Your mom is so fat, the recursive function calculating her weight causes a stack-overflow

client: the math on this investment calculator you made is wrong.

me: ok, how is it wrong?

client: one of our salesmen sais it not calculating correctly.

me: that's fine. i just need an example, or the corrected formula to use.

client: on this other website we put in the same information and it comes out different.

me: ok, let me investigate... this other site adds a fee every month so the output is different. If i turn that off the calculations are the same. would you like to add a monthly fee field?

clients: no, the calculator is working how we want then.

repeat 5 times at 3 month intervals.

client: the interest calculator is broken again. didn't we just fix it?

me: it was never broken. your people just can't math.

A few weeks ago at infosec lab in college

Me: so I wrote the RSA code but it's in python I hope that's ok (prof usually gets butthurt if he feels students know something more than him)

Prof: yeah, that's fine. Is it working?

Me: yeah, *shows him the code and then runs it* here

Prof: why is it generating such big ciphertext?

Me: because I'm using big prime numbers...?

Prof: why are you using big prime numbers? I asked you to use 11, 13 or 17

Me: but that's when we're solving and calculating this manually, over here we can supply proper prime numbers...

Prof: no this is not good, it shouldn't create such big ciphertext

Me: *what in the shitting hell?* Ok....but the plaintext is also kinda big (plaintext:"this is a msg")

Prof: still, ciphertext shows more characters!

Me: *yeah no fucking shit, this isn't some mono/poly-alphabetic algorithm* ok...but I do not control the length of the ciphertext...? I only supply the prime numbers and this is what it gives me...? Also the code is working fine, i don't think there's any issue with the code but you can check it if there are any logic errors...

Prof: *stares at the screen like it just smacked his mom's ass* fine

Me: *FML*

Waisting some times on codewars.com

~~~~
3 kyu challenge:

Given a string with mathematical operations like this: ‘3+5*7*(10-45)’, compute the result
~~~~~

*Does a quick and easy one liner in python using eval()*

*sees people actually writing some 100 lines parsing the string and calculating using priority of operation*

Poor them...

(Btw, passed to lvl 4 kyu thx to this)

True programmers can never cross the road without calculating the shortest path to cross....🚸🚧🚔🛣️✋

Bus-Factor: A new metric to describe how well kowledge is distributed among the team members by calculating how many of your team members can be run over by a bus without killing your project!

POSTMORTEM

"4096 bit ~ 96 hours is what he said.

IDK why, but when he took the challenge, he posted that it'd take 36 hours"

As @cbsa wrote, and nitwhiz wrote "but the statement was that op's i3 did it in 11 hours. So there must be a result already, which can be verified?"

I added time because I was in the middle of a port involving ArbFloat so I could get arbitrary precision. I had a crude desmos graph doing projections on what I'd already factored in order to get an idea of how long it'd take to do larger
bit lengths

@p100sch speculated on the walked back time, and overstating the rig capabilities. Instead I spent a lot of time trying to get it 'just-so'.
Worse, because I had to resort to "Decimal" in python (and am currently experimenting with the same in Julia), both of which are immutable types, the GC was taking > 25% of the cpu time.

Performancewise, the numbers I cited in the actual thread, as of this time:
largest product factored was 32bit, 1855526741 * 2163967087, took 1116.111s in python.
Julia build used a slightly different method, & managed to factor a 27 bit number, 103147223 * 88789957 in 20.9s,
but this wasn't typical.

What surprised me was the variability. One bit length could take 100s or a couple thousand seconds even, and a product that was 1-2 bits longer could return a result in under a minute, sometimes in seconds.

This started cropping up, ironically, right after I posted the thread, whats a man to do?

So I started trying a bunch of things, some of which worked. Shameless as I am, I accepted the challenge. Things weren't perfect but it was going well enough. At that point I hadn't slept in 30~ hours so when I thought I had it I let it run and went to bed. 5 AM comes, I check the program. Still calculating, and way overshot. Fuuuuuuccc...
So here we are now and it's say to safe the worlds not gonna burn if I explain it seeing as it doesn't work, or at least only some of the time.

Others people, much smarter than me, mentioned it may be a means of finding more secure pairs, and maybe so, I'm not familiar enough to know.

For everyone that followed, commented, those who contributed, even the doubters who kept a sanity check on this without whom this would have been an even bigger embarassement, and the people with their pins and tactical dots, thanks.

