And so but anyway, did I ever tell you about my most humiliating experience as a skilled and successful computer programmer?
-
@GeePawHill Have you read the Aubrey/Maturin series? It’s partly an extended essay on the knife’s-edge dance between the corrupting effects of inviolate power and being a social animal. And power due to position vs. power due to individual accomplishment.
@marick I have, actually, at least in part because you liked it so much. I've read them all.
-
@GeePawHill there was a story about a couple of scientists in WW2 assigned to improve U boat detection and destruction rates. 1 read reports and did calcs at a desk. The other went out on patrol and saw how hopeless reports were at conveying reality. It is a danger all disciplines of engineers can encounter and we often need to go and visit the 'workplace' to understand how the work is done and the reality of any equipment and automation. Oh, and add on human factors too.
@GeePawHill just checked my reference. It was German magnetic mines not U boats and patrol was on a minesweeper. Quoted in "Dispelling Chemical Engineering Myths" by Trevor Kletz. Original source appears to be R.V. Jones 1978, Most Secret War p353.
Anyway concept still holds if not detail -
@GeePawHill @mayintoronto and talk to the end-user, who may not be the same person!
@pozorvlak @GeePawHill @mayintoronto I just have three questions:
Am I allowed to talk to an end user?
Is it actually possible for me to talk an end user?
Is the end user willing to talk to me?
It's like herding genies. -
@pozorvlak @GeePawHill @mayintoronto I just have three questions:
Am I allowed to talk to an end user?
Is it actually possible for me to talk an end user?
Is the end user willing to talk to me?
It's like herding genies.@drgroftehauge depends on the context, but that's what a lot of my work is about. Getting people access to context so they can make better decisions.
-
And so but anyway, did I ever tell you about my most humiliating experience as a skilled and successful computer programmer?
@GeePawHill
Fantastic story!Another way of summing up the issue: When doing data analysis, get real world data as quickly as possible
-
And, for the record, I have been a successful professional programmer, an independent, for 45 years. I've failed more times than most people have even tried.
Some days you get the bear.
Some days the bear gets you.
Find joy in it. Without joy, why are we even doing this shit?
@GeePawHill I should totally write up probably my greatest failure as a programmer, where I may have quietly killed a professor's research project (well, one of them) by developing on a too-small machine. And I got to learn that I'd developed on a too-small machine in person in an awkward way.
(Someday I also have to write up my most successful program, which was more or less an accident that snowballed and I didn't even discover how successful until years and years later.)
-
And so but anyway, did I ever tell you about my most humiliating experience as a skilled and successful computer programmer?
@GeePawHill Thanks! -
And, for the record, I have been a successful professional programmer, an independent, for 45 years. I've failed more times than most people have even tried.
Some days you get the bear.
Some days the bear gets you.
Find joy in it. Without joy, why are we even doing this shit?
@GeePawHill this was a great read, thanks for sharing
-
And, for the record, I have been a successful professional programmer, an independent, for 45 years. I've failed more times than most people have even tried.
Some days you get the bear.
Some days the bear gets you.
Find joy in it. Without joy, why are we even doing this shit?
@GeePawHill Thanks for sharing! It’s always good to be reminded of how many ways clever software can easily be defeated by real world conditions!
-
And, for the record, I have been a successful professional programmer, an independent, for 45 years. I've failed more times than most people have even tried.
Some days you get the bear.
Some days the bear gets you.
Find joy in it. Without joy, why are we even doing this shit?
@GeePawHill Amazing! Thank you for sharing!
-
So, for my juniors, when I tell you "typing is not the bottleneck", I know what I'm fucking talking about.
It took me a couple of weeks to re-create 4 months worth of work. If I had to bet, I'd bet my second edition was *better* than the edition I lost.
So we come down to the day, and I am ready.
@GeePawHill this is what Artur Grabowski from OpenBSD calls "quality through destruction". When he worked on file systems he would test on the disks that held his changes.
When working on things you often realise something should be done differently, but "it works now". When starting over, v2 gets the cleaner code.
Great story, envious of your adventure!
-
And, for the record, I have been a successful professional programmer, an independent, for 45 years. I've failed more times than most people have even tried.
Some days you get the bear.
Some days the bear gets you.
Find joy in it. Without joy, why are we even doing this shit?
