Live data from Hacker News

Ask HN: What's the hardest problem you've ever solved?

news.ycombinator.com

321–330 of 441 posts

Re: Ask HN: What's the hardest problem you've ever solved?

#321

Earlier quoted context omitted.

You rejected ethical behaviour 'on principle', you even seem proud of that. Have you kept doing that in life? Do you feel like an impostor, presumably with your degree that you didn't think you could acquire without cheating, or couldn't be bothered? A surprising number of the replies to your comment also seem to think cheating like that at university is perfectly fine.

Why isn't it fine, though? What serious damage did his actions cause? Which important societal construct rests solely on the assumption that no students cheat and, once this is violated by a single student, comes crashing down? I would suggest that if such a construct exists, then maybe the problem is its very conception and this is what should be rethinked. I think the implication that he should feel like an imposto…

"Which important societal construct rests solely on the assumption that no students cheat "

Society does not depend on wether or not a few students cheat.

But society definitely depends on the social and moral contract that we do not cheat, and that it is wrong. If most of us cheated, the system would definitely fall apart.

Have you ever visited a developing nation? Where outside of small villages it seems like everyone is cheating at everything, all the time?

It's like pouring sand into the gears of an engine: everything starts to break down.

I would not hire someone who admitted cheating during an interview unless they talked at length at their remorse, how they learned from it, how they grew from the experience, and there were exceptional circumstances.

Re: Ask HN: What's the hardest problem you've ever solved?

#322

Earlier quoted context omitted.

I have to steal that idea next time someone talks about time delays and says "it's not a big deal!" Awesome!

This is 50ms. This is 200ms. doesn't quite work as well. Depending on the situation there has to be some theshhold of diminishing returns.

you can use something like quickly flipping through some pages of what ever kind to find something, whet the page change is delayed by that much

if it's about computation you could make a bunch of objects load one after another

Re: Ask HN: What's the hardest problem you've ever solved?

#323

Earlier quoted context omitted.

Admitting to significant academic fraud during a job interview is a very, very bad idea. Furthermore, I would question the ethics and morality of any company that said "Wow, you solved an incredibly tough problem in order to commit fraud. Boy, are we the place for you!"

> Admitting to significant ... fraud during a job interview is a very, very bad idea. After serving his time, that's exactly how Frank Abagnale got a job in the FBI. https://en.wikipedia.org/wiki/Frank_Abagnale

That's different. According to that page, Abagnale served 5 years in prison (6 months in solitary confinement), and had to work for the feds for a while "without pay". Then when he first got legitimate work, he was usually fired.

I'd be willing to hire someone who cheated and was caught and served time, but not someone who cheated without consequence and boasted about it.

The commenter here doesn't sound like they segued that into a career catching academic cheaters. They just used it as a stepping stone to their next personal accomplishment.

Re: Ask HN: What's the hardest problem you've ever solved?

#324

Earlier quoted context omitted.

Admitting to significant academic fraud during a job interview is a very, very bad idea. Furthermore, I would question the ethics and morality of any company that said "Wow, you solved an incredibly tough problem in order to commit fraud. Boy, are we the place for you!"

I would allow it (if these kind of skills to design all of this system is the skills that is useful for this job). You are honest to admit it, and if you can make such thing, it is good. Of course that was cheating, but that does not mean that such thing is only to make fraud; you might do similar thing with other thing too.

Once you admit to fraud, though, they have no idea what else you might be lying about -- including the admission itself. They can't exactly call his old professor to confirm that he built and used a cheating machine.

Plus, after you admit to doing it once, if I weren't the sort to reject you outright, I'd still call every single person and institution on your resume to confirm all of those were real. Right out of the gate, you're proving yourself to be a lot more trouble than anyone else I'm interviewing.

It sounds like mostly luck that they weren't caught before. Are they going to try again, at my company? I don't want my company's legal footing to rest on one person thinking they'll never get caught.

Re: Ask HN: What's the hardest problem you've ever solved?