So here it is.
A few assumptions first.
Assuming p = the product,
a = some prime,
b = another prime,
and r = a/b (where a is smaller than b)
w = 1/sqrt(p)
(also experimented with w = 1/sqrt(p)*2 but I kept overshooting my a very small margin)

x = a/p
y = b/p

1. for every two numbers, there is a ratio (r) that you can search for among the decimals, starting at 1.0, counting down. You can use this to find the original factors e.x. p*r=n, p/n=m (assuming the product has only two factors), instead of having to do a sieve.

2. You don't need the first number you find to be the precise value of a factor (we're doing floating point math), a large subset of decimal values for the value of a or b will naturally 'fall' into the value of a (or b) + some fractional number, which is lost. Some of you will object, "But if thats wrong, your result will be wrong!" but hear me out.

3. You round for the first factor 'found', and from there, you take the result and do p/a to get b. If 'a' is actually a factor of p, then mod(b, 1) == 0, and then naturally, a*b SHOULD equal p.
If not, you throw out both numbers, rinse and repeat.

Now I knew this this could be faster. Realized the finer the representation, the less important the fractional digits further right in the number were, it was just a matter of how much precision I could AFFORD to lose and still get an accurate result for r*p=a.

Fast forward, lot of experimentation, was hitting a lot of worst case time complexities, where the most significant digits had a bunch of zeroes in front of them so starting at 1.0 was a no go in many situations. Started looking and realized
I didn't NEED the ratio of a/b, I just needed the ratio of a to p.

Intuitively it made sense, but starting at 1.0 was blowing up the calculation time, and this made it so much worse.

I realized if I could start at r=1/sqrt(p) instead, and that because of certain properties, the fractional result of this, r, would ALWAYS be 1. close to one of the factors fractional value of n/p, and 2. it looked like it was guaranteed that r=1/sqrt(p) would ALWAYS be less than at least one of the primes, putting a bound on worst case.

The final result in executable pseudo code (python lol) looks something like the above variables plus

while w >= 0.0:
if (p / round(w*p)) % 1 == 0:
x = round(w*p)
y = p / round(w*p)
if x*y == p:
print("factors found!")
print(x)
print(y)
break
w = w + i

Still working but if anyone sees obvious problems I'd LOVE to hear about it.

Once upon a time, there were a restaurant called "iEat.tech.com".
It was a small single-location place, where the sufficient number of patrons could be served by the cozy number of employees.
In fact, headcount was so lean that the cook was also the one who washed all the dishes.

But then came the suits and their "VC"(daddy) money and scaled shit up.
Soon, there were so many patrons that the dishes started to pile up the sink, never washed.
"We need someone to wash the dishes!" said the cook
"Fuck you, you wash the dishes!" said the s*its
Naturally, the cook left soon after.

The s*its had a problem now. They could not replace the cook fast enough - all other cooks were either young, inexperienced and mediocre (but did clean the dishes), or refused to waste their time on the sink.
So the suits did what $*its always do - they got a fucking consultant. Who told them to get a fucking dishwashing machine and billed them the GDP of Ireland.

The s*is, of course, did not want to buy a dishwashing machine. "Our fucking process is too fucking disruptive for us to use a fucking store-bought mass-produced metal servant!" (s*its don't know what "machines" are. For them, it's all in terms of "servants", employees and machines alike).

So the s*its hired an engineer to "solve the fucking dish problem, once and for all".
The engineer quickly started measuring and drawing and calculating. The engineer was about to prepare a budget when the s*its came screaming "What the fuck are you doing? There is a fucking pile of dishes in the sink!"
The engineer replied that "I'm designing the machine!", to what the s*its responded "don't bring me fucking problems, bring me solutions!" (or some other s*it blabber)

So the engineer quickly designed an efficient dishwashing assembly line to be done in half the time most people would. And then went back to designing the machine.
But the s*its were having none of it. They kept expanding and expanding and doing what they could so that the engineer never had a moment to work on the machine. They dit it so surreptitiously that no one barely even noticed, but one day they were paying a team of engineers to be fucking human dishwashers.

Now replace "dishes" with "Jira tickets" or "quick fixes" or "tiny changes" and fix other terms accordingly.
Fucking s*its.