Awesome story!
Did anyone then suggest "what if we added a Kalman filter to integrate the sources?"
-
R relay@relay.mycrowd.ca shared this topic
-
Not, I repeat, my only great failure as a geek.
But, *damn*, that was humiliating.
I wrote an *excellent* program that *brilliantly* displayed data coming from hardware that didn't work.
It was a gig. I got paid. That's not the point. I was a pro, and pro's deliver *value*.
All I delivered was a good laugh.
@GeePawHill having myself given a hardware demo where almost everything went wrong in front of a few hundred people and god knows how many more on the live stream, any presentation where you get the whole room laughing is a good presentation
-
And, for the record, I have been a successful professional programmer, an independent, for 45 years. I've failed more times than most people have even tried.
Some days you get the bear.
Some days the bear gets you.
Find joy in it. Without joy, why are we even doing this shit?
@GeePawHill Thanks for the great story! It's all about having fun.
-
@GeePawHill Wow. What a story. Awesome.
But somebody hired you. This wasn’t your idea. You didn’t say “I have an idea: let’s bring these 3 devices together on an icebreaker.”
So somebody knew enough about these 5 things: icebreakers, gps, speed logs, radar, and computer programmers. They knew enough to imagine what each could do, but not enough to know that this wasn’t going to work at all.
And the supreme irony that you forgot to mention: all 4 ships, the icebreaker and its 3 ships behind, all made it safely to where they were going even while your thing didn’t work at all.
Brilliant story though. Humbling and hilarious.
@paco @GeePawHill i am guessing that the reason this story was presented is that to write software well you need a good understanding of the process and empathy for the people involved, and this test showed some fundamental misunderstanding that could have been resolved much earlier had anyone thought to ask, which is the embarassing miss, but at least it was caught during testing.
A lot of software today is written very arrogantly where there is little attempt to understand the problem or empathy for the users, because being *useful* is so decoupled from "money line go up".
-
And I was sub-contracted to do that. It was about a six month long project. I wrote an entire windowing system on top of DOS to use VGA to show the display.
(I'm a good fucking programmer, and that's not the only time I've written a graphical UI from scratch.)
And. A comical note: about six weeks before the project was due, my hard drive died. And. My backup drive died.
All I had were some two-month old printouts.
@GeePawHill i dont understand what we are doing differently with all of the tools we have today, because i think back to how much i can do in 6mo and it doesnt include writing an entire graphics system on top of whatever the actual goal is. We have web browsers and html/css/svg with openstreetmap and im not sure i could make something with the same featureset in the same timeframe. I dont have a complete clear thought but im not sure we have the right level of abstraction and code reuse and might instead be making things harder and more comples for minimal benefit. Im not saying we need Amish / acoustic pixels or anything, but it seems like this would still be a 6mo project and im not sure how much the layers of functionality between then and now are buying us. There are many things i wouldnt want to do without, like unicode, but it seems we built everything that is needed and useful 20y ago amn its been downhill and bloat ever since.
Sorry for the stream-of-conciousness, trying to get a thought in order
-
@billseitz They were all already in place, with displays, on the bridge. I suspect they were often ignored, cuz they didn't work very well.
@GeePawHill @billseitz “I suspect they were often ignored” ought to be posted in giant letters where anyone involved in a data visualization or consolidation project should see them.
A neat experiment: how would putting that in the startup instructions for an LLM affect the resulting product?
-
@GeePawHill as someone who once worked on calibrating software for inertial navigation units...
Yeah.
Your story is epic.
@knowprose
Tell us more, please. Which depths lie there unseen?
@GeePawHill -
@knowprose
Tell us more, please. Which depths lie there unseen?
@GeePawHill@yala @GeePawHill in that role and others, I was always adjacent to those things, never a participant.
I was the one slaying dragons while the cool kids were out doing the fun stuff.
But I got stories. None of them mine.
-
@GeePawHill here's an illustration of another good point : go on the field to see how shit works before coding any line of code that's suppose to fix that shit.
@Fangh @GeePawHill by now I don't even remember how many times my teams coded some kind of quick, hacked together version instead of multi-month solution, to learn that domain doesn't work as they expected and they would have wasted many months if they went with perfect plan.
Difficult to coach or explain to people who never experienced it.
