrongs. Many cultures use rituals to shed bad mojo. In this case, it could be a meeting to air the dirty laundry, put it in a box, and burn the box.”

  • 100 tiny changes to transform your life: from the one-minute rule to pyjama yoga - “’By asking: “If it’s irrational, why do it?” I stop doing it’” - this is the problem, though: how do you get yourself to actually do the rational act of following the rational? Perhaps it’s something like: you have to trust that things will be better. If you still end up unsatisfied and struggling after doing the rational, what was the point? Instead you can live in the moment and have a better chance of some seconds of happiness and relief.

  • Paying Netflix $0.53/h, etc. - This is a fun way to look at streaming service pricing.

  • Tetragrammaton - Rick Rubin interviews Marc Andreessen - This is an example of where a 3+ hour podcast works well, and is better than a 30 minute one. Normally, you just get tiny context-free chunks of Marc Anderson, and he comes off as a rich guy sort of floating above human moral concerns and everyday life. He doesn’t exactly get grounded in this interview - he’s ardently techno-positive and defensive about hating the tech industry. But, you at least get to understand his thinking so much more than you would in a WSJ column, or even a standard 30 to 40 minute podcast interview.

  • How bad are search results? Let’s compare Google, Bing, Marginalia, Kagi, Mwmbl, and ChatGPT - Testing out search across search engine and AIs. Not good!

I have a backlog of more links to share. Rather than overwhelm you, dear reader, I shall take the risk of saving some for next time, more aged as they may become. Do you like a long list, or a shorter one?

Garbage Chairs of Amsterdam, New Year’s in Austin edition.


  • “‘Failures’ are usually just incremental steps.” Here.

  • Related: I’m too busy being upset at myself fucking up yet again to care about you fucking up yet again.

  • The Dutch like to use cake as a pie-filling.

  • “[A] Minotaur’s labyrinth costs nothing to enter. What we need is string to find our way out again.” Here.

  • “Yeah, we know how magnets work. But they’re still incredible.” Here.

  • “Fecosystem.” Here.

  • “The problem is, everyone we know really only has stones.” RoTL #521.

  • “Do you believe that? I’m not asking if you can argue it, I’m asking if you believe it.” Here.

  • “Since having kids, deadlines are a mere illusion, shadows of their former selves.” Here.

  • “Temporal illegibility.” Here.

  • It was only dark when it was night.

  • For me, the long winter vacation is about relaxing, “recharging.” But the real trick is figure out how to continue that chilled out mindset once back in the fray of life. I rarely can.

  • “We’d be showing them a pipeline of pipelines…redacting things that we don’t want to expose.”


I have a new theory for these little videos. The views on them are so uneven - that one above has over 1,000 views, while others have, like, 79. I think what you’re supposed to do is (1) only post one a week or so, and, (2) re-post the ones that get low-views, maybe even varying the videos a little? Seems like a lot of work, but not really if you have a good workflow and a tool like Descript.

Also, we should get lunch - hahahah…jokes.

(Also: Kenji is so inspirational as a video maker. His content, of course, is great, though it has nothing to do with my content [except the chip tasting and Nutella nonsense, sort of]. What’s inspiring is that, well, if you make videos, you can imagine his workflow. I’m sure each video takes a long time, but you can totally see how he’s optimized a one-man [I’m guessing] workflow from content planning, filming, editing, post-production and posting. You can tell he has a “stop planning and just get out the camera and start filming” attitude that you have to have for this kind of work. It’s also a good example of how much “pro-am” equipment [I think Go Pro’s and iPhones] make it possible for individuals to do good, important, etc. content. In all the nonsense about TikTok destroying the world, or whatever us old people worry about, it’s easy to overlook how this equipment, contemporary computers with video editing and rendering horse-power, mics [you can see him using a Rode Wireless Go here, one of the best], and video production have “democratized” video production. There may worrying about other types of content [text, music, etc.] being shittified by the Internet, but as a consumer, video is so much better than when there were just three channels, the PBS and the UHF channel. Also, I rewatched my goofy how to eat videos recently - I mean, first, they’re great! And, second, it’s adorable to see my kids grow up over the course of three years. Also, gravy is always a winner.)


One of the key things people miss (or mis-use?) from [the hype cycle]( is the fall from the peak of inflated expectations. People feel let down when a technology fails to live up to its promise, they think less of the technology. The point of that fall doesn't have anything to do with the goodness of that technology, it's all about the failure of humans to know understand the technology and the failure of boosters to moderate claims about the technology. When a technology slides down to the trough of disillusionment, it means that people's perceptions are more _correctly_ set. People being people (overreacting), the slide goes down too much, and it eventually re-sets, just a little bit up, hence, the Plateau of Productivity. Wow! Maybe there's something to be typed up here with mis-uses of the hype-cycle. The other dangerous use is to think t[he diffusion of innovation curve]( is highly related - can be overlayed even - on the hype-cycle. And then you throw in [all the Disruption lore](, and you're off the races of chaos-strategy (applying the lessons of steel mill business strategy to software, it turns out, is dangerous).


One of the principles of this newsletter is that I don’t have time to write the perfect thing. I bet there is such work going on - maybe it just needs to be spread more widely. I don’t know, get Martin Fowler to curate a series of blog posts that turn into a book, or something ¯\_(ツ)_/¯.


After the Broadcom acquisition, my organization is now called "Tanzu by Broadcom," or, just "Tanzu." This is how most people would refer to it anyway, but it's noticeable sign throwing that it's no longer "VMware Tanzu." We're, you know, our own thing now, within the conglomerate of Broadcom.,, @cote,,