Martin Peck

Our users asked, we delivered. 🚢 🗂 Structured Generation Type /structured generation to get output in structured JSON format.  👧🏻 New Models • Claude 3.5 Sonnet from Anthropic • Gemma 2 9B form Google Try them out and post your apps in the comments! (see also what Robert Chandler did with Claude 3.5 in a few minutes) 👁️‍🗨️ Follow Sub-Prompt Generation (see video) Click a running prompt to see how it generates, then return to the parent prompt. Check the Runs logs for more details.

10

3 Comments

Earlier this week, I got some lovely feedback on Learn Go with Tests (https://lnkd.in/exspPfsg) The best part is that it shows that once you get past the tutorial stage, you start to see the real benefits of TDD. So many people are stuck in the mindset that TDD is about test coverage, but the writer of this feedback has read my material and made the leap into the realm of productivity with TDD. It was very gratifying to read it. You can find the feedback here. https://lnkd.in/enCiKWqV #golang #tdd #nothumblebrag

37

3 Comments

This week we revisit the Jan 2022 article on What Influences Application Development Costs - 42 Interactive https://bit.ly/3nDdNhQ The primary factors influencing the cost of developing an application. Key elements include project complexity, platform type and quantity, backend infrastructure and APIs, security requirements, and quality assurance. Each factor affects the overall budget, with the complexity and functionality being the most significant cost drivers. Native apps for multiple platforms are generally more expensive than cross-platform solutions. Proper planning and understanding these variables can help manage costs effectively. #AppDevelopment #CostFactors #ProjectComplexity #PlatformChoice #BackendInfrastructure #AppSecurity #QualityAssurance

2

Hard to believe the CQRS pattern is over 21 years old and yet it is still widely misunderstood by developers writing code *today*. This is because its introduction in Eric Evans’ book, Domain Driven Design, was focused on the space *between* a presentation layer and repository layer and how CQRS fits there, and it was not clear *then* that REST (Representational State of Resource) would eventually become *the* standard in presentation layers. Such a close up view on CQRS left many naive developers like myself to design CQRS implementations around REST designs, and not the other way around, leading many to presume CQRS is either just obsolete boilterplate or a slippery slope into RPC over HTTP (in other words, anti-REST.) Let me keep this simple: * If your CQRS objects mirror your REST endpoints, that could be a “design smell” indicating you’re not getting your bang for your buck on CQRS. * If you find yourself setting up REST endpoints *before* creating CQRS objects to serve them, that could be a design smell. * If your CQRS objects know how to present themselves *directly* to the request or, you might have a design smell. If you have concerns about how REST or CQRS is being implemented in your shop, please reach out! Happy to discuss your situation personally or even speak to your local community about tips and tricks to get the most out of both. The consequences for a sub-par REST or CQRS layer are tremendous: from significantly wasted developer energy to tightly-coupled software layers, it all adds up to the same price: incrementally higher dev costs coupled with slower velocity.

22

Optimize your performance and reduce the cost of running #AI workloads by powering them entirely with CPUs. With Aible’s serverless computing and the efficiency of #IntelXeon processors, you can improve your end-to-end process. Read the release to learn how to tap into enhanced AI. https://intel.ly/3XG7cpo #IAmIntel

1

I just updated my first table format comparison blog post to fix some missing details of Apache Hudi. https://lnkd.in/dwSUS6qS

5

Prof. Felix Hovsepian and I share concerns about the state of #software development, specifically an alarming disappearance of actual #engineering — which requires a multidisciplinary approach that includes ... 🔹understanding user requirements 🔹design 🔹implementation 🔹testing 🔹debugging 🔹documentation 🔹deployment 🔹updates (fixing bugs, adding features) 🔹support 🔹training 🔹marketing Recently, I met with Noah, a #SoftwareEngineering student who is an intern for the NASA - National Aeronautics and Space Administration Europa Clipper mission. We began an engaging, thought-provoking discussion about the state of #software development — comparing how it was done in the 1980s-90s with how it's done today. 🤔 https://lnkd.in/gKzzQFuB Noah and I plan to continue this discussion. 🙂 #SoftwareEngineering #ExperienceMatters #MentorsMatter

26

7 Comments

Absolutely wrong. Second only to driving the clarity of your code, tests are notifiers of change. Any change. Yes, behavior is important, but so are the internal semantics at every layer of abstraction in your system. If your tests aren't blowing up when you make a change - any change - you are most definitely doing it wrong.

2

11 Comments

📣Build 101 of VRGS 3.1 is now available. This week's build improves the performance of attribute generation on meshes and point clouds, with a speed improvement of up to 7x in tests. 😎🎉 To get started using attributes see this quick tip https://lnkd.in/eVtKvPXY

10

It's not all non-symbolic AI... I presented today to Accord Project Technology Working Group on recent enhancements to the Template Engine and how TypeScript expressions within templates are compiled and evaluated. Accord Project uses TypeScript over Concerto types (with Concerto also compiled to TypeScript) to provide type-safe logic/calculations for templates; ensuring that a wide array of programming errors are caught at template compilation time, not at runtime. #typesafety #soundness #typescript #templating #legaltech

17

2 Comments

Have you used Wardley Mapping in your teams? Maps can be used to help you find a route to your destination. Many of our other diagrams in technology either document things as they are, or describe an ideal future state. Wardley maps can help you leave that 2 hour meeting with clear next steps. They also help us ride the architect elevator up a few floors and tell compelling stories to our business stakeholders about why all this Serverless modernization stuff matters to them. In the new episode of Serverless Craic, David Anderson, Michael O'Reilly and Mark McCann talk about their experience after a decade of using Wardley Maps to get buy in, and go on to execute some of the most successful Serverless First transformations anywhere on AWS. Well worth a listen, and I’m looking forward to part 2.2.