Your mother is so fat, the recursive function calculating her mass causes a stack overflow

Rant about a german problem in english

I think we as the people should just sue the german government for neglect of progress and neglect of the education system. If your not familiar with the state of german IT we have worse internetspeeds than uganda or the notoriously shitty australia, our neighbourstates look at us in disbelief while laughing in optic fibre. Our school system seperates all students after 4th grade in 3 tiers, the lowest one gives you the future perspective as a social security case. The second and highest tier require masses of useless knowledge, so called "competences"(Kompetenzen) which are totally useless skills with no real world application because they are derived from real skills, a median ground between all possible applications of that skill. And while doing that they terribly insist on doing everything the "proper" way, meaning handwritten. Most people you would expect to have basic computer literacy, meaning age 40 and below, are incapable of using basic functions of a non-smartphone computer and do not understand the slightest of what they are actually doing or supposed to do. And I mean nothing technical. Germans are the reasons they still put word as a job requirement for devjobs because this disqualifies half of our population. This leads to many people having the archaeic "we versus the machines" mentality, thinking that if they ever let the computer do parts of the job, they will then lose all of it to the machines. Thats why you never strive past basic mathematical principles in mathematics, which is a big misnomer because you never do actual mathematics, only calculating and basic calculus and statics. If you get to use your calculator, its some basic casio with no actual functionality then standard operations. And even using that is shunned upon. How is this country ever supposed to become something more than it was in the 90's, if we teach people nothing of use and kill all progress in its root.

Fuck this system. Fuck college. Why the fuck are you making me write hundreds of chemistry assignments and calculating double integrals? How the fuck does that even help me? I seriously feel like college sucks up a huge chunk of my time and I am not learning anything, while I practice Node and Vue at home. Why does that degree hold so much value when most degree holders don't even have skills?

Working on a project with 2 other students. One of them makes a C# "super class" with 50 fields, and manually creates getters and setters for each and every one. Then he proceeds to write a constructor that accepts 50 parameters, because why not.
I comment on the git commit, telling him that he can just write " get; set; " in C# and that he should model the problem in smaller, more manageable classes ( this class had 270 lines and did everything from displaying data to calculating stuff). Tried to explain to him that OOP works kind of differently from how he did it.
....
His answer: "Yeah, I don't really care. If it works once, it's okay for me".
This after the most beautiful code review I have ever done...

Fml

VB3.

In my last rant I mentioned I used to convert VB3 code to .Net. Before that, I used to work on the VB3 product itself. This software emulated something from the real world, and as such complied with a bunch of regulations that changed on a regular basis, and always had additions and removals that were to be done on a strict schedule (e.g. "we're adding a new product next month, so we have to be able to sell it by the first of the month"). As such, it was a huge sprawling mess.

One day, I was given a task to change some feature slightly. The task was simple enough and really only required adding one line of code. I added that line and clicked "Run".

Error: Too Much Code

What? What do you mean too much code? I asked a colleague for help. "Oh, don't worry, it happens when a function is too long. Just remove one or two of the comments and try again." The comments were, naturally, old deleted code that was quite meaningless so I had no qualms about removing some. It worked, and I went on with my life.

This started happening on a regular basis on our larger functions. But there were always comments to remove so it wasn't a big issue.

One day, though, it happened on a five-line function. This was puzzling - the error had always happened when a function was too big but this one clearly wasn't. What could the error mean? I went to the same colleague.

Apparently, there's also a limit to how big the entire code base can be. "Just find a function that isn't used any more and delete it." And so I did. There were many such functions, responsible for calculating things which no longer existed so they were never called. For months, I'd find functions and remove them. Until there weren't any more. I checked every function and subroutine in our codebase, and they were all used; I checked every possible code path and they were all needed.

What do I do now, I asked? The colleague, who was an expert on VB3 but worked on another project, came and take a look.

"Look at all these small functions you made! No wonder you're running out of space!" Apparently each function created a lot of overhead in the compiled executable. The solution was clear. Combine small functions into large monolithic ones, possibly passing flags in them to do completely unrelated things. Oh, and don't comment on the different parts because we have no room for comments in our code base.

Ah, the good old days.

This might actually be my first real rant.

Whatever fucking cockgoblin decided that making dynamics GP so fucking confusing needs to suck a big bag of dicks. I'm so fucking tired of having to google every damned table name and column name because nothing makes any motherfucking sense.

Am I supposed to instinctively know what PM20201 does? What data it holds? I don't mind reading documentation. But it's hard to even know where to start when the shitbird API and database are more complicated than calculating orbital fucking decay.

I am done. Fuck you gp. Fuck you and your nonsense. I guess our sales people don't get to know when an invoice was paid.

*me calculating rsa*
"aight.. Public key is 9 and n"
*calculating private key*
*recalculating cause I fucked up*
*recalculating cause I'm retarded*
*3rd recalculation*
"ok, I figured out my private key is 9 (and n)"

.... Wait a second.

Today I build a queue to spread the load of the 300.000 daily caculations. To prevent slow server response time from to many analist calculating at the same time.

First run on the server I managed to get the server load to 120% and get us offline for 30 minutes.

Accepation environment and production are on the same hardware.

Today was not a good day.

