DEV Community: Jasper Blank

Teun

Jasper Blank — Thu, 03 Sep 2026 09:20:12 +0000

Teun

I ask mine about everything. Courgettes, the mole on my back, whether to take the ring road. In eleven years I have never asked it about Teun.

People find this strange, but they find the wrong part strange. They think the strange part is Teun, who doesn't use his at all. His sits in the flat doing nothing, the way other people have a piano. It's on the shelf above the fuse box. I dust it.

I price things for a living. I price how long things last. Roofs, hips, container ships, marriages. You'd be surprised how similar the curves are.

On Tuesday I priced a bridge. They wanted sixty years. I gave them thirty-one. They took it badly. People always take it badly. They think a number is an opinion you can argue with.

Mine found eleven hundred euros the state had owed my mother since 2031. It read the lease on the Zuilenstraat flat and made them strike clause 14. I wouldn't have understood clause 14. It does my mother's prescriptions. It found the leak under the bath before there was a stain. When the neighbour's tree came down on the shed it wrote the claim, and I read the claim, and it was better than mine. I do this for a living.

When Teun's father was ill it gave me the four treatment options with the numbers on them. I sent them to Teun's sister. She called me crying and said she didn't know how to thank me. I said it was nothing. That was true. It took nine seconds.

Teun puts his hands on people. That's the whole job. Three times a week for four months, and then a woman who couldn't lift a shopping bag can lift a shopping bag. There's no faster way. Nobody has ever found one.

There's a woman called Ans who comes Tuesdays. She's seventy-something. In September she couldn't reach a tin on the top shelf. In January she could. That's it. That's the whole thing that happened.

He's proud of this in a way he'd deny if you said it out loud.

Everybody asks. My mother asked about my father in 2028. Sixty-one per cent. She stayed anyway and tells it at parties. Wies runs everyone she meets before the second date and says it's hygiene, like washing lettuce.

She asked me once, at this table, whether I'd done Teun.

"No," I said.

She waited. People always wait. They think no means not yet.

"You should," she said. "Then you'd know."

"I know," I said. That wasn't an answer. She let it go, because she's kind.

Last year at a wedding a man I didn't know told me he'd run his own children. Both of them. Eleven and nine, he said. He leaned in when he said it, the way people do when they want you to tell them it's fine. Then he ordered another drink. He didn't look sorry. He looked relieved, which was worse.

I could ask in four seconds. That's the part nobody understands. It isn't difficult and it isn't expensive and nobody would know. It's four seconds, twice a day, for eleven years.

In February Teun came home and said he wanted to buy into the practice.

"Twenty years," he said. "They'd want both names on it."

"I know how loans work."

"I know you know. That's why I'm asking you and not it."

He meant it as a joke. He's not a cruel person; he just doesn't know which things are made of glass.

So I asked. Not about him. About the loan. That's an ordinary question. It's the most ordinary question there is. I ask it about loans for strangers forty times a day and I have never once thought about it afterwards.

It came back before I'd put the kettle on. Three structures, ranked, tax handled, the practice's books pulled and checked. Recommended: take it, fixed, restructure at year seven.

Then underneath, in the assumptions, where nobody looks except people like me:

Household horizon: 9 years.

I read it four times. Not because I didn't understand it. Because it was a nine and the loan was a twenty, and I have spent my entire working life explaining to people what it means when those two numbers don't match.

I didn't ask it to explain. It would have. That's the thing about it — it always explains, it has never once refused me anything, it has been good to me for eleven years and it was good to me that afternoon. I asked a question and it answered the question. There was no other honest way to answer it.

Teun was in the kitchen standing on one leg. He does that while the kettle boils and he doesn't know he does it. Eleven years and he has never once noticed.

"Well?" he said.

"It's fine," I said. "Fixed. We restructure at seven."

He said great, and put his arms around me from behind, and I could feel the calluses through my shirt. They're on the pads under his fingers, not the palms. That's from the work. Everybody gets them somewhere different.

I haven't told him.

Not to protect him. If I tell him what it said, I have to tell him that I've been not asking. On purpose. Every day. For eleven years. And he doesn't know that. He thinks I just never got round to it, the way you never get round to the loft.