23

3 Comments

📚 Ever wanted a modern, open-source solution for your library's integrated management system? Meet RERO ILS, a large-scale Integrated Library System (ILS) designed for library networks! 🚀 It includes most standard modules such as circulation, acquisitions, cataloging, and more. With a web-based public and professional interface, RERO ILS is perfect for libraries looking to provide smooth access to resources while enjoying the benefits of open-source software. Being a developer and content creator, I've had the pleasure of exploring RERO ILS, and I'm impressed by its powerful features: ⚡ Fast and ergonomic interface 🌍 Consortial model built for library networks 📚 Cataloguing and serial management 🗂️ Data interaction and openness via JSON-formatted bibliographic data and API 🔍 Online catalog with user-friendly search and filtering functions RERO ILS is built with popular technologies including Invenio, Flask, Angular, ElasticSearch, and Ngx-Formly. With active development since 2017 and a dedicated team from Switzerland and Belgium, RERO ILS is continually improving to fulfill users' Let's discuss RERO ILS and its potential for your library. Share your thoughts, questions, or experiences with me here in the comments! #REROILS #IntegratedLibrarySystem #OpenSource #LibraryManagement

5

Just a reminder, that my course of the foundations of messaging, Practical Messaging, is running next at NDC London in 2025. - Day 1: Messaging Patterns. We will cover the fundamentals of using messages for interoperability in distributed systems - Day 2: Conversation Patterns. How to compose the building blocks of Day 1 into Event Driven Architectures. Takeaways: how to write messaging code; how to compose event driven architectures See the link in the comments.

30

2 Comments

Observability is not just for app behaviour. It can be used to open the black box of costs. I didn't see this for years. When engineers and other technical folk in team don't have visibility on costs... ...that can be solved with observability techniques. Instrumenting which costs come from where. Then showing them in graphs, allowing teams to drill down on costs with a series of "why" questions. My vision for Minor Cloud Costs - my latest product idea - is to present the most often used and most useful information without any complex filters. We should be able to see, at a glance, within 15 seconds a day: * The most unpredictable costs * The largest costs ...and drill down into the answers to questions like: WHY is this cost being accrued? HOW can I stop this happening? WHAT is being used and what is not? DM me if you'd like to share your experiences of managing costs in tech orgs. No selling from me. Pinky promise. I just want to understand the problem better.

6

Building trust and fostering a sense of safety within software engineering teams is essential for success. Graeme shares how setting clear expectations and reducing the number of "in-flight" tasks helped his team focus on overall productivity. While some initially felt their individual output slowed, the bigger picture revealed a significant boost in team efficiency and collaboration. His message: it's not just about doing more—it's about doing it better, together. Curious to see how? Watch the full video on our YouTube channel! #SoftwareEngineering #TeamCollaboration #ProductivityBoost #AgileDevelopment #EngineeringLeadership #TechInnovation #TeamEfficiency #WorkInProgress #LeadershipTips #ProjectManagement #DevTeamSuccess #EngineeringProductivity #TeamCulture #SoftwareDevelopment #WorkplaceSuccess #TechLeadership #TeamDynamics #EfficiencyMatters #EngineeringExcellence #InnovationInTech #AgileMindset #LeadershipInTech #EngineeringManager #TeamTrust #TechCommunity #ProgrammingLife #CodeCollaboration #TeamGrowth #DevLeadership #TechTalk #ProductivityHacks

2

In this article I would like provide some high-level guidance for domain (data) modelling. This article is for you if you are relatively new to domain or concept modelling. It uses Accord Project Concerto but the principles should apply to many modelling languages. #datamodels #modelling #concerto https://lnkd.in/dxgbXdkZ

7

4 Comments

What is HiGHS? How is it used? And what’s next for this open source project? We revisit a conversation and Q&A hosted by Ryan O'Neil with Ivet Galabova and Julian Hall from HiGHS to find out: https://hubs.la/Q02RfKCN0 #orms #optimization #decisionscience

32

5 Comments

Top tip for "No Code" integration solutions: Only use them if it can be done internally (i.e. someone in the business is has the technical ability to build and manage them) and the monthly platform bill is < £500. If you are paying a third party to build no code integrations - you may as well get them custom built so that you avoid the high ongoing fees and scalability issues that come with a no code platform. As the business scales you'll reach a point where it is more cost effective to do this anyway.

4

SLMs - Small Language Models I saw Microsoft CTO Kevin Scott introduce this term at their Build developer conference last week. Are they trying to make “SLM” a thing? In any event, their point is well taken. An AI model that’s less capable than a big-iron, frontier model, can be very useful nonetheless. A smaller model can be efficient and versatile: ✅ Run Locally: operate on a phone, laptop, clothes-pin, or digi-glasses ✅ Reduce Latency: provide quick response for good UX ✅ Enhance Security: keep personal data on your device Microsoft curated a short list of notable SLMs: Phi, Gemma, Mixtral, and Llama. Quality, capability, application, and choice are growing rapidly! Please practice with AI every day - it’s the real deal. Incorporating it into your daily routine will help you gain an edge among your competition and your peers. Tell me how I can help you. Let’s do something awesome together!

5

2 Comments