Sign in

Mark Seemann

@ploeh.dk
211 followers 2 following 243 posts
PostsRepliesMedia
Mark Seemann @ploeh.dk · 26/09/2026
Yes, it was quite difficult. Thank you, both for the response and the exercise.
110
Mark Seemann @ploeh.dk · 25/09/2026
I'll be in London November 23-27, and could probably swing by for half a day or so if you think it would be useful for your organisation. blog.ploeh.dk/hire-me
blog.ploeh.dk
Hire Me
Would you like help with software design, TDD, code review, process, or programming? If so, I may be able to assist.
000
Mark Seemann @ploeh.dk · 24/09/2026
@hillelwayne.com I'm embarking on a series of blog posts tackling your PBT / Example Testing comparison problem. Is there any context related to that gist? I mean, why did you originally post it? I tried asking Brian Marick back when you were taking a break: mastodon.social/@ploeh/11600...
mastodon.social
Mark Seemann (@ploeh@mastodon.social)
@marick@mstdn.social Sorry for the extremely belated reply, but in https://www.crustofcode.com/arriving-at-properties/ you refer to a problem posed by Hillel Wayne. Do you know anything about the cont...
100
Mark Seemann @ploeh.dk · 16/09/2026
New article: On learning programming in an age of LLMs. blog.ploeh.dk/2026/09/16/o...
blog.ploeh.dk
On learning programming in an age of LLMs
Open answers to a reader's letter.
032
Mark Seemann @ploeh.dk · 08/09/2026
New article: Worse is better: C# versus F#. blog.ploeh.dk/2026/09/08/w...
blog.ploeh.dk
Worse is better: C# versus F#
C# is a great programming language, but F# is better.
041
Mark Seemann @ploeh.dk · 26/08/2026
Among many other things, I'm also interested in helping improve legacy code bases (C#, F#, or Haskell; possibly Python), as this is typically a really difficult problem. So get in touch if you think I can help.
045
Mark Seemann @ploeh.dk · 24/08/2026
The medical open-source project that hired me to help seems to have unexpectedly run out of money, so it looks as though I'm available for hire until November. Get in touch if you think I can help your software engineering organisation. blog.ploeh.dk/hire-me
blog.ploeh.dk
Hire Me
Would you like help with software design, TDD, code review, process, or programming? If so, I may be able to assist.
031
Mark Seemann @ploeh.dk · 18/08/2026
New AI DevOps Podcast episode, this time with yours truly as a guest, in which we discuss code quality related to LLMs, among other things: azuredevopspodcast.clear-measure.com/mark-seemann...
azuredevopspodcast.clear-measure.com
AI DevOps Podcast: Mark Seemann: AI Code Quality - Episode 415
Mark Seemann is a self-employed programmer and software architect based in Copenhagen, Denmark, and the author of Code That Fits in Your Head and Dependency Injection Principles, Practices, and Patter...
010
Mark Seemann @ploeh.dk · 18/08/2026
ISTR seeing a figure like my poor attempt at a reproduction posted here. It was, of course, much nicer done, and also IIRC had more details and colour. Can anyone help me track down the original?
000
Mark Seemann @ploeh.dk · 15/08/2026
On that note 😉 blog.ploeh.dk/2026/03/30/p...
blog.ploeh.dk
Programming languages for AI
Which programming language is best suited for LLM-based generation?
000
Mark Seemann @ploeh.dk · 14/08/2026
People have their ways; I do, too. This is all good. The thought experiment here is: What if you're no longer writing any code? In that case, is syntax highlighting and code completion still relevant?
100
Mark Seemann @ploeh.dk · 13/08/2026
New article: Why use ORMs if LLMs write code? blog.ploeh.dk/2026/08/13/w...
blog.ploeh.dk
Why use ORMs if LLMs write code?
The return of SQL?
120
Mark Seemann @ploeh.dk · 23/07/2026
New article: Does code quality still matter? blog.ploeh.dk/2026/07/23/d...
blog.ploeh.dk
Does code quality still matter?
I honestly don't know, but I guess we'll find out.
022
Mark Seemann @ploeh.dk · 21/07/2026
Min kone fik en ide, der er så god, at jeg var nødt til at dele den. Så nu har vi oprettet vores første borgerforslag: Fjern eller sænk moms på økologiske varer. Denne ide er lidt af et columbusæg, der løser mange problemer på en gang. Støt forslaget her: www.borgerforslag.dk/se-og-stoet-...