He thinks I'm somebody who didn't. I'm somebody who wouldn't. Those are not the same and I have never worked out how to say the difference out loud without it sounding like an accusation, or a gift, and it isn't either of them, or it's both.

The loan goes through in April. I've read the papers twice and they're good papers.

In the morning I asked about tiles for the bathroom. It gave me three and they were all fine.

Teun picked the wrong one. I let him, because it's a bathroom.

One day with acados: 8 errors I hit and what they meant

Jasper Blank — Fri, 10 Apr 2026 09:23:22 +0000

Yesterday I spent a day learning acados.

I was not trying to become an expert in nonlinear MPC in one sitting. I wanted to see where the friction actually was when going from "the examples run" to "I can change things and still understand what is happening."

I used the Python examples, changed them, broke them on purpose, and kept notes. I also used Claude and Codex the whole time, but mostly as pair engineers: terminal output in, next step out, then back to the code.

The main surprise was that the hardest part was usually not the control part. It was setup, code generation, environment state, and figuring out which changes were runtime-safe versus which changes forced regeneration.

These were the eight most useful failures from that day.

1. `EOFError` after "Tera template render executable not found"

The first Python example did not just fail. It failed in a misleading way:

Tera template render executable not found ...
Do you wish to set up Tera renderer automatically?
...
EOFError: EOF when reading a line

The last line points you toward Python input handling. The real problem was simpler:

t_renderer was missing
the script tried to prompt for setup
the shell was non-interactive

Fix:

install t_renderer

This one set the tone for the rest of the day. The raw error was not fake, but it was also not the useful diagnosis.
Yesterday I spent a day learning acados.

These were the eight most useful failures from that day.

1. `EOFError` after "Tera template render executable not found"

The first Python example did not just fail. It failed in a misleading way:

Tera template render executable not found ...
Do you wish to set up Tera renderer automatically?
...
EOFError: EOF when reading a line

The last line points you toward Python input handling. The real problem was simpler:

t_renderer was missing
the script tred to prompt for setup
the shell was non-interactive

Fix:

install t_renderer

This one set the tone for the rest of the day. The raw error was not fake, but it was also not the useful diagnosis.

2. Missing `.so` really meant `LD_LIBRARY_PATH` was wrong

Later I unset LD_LIBRARY_PATH and reran a working example. The failure was:

OSError: libqpOASES_e.so: cannot open shared object file: No such file or directory

That points at one specific shared library. In practice the right action was:

add <acados_root>/lib back to LD_LIBRARY_PATH

This is a pattern I see a lot in technical tooling. The message is factually correct and still not the fastest route to a fix.

3. Copying a getting-started example was not enough

I copied minimal_example_closed_loop.py into my own folder so I could experiment on it without touching the original. It failed immediately:

ModuleNotFoundError: No module named 'pendulum_model'

The script looked standalone. It was not.

Fix:

copy pendulum_model.py and utils.py too
or keep running from the original example layout

This is small, but it matters. It is exactly the kind of thing that makes a tool feel brittle when you are new to it.

4. WSL clock skew looked worse than it was

After changing N_horizon, I ran the copied example from /mnt/c and got:

make: Warning: File 'Makefile' has modification time in the future
make: warning: Clock skew detected. Your build may be incomplete.

At first glance that looks like generated code might be corrupted. In my case it was mostly a filesystem issue:

active build artifacts were on a Windows-mounted path
build tools inside WSL did not like the timestamps

Fix:

keep active generated code on the Linux filesystem

This is one of those cases where a good tool should say "probably environment noise" before the user goes hunting inside the OCP.

5. `yref` shape errors were understandable, but late

I injected a bad yref shape and got:

AcadosOcpSolver.set(): mismatching dimension for field "yref" with dimension 5 (you have 6)

This message was actually decent. The annoying part was somewhere else:

the example still did startup codegen/build first
only after that did the local shape problem show up

In this case the fix was simple:

use the right path-stage yref length

So this is less "decoder magic" and more "tell me this earlier."