#325

I started with the glimmer of a hope that perhaps network sync could be made stateless, went down a months-long rabbit-hole of research, and ended up writing a novella-length article about CRDTs: http://archagon.net/blog/2018/03/24/data-laced-with-history/ So many days spent thinking, sketching, trying to swallow a concept that seemed far to big for my jaws—only to suddenly find myself on the other side, with this ar…

That might be mine as well, https://medium.com/@raphlinus/towards-a-unified-theory-of-op... . It's very much unfinished work though. Certainly wrapping my head around the OT and CRDT literature took huge amounts of brainpower exertion, comparable to what I did for my PhD. Also, I have your essay in my queue to read more thoroughly, it's high on my list of stuff to study as we possibly rethink the way this stuff works…

Hey Raph, thanks for the vote of confidence! I spent a lot of time looking through your research on OT and CRDTs (plus your Rope Science series) in the course of writing my article, especially as documented and implemented in xi. Really fascinating stuff. I love how many different ways there are to approach this problem.

Re: Ask HN: What's the hardest problem you've ever solved?

#326
As far as technical problems, there are plenty of optimization sorts of things that come to mind. But I'll go with a different sort.

A while back I was working on an embedded 3D scanner at school. Part of this was using a freescale ARM processor, and a USB interface to transfer scanned models back to the computer. But a weird problem was happening: only when trying to use the USB, every once in a while the processor would have a fault, and the fault was often different and seemingly unrelated.

After a week of constantly banging my head on the problem, I came upon a stroke of luck. We were smart enough to have built in a programmer on the board that supported gdb debugging (we built the circuit and populated the PCB). While in a gdb session, I read from the code flash section and noticed that there was a zeroed-out word (4-bytes) that definitely shouldn't have been. More intriguing, a subsequent read was correct, or had a different word zeroed! So the random faults were when one of these zeroed words occurred while code was read (and would have a bad effect depending in the instructions).

Now that I had found the problem, the solution wasn't too far, but still took some fiddling. Eventually I found that reducing the flash memory clock fixed the issue. We had all the clocks set to a valid configuration according to the documentation, and they did work fine when the USB module wasn't enabled. But as soon as it was enabled, it apparently interacted with the flash in weird ways.

I was glad to find a fix, and luckily we didn't necessarily need the flash to be fast. Embedded systems (and other complex black boxes) can cause some real frustration at times. But it's always fun, and I still found it rewarding :)

Re: Ask HN: What's the hardest problem you've ever solved?

#327

Fixing a SRAM 11 speed GX rear shifter on my mountain bike. I was on a downhill MTB holiday and fell off, snapping the thumb trigger on my rear gear shifter. Replacing the shifter is usually like a £30 part, but I was in an alpine resort in summer so the shops that stocked them were charging nearly £100! I had my tools and some super glue so sat down to repair it. The snapped part fixed in two minutes but getting tha…

That's pretty funny and I'm glad to hear something other than another barely believable "I did this god mode hack" story... The mechanical realm doesn't get enough love. Indexing shifters have become better, but progressively more miniaturized (and ironically flimsier IMO) over the past 30+ years. Good work!

Re: Ask HN: What's the hardest problem you've ever solved?

#328

I convinced my wife to marry me. She was very pessimistic, thinking that I would leave her. It took six months to get a positive response to my proposal and another year of reassurance before we married. We have been married 41 years. The courtship was helpful experience for the marriage.

Did her pessimism ever subside or did you just learn to accept it as part of her?

Re: Ask HN: What's the hardest problem you've ever solved?

#329
I wrote a transactional object store for an early PDA that could use both vanilla battery-backed RAM and flash media. Took about six back-to-back hundred hour weeks (and had a surprisingly short tail of bugs).

http://www.dadhacker.com/blog/?p=948

Later, I hooked the store up to the NewtonScript garbage collector in a way that let us have garbage-collected semantics for persistent memory-mapped objects. This let us do things like "receive a FAX" on a system with not a bunch of free RAM.

Re: Ask HN: What's the hardest problem you've ever solved?

#330

Getting well with a genetic disorder while the world harangues me and accuses me of making that up. Next up: Figuring out how to talk to the world about it effectively. I'm not sure that's solvable. "Curing CF" seems easier.

I don't understand. People try to tell you that cystic fibrosis isn't real?
Post reply on HN