011
Mark Seemann @ploeh.dk · 17/07/2026
My 14th consecutive Microsoft MVP award: How lovely 😊
030
Mark Seemann @ploeh.dk · 07/07/2026
It's not? I admit that I've never measured either, as they seem to be more than adequately efficient. It of course depends on context, but I've never noticed (de)serialisation to be a bottleneck in the systems I've worked on. Is the difference significant?
110
Mark Seemann @ploeh.dk · 06/07/2026
New article: Worse is better: JSON versus XML. blog.ploeh.dk/2026/07/06/w...
blog.ploeh.dk
Worse is better: JSON versus XML
JSON is not a good data-interchange format.
132
Mark Seemann @ploeh.dk · 29/06/2026
New article: Worse is better. blog.ploeh.dk/2026/06/29/w...
blog.ploeh.dk
Worse is better
The most popular technology may not be the best.
010
Mark Seemann @ploeh.dk · 16/06/2026
Coming off a successful undertaking, I have time for new endeavours. Potentially full time until end of August, and perhaps a few months more. Part-time with no definite end. I may be particularly interested in improving legacy code. blog.ploeh.dk/hire-me
blog.ploeh.dk
Hire Me
Would you like help with software design, TDD, code review, process, or programming? If so, I may be able to assist.
030
Mark Seemann @ploeh.dk · 15/06/2026
This was the most interesting episode in a long time. Worth listening to, despite the bad audio quality. Bear with it; the content is good.
020
Mark Seemann @ploeh.dk · 10/06/2026
Are you using LLMs to generate so much code that reviews or other forms of quality assurance have become a bottleneck? I may be able to help. blog.ploeh.dk/hire-me
blog.ploeh.dk
Hire Me
Would you like help with software design, TDD, code review, process, or programming? If so, I may be able to assist.
020
Mark Seemann @ploeh.dk · 31/05/2026
Is there anything you would like to see?
100
Mark Seemann @ploeh.dk · 30/05/2026
Indeed, I think so too. But then, my blog already has 800+ articles, many on the topic of software engineering. Does the world need more? Perhaps. Does it want more? It doesn't feel like that at the moment.
110
Mark Seemann @ploeh.dk · 29/05/2026
As impressive and occasionally useful as LLMs are, the hype means that I'm regularly disappointed when I try to get it to do something useful. Latest example: I want some LLM to summarise a podcast episode. I can even provide a transcript. So far, no such luck. Suggestions?
200
Mark Seemann @ploeh.dk · 29/05/2026
Thank you for asking. I'm busy, and simultaneously somewhat disenchanted with the direction software development is taking, so no longer inspired. I sent out a longer explanation to blog supporters, but for now, I plan to go back to a once-a-month publication cadence.
110
Mark Seemann @ploeh.dk · 28/05/2026
Recently, I've been feeding LLMs photos of hand-drawn diagrams, asking them to turn them into TikZ drawings. This seems to be at the frontier of what LLMs can do. It's not an entire fiasco, but on the other hand not going particularly well, either.
111
Mark Seemann @ploeh.dk · 05/05/2026
Good point. Some of the pains you mention were always a part of the game, but I've managed to find workplaces with only a minimum of that going on. Mainly because there were still organisations where correctness was more important. I hope some of those places remain.
010
Mark Seemann @ploeh.dk · 04/05/2026
New article: Motivated reasoning. blog.ploeh.dk/2026/05/04/m...
blog.ploeh.dk
Motivated reasoning
Most of my concerns about AI are probably irrelevant, but what if one of them is not?
120
Mark Seemann @ploeh.dk · 01/05/2026
I think I understand the gist of what you're writing, although it sounds backwards to me. I thought that fuzzing was mostly unguided random garbage creation, with simple crash/didn't crash observations. But it's been decades since I had a system subjected to that, and someone else set it up.
110
Mark Seemann @ploeh.dk · 01/05/2026
Yes, that sounds like a really good idea. It's possible that I misunderstand what you mean, but I usually call that 'encapsulation', which is a topic I've had a lot to say about over the years.
000
Mark Seemann @ploeh.dk · 30/04/2026
Could you elaborate on your 'inserted assertions' comment?
210
Mark Seemann @ploeh.dk · 29/04/2026
I've mostly written systems in .NET (F# and C#) and Haskell, and never found any value in fuzz testing. Should I? I've always seen fuzz testing as a tool to reveal buffer overruns or similar, which seems unrelated in languages that don't do pointer arithmetic. Am I missing something?