6. `W` shape errors leaked internal details

Then I broke the running cost matrix:

AcadosOcpSolver.cost_set(): mismatching dimension for field "W" at stage c_int(0) with dimension (np.int32(5), np.int32(5)) (you have (6, 6))

This is the kind of output experienced users can parse and newer users still hate:

it tells you the shape is wrong
it also dumps internal type details into the message

The real takeaway was just:

expected 5x5
got 6x6

That is a good example of where normalization helps even if the solver is technically being precise.

7. `status 4` was not one thing

One of the strongest cases came from contradictory bounds. The output looked like this:

QP solver returned error status 3 (ACADOS_MINSTEP)
...
acados returned status 4

This is the point where the stack stops being beginner-friendly:

one layer is talking about the QP
another layer is talking about the OCP solve
the mapping between them is not obvious if you are still learning the tool

In this case the real issue was:

contradictory bounds
specifically lbu > ubu

So the useful diagnosis was not just "status 4." It was:

the QP failed
start by checking feasibility of the bounds

8. `ACADOS_MAXITER` meant different things in different contexts

At first I had a clean example:

set nlp_solver_max_iter very low
get status 2

That made ACADOS_MAXITER look simple.

Then I moved to a planar quadrotor MPC example and it stopped being simple.

I found at least three versions of "status 2":

the iteration cap was genuinely too low
I tightened runtime bounds and the maneuver became too hard
a constrained maneuver failed on a short horizon, then succeeded when I lengthened the horizon

The third case was the useful one.

I had a tight active angle bound. A maneuver failed. The first guess was obvious:

the constraint is too tight

I softened the constraint.
Still failed.
Slack stayed zero.

Then I increased the horizon.
The same maneuver succeeded.

That changed how I think about debugging these failures. A useful tool should not flatten status 2 into one canned explanation. Sometimes the right next step is:

try a longer horizon before relaxing the constraint

That is much closer to what an experienced user would actually do.

What I actually found useful about Claude and Codex

The part that worked was not "AI solves control engineering for you."

What worked was much narrower:

paste exact terminal output
ask for the next concrete step
go run it
write down what happened

Every time something useful happened, I turned it into one of four things:

an incident note
an error/fix pair
a recipe
a regression test case

That kept the whole day grounded in real failures instead of vague "can AI help with MPC?" discussion.

What I had by the end of the day

By the end of the sprint I had:

a small local acados decoder CLI
a recipe layer for setup and runtime-vs-rebuild questions
a semi-realistic trial pack of pasted logs
a regression evaluator

More importantly, I had a better picture of the actual scope.

The useful thing here is probably small:

decode confusing acados errors
suggest the next check
explain a few common runtime-vs-rebuild boundaries

That is enough to be useful. It does not need to be a universal control assistant.

One external signal that mattered

After I posted about this on the acados forum, one person replied to say:

yes, the segfault was real
no, they never solved it
they stopped using acados because of it

They did not have the old log anymore, so it was not a good eval case for the tool.

But it was still useful. It showed that a confusing first failure can be enough to make someone leave the tool entirely.

If you use acados

I would like to know whether the same pattern shows up for you:

setup errors that surface as misleading Python failures
status code combinations that are hard to interpret
runtime-vs-rebuild confusion
failures that look like infeasibility but turn out to be horizon or authority issues

If you have an acados log that confused you, send it to me. Those are the cases I care about most right now.

DEV Community: Jasper Blank

Teun

One day with acados: 8 errors I hit and what they meant

1. EOFError after "Tera template render executable not found"

1. EOFError after "Tera template render executable not found"

2. Missing .so really meant LD_LIBRARY_PATH was wrong

3. Copying a getting-started example was not enough

4. WSL clock skew looked worse than it was

5. yref shape errors were understandable, but late

6. W shape errors leaked internal details

7. status 4 was not one thing

8. ACADOS_MAXITER meant different things in different contexts

What I actually found useful about Claude and Codex

What I had by the end of the day

One external signal that mattered

If you use acados

1. `EOFError` after "Tera template render executable not found"

1. `EOFError` after "Tera template render executable not found"

2. Missing `.so` really meant `LD_LIBRARY_PATH` was wrong

5. `yref` shape errors were understandable, but late

6. `W` shape errors leaked internal details

7. `status 4` was not one thing

8. `ACADOS_MAXITER` meant different things in different contexts