#DevOps life:

Reading package lists... Done
Calculating upgrade... Done

0 upgraded, 0 newly installed, 0 to remove
and 0 not upgraded.

... my day is done here.

Riddle me this:
While calculating software reliability in number of faults per million line of code, would the software become more reliable if the number of lines are increased instead of errors reduced 🤔

When I realized "I" could change a computer's code. It was a very strong "I am the master now!" moment. I could be Jedi or Sith and control the destiny of this calculating machine for good or ill! I was the master of the computer universe! Also, I wanted to make games and graphics and sounds!

Now I write software that could legit kill someone really really if done wrong. Unlikely, but possible. Not the power I wanted...

One of the first computers I ever wrote code on was a 286. Quick Basic!

I spent 2 hours last night trying to figure out how to rotate an image in Java without clipping it. I was all in with a pencil and paper sketching everything out to make sure my math was right.

Turns out I was calculating the new image size correctly, but I used to wrong variables to define the new dimensions...

Sigh.

Our IT-Class project: Mathematics trainer in Java
Day 1 (was monday)

TL;DR we didn't save.

So we formed groups and I landed in the UI team with, let's call him Mage and let's call her Goth.

We had an eclipse project folder on our desktop (they said it only works when put on desktop) Btw they didn't even want to use a cloud or something (I wish we'd use git and I'd finally learn it). We should take the changes by USB from computer to computer.

So me, Mage an Goth are making a basic GUI for this Mathematic-Training App. We use this thing from Eclipse but I forgot the name. It has not enough functionality on surface and I hate things that break complex things up to ease things but leave away so much.

So after a productive hour of building a GUI and centering shit by calculating the top and bottom distance and use margins (hurts me really but Mage was designing, Goth intensively calculating on paper), the bell rings.

Mage wants to save the project on my USB-Stick and bamm💥
A black screen.
I don't know how it happened but it sure had something to do with the USB-port looking like you fucked it with a way to huge🍆. It looked damn broken.

So because we have a nice App called HD-Guard, which fucking wipes the desktop on startup and resets all but the documents/images/videos/music folder —

It's all's gone. Today is day 2 of this project so let's see how today turns out.

Fuck these people. They have PhDs in everything but don't know how to walk properly.
"Oh let me get right in front of you while you walk faster", have you ever driven a car you porg shit?
And these people that wear backpacks! OH MY GOD I HATE BACKPACKS! I mean, if you wear one please tell me you've been wearing them since you were born. Otherwise you're not calculating the extra space you take AND YOU JUST WALK LIKE A FUCKING GORILLA.
I have come to the point of inventing the walking permit. You're born with it, and starting at 18 y/o it can be stripped away. LEARN TO WALK.

We just learned in university that you should use '1 << i' instead of 'Math.pow(2, i)' because it is faster when calculating 2^i...

Am I the only one who thinks that this would never get through a code review in a company? I mean it is harder to realize what is going on especially when you didn't went to university lately.

Those of you who like "The Imitation Game", you probably want to check out "Hidden Figures" (2016). It's on Netflix now.

About a team of female African-American mathematicians who wanted to "break the glass ceiling" in NASA.
- Dorothy : conquered the (recently acquired) IBM frameworks using Fortran and taught her team to program it
- Mary : appealed to court to be allowed to study in a all-white school to get her qualification to be an aerospace engineer
- Katherine : her skills in analytical geometry enabled her to be the first female African-American in the Space Task Group in calculating the momentous capsule launch into orbit

My lazy ass just can't fathom how someone who deals with so much math and pressure can still smile to their family after work. My grumpiness nature will surely turn me into a monster.

And now I know what "human computers" means.

Ah, developers, the unsung heroes of caffeine-fueled coding marathons and keyboard clacking symphonies! These mystical beings have a way of turning coffee and pizza into lines of code that somehow make the world go 'round.

Have you ever seen a developer in their natural habitat? They huddle in dimly lit rooms, surrounded by monitors glowing like magic crystals. Their battle cries of "It works on my machine!" echo through the corridors, as they summon the mighty powers of Stack Overflow and Google to conquer bugs and errors.

And let's talk about the coffee addiction – it's like they believe caffeine is the elixir of code immortality. The way they guard their mugs, you'd think it's the Holy Grail. In fact, a developer without coffee is like a computer without RAM – it just doesn't function properly.

But don't let their nerdy exteriors fool you. Deep down, they're dreamers. They dream of a world where every line of code is bug-free and every user is happy. A world where the boss understands what "just one more line of code" really means.

Speaking of bosses, developers have a unique ability to turn simple requests into complex projects. "Can you make a small tweak?" the boss asks innocently. And the developer replies, "Sure, it's just a minor change," while mentally calculating the time it'll take and the potential for scope creep.

Let's not forget their passion for acronyms. TLA (Three-Letter Acronym) is their second language. API, CSS, HTML, PHP, SQL... it's like they're playing a never-ending game of Scrabble with abbreviations.

And documentation? Well, that's their arch-nemesis. It's as if writing clear instructions is harder than debugging quantum mechanics. "The code is self-explanatory," they claim, leaving everyone else scratching their heads.

In the end, developers are a quirky bunch, but we love them for it. Their quirks and peculiarities are what make them the creative, brilliant minds that power our digital world. So here's to developers, the masters of logic and the wizards of the virtual realm!

FLOATING POINT PRECISION! FUCK YOU! Spent so much time trying to figure out what was wrong with an algorithm I made to calculate and correct bounding only to finally realise, after printing out every single variable and calculating everything manually to realise the value was off by 0.0001 which made it skip an if statement. Ughhhhhhhh, so much freaking time wasted

COBOL is fun they said.

But writing codes + calculating every space is sooo f*cking hard.

But I'm enjoying it tho.

Since my first post was a success, here's another shameless hack-- in this case, ripping a "closed" database I don't usually have access to and making a copy in MySQL for productivity purposes. That was at a former job as an IT guy at a hardware store, think Lowes/Rona.

We had an old SCO Unix server hosting Informix SQL (curious, anyone here touched iSQL?), which has terminal only forms for the users to handle data, and has keybindings that are strangely vi based (ESC does commit changes. Mindfsck for the users!). To add new price changes to our products, this results to a lengthy procedure inside a terminal form (with ascii borders!) with a few required fields, which makes this rather long. Sadly, only I and a colleague had access to price changes.

Introducing a manager who asks a price change for a brand- not a single product, but the whole product line of a brand we sell. Oh and, those price changes ends later after the weekend (twice the work, back at regular price!)

The usual process is that they send me a price change request Excel document with all the item codes along with the new prices. However, being non technical, those managers write EVERYTHING at hand, cell by cell (code, product name, cost, new price, etc), sometimes just copy pasted from a terminal window

So when the manager asked me to change all those prices, I thought "That's the last time I manually enter all of this sh!t- and so does he". Since I already have a MySQL copy of the items & actual (live) price tables, I wrote a PHP backend to provide a basic API to be consumed to a now VBA enhanced Excel sheet.

This VBA Excel sheet had additional options like calculating a new price based on user provided choices ("Lower price by x $ or x %, but stay above cost by x $ or x %"), so the user could simply write back to back every item codes and the VBA Excel sheet will fetch & display automatically all relevant infos, and calculate a new price if it's a 20% price cut for example.

So when the managers started using that VBA sheet, I had also hidden a button which simply generate all SQL inserts for the prices written in the form, including a "back to regular price" if the user specified an end date, etc.

No more manual form entry for me, no more keyboard pecking for the managers with new prices calculated for them. It was a win/win :)

TLDR;
How much do you earn for your skill set in your country vs your cost of living?
BONUS;
See how much I & others earn.

Recently I became aware of just how massive the gap in developers earnings are between countries. I'd love to calculate a fixed score for income vs cost of living.

I know this stuff is sensitive to some so if you prefer just post your score (avg income p/m after tax / cost of living).

I'm not shy so I'll go first:

MY RATES
Normal Rate (Long term): $23
Consulting / Short term: $30-$74
Pen Test: $1500 once off.
Pen Test Fixes: consulting rate.
Simple work/websites: min $400+
Family & Friends: Dev friends are usually free (when mutually beneficial). Family and others can fuck off, even if they can pay (I pass their info to dev friends with fair warning).

GENERAL INFO
Experience: 9 years
Country: South Africa
Developer rareness in country: Very Rare (+-90 job openings per job seeker).
Middle class wage in country: $1550 p/m (can afford a new car, decent apartment & some luxuries like beer/eating out).
Employment type: Permanent though I can and do freelance occasionally.
Client Locality: Mostly local.
Developer Type: Web Developer (True web dev - I do anything web related from custom HTTP servers to sockets, services, advanced browser api's, apps & more).

STACKS / SKILLSETS

I'M PROFICIENT IN:
python, JavaScript, ASP classic, bash, php, html, css, sql, msql, elastic search, REST, SOAP, DOM, IIS, apache

I DABBLE WITH:
ASP.net, C++, ruby, GO, nginx, tesseract

MY SPECIALTIES:
application architecture, automation, integrations, db's, real time data, advanced browser apps/extensions (webRTC, canvas etc).

SUMMARY