320
Mark Seemann @ploeh.dk · 20/04/2026
New article: In defence of bureaucracy. blog.ploeh.dk/2026/04/20/i...
blog.ploeh.dk
In defence of bureaucracy
Objectivity, meritocracy, high bus factor. What's not to like?
041
Mark Seemann @ploeh.dk · 13/04/2026
New article: Secret agentic AI. blog.ploeh.dk/2026/04/13/s...
blog.ploeh.dk
Secret agentic AI
A scenario.
110
Mark Seemann @ploeh.dk · 10/04/2026
I find Robert C. Martin's definition useful: "Abstraction is the elimination of the irrelevant and the amplification of the essential", from 'Designing Object-Oriented C++ Applications Using The Booch Method', ch. 00.
000
Mark Seemann @ploeh.dk · 09/04/2026
In both 2023 and 2024 I gave keynotes related to AI and software development: youtu.be/bLpwNWWs5EY?... youtu.be/X_SsnmZ4Dlc?... Since they are of a general nature, I still find them relevant in 2026. Let me know if you need a keynote for your conference. blog.ploeh.dk/hire-me
youtu.be
IT Days - Epistemology of software - Mark Seemann - Programmer and Software Architect
YouTube video by Today Software Magazine
030
Mark Seemann @ploeh.dk · 09/04/2026
Zawinski's Law strikes again 😉
020
Mark Seemann @ploeh.dk · 07/04/2026
Lately, I've observed that engagement on all platforms have all but dried up. Is this a general phenomenon, or am I just writing for no-one? If you still read and get value out of my blog, please let me know. You could, for example, support it: blog.ploeh.dk/support
blog.ploeh.dk
Support the blog
Ploeh blog is free, and always has been. Keep it free.
022
Mark Seemann @ploeh.dk · 06/04/2026
New article: The hailo effect. blog.ploeh.dk/2026/04/06/t...
blog.ploeh.dk
The hailo effect
LLM friendliness does not entail competency.
040
Mark Seemann @ploeh.dk · 30/03/2026
New article: Programming languages for AI. blog.ploeh.dk/2026/03/30/p...
blog.ploeh.dk
Programming languages for AI
Which programming language is best suited for LLM-based generation?
020
Mark Seemann @ploeh.dk · 29/03/2026
Smilodon?
000
Mark Seemann @ploeh.dk · 24/03/2026
Yes, and my current view is that this question will be decided in court. I am not a lawyer, though, so I am open to other viewpoints. (I try to be that, in general.)
110
Mark Seemann @ploeh.dk · 23/03/2026
New article: Will you go to prison for an AI? blog.ploeh.dk/2026/03/23/w...
blog.ploeh.dk
Will you go to prison for an AI?
Who is liable for code written by LLMs?
120
Mark Seemann @ploeh.dk · 10/03/2026
Indeed, but even in the first case, I think there is more to say. I plan to follow up on this later.
020
Mark Seemann @ploeh.dk · 10/03/2026
ε, usually 😉
000
Mark Seemann @ploeh.dk · 09/03/2026
New article: Which of these languages are best for AI? blog.ploeh.dk/2026/03/09/w...
blog.ploeh.dk
Which of these languages are best for AI?
A rhetorical poll.
111
Mark Seemann @ploeh.dk · 06/03/2026
#repostfriday Keeping cross-cutting concerns out of application code. blog.ploeh.dk/2024/09/02/k...
blog.ploeh.dk
Keeping cross-cutting concerns out of application code
Don't inject third-party dependencies. Use Decorators.
000
Mark Seemann @ploeh.dk · 05/03/2026
Worth mulling over: www.rockoder.com/beyondthecod...
rockoder.com
Cognitive Debt: When Velocity Exceeds Comprehension | rockoder
A systems analysis of how AI-assisted development creates a gap between output speed and understanding, and why organizations cannot see it happening.
020
Mark Seemann @ploeh.dk · 04/03/2026
That is a really good question 🤔 www.rockoder.com/beyondthecod...
rockoder.com
Cognitive Debt: When Velocity Exceeds Comprehension | rockoder
A systems analysis of how AI-assisted development creates a gap between output speed and understanding, and why organizations cannot see it happening.
000
Mark Seemann @ploeh.dk · 04/03/2026
Other readers are showing their appreciation of ploeh blog by supporting it. You, too, can do that. blog.ploeh.dk/support
blog.ploeh.dk
Support the blog
Ploeh blog is free, and always has been. Keep it free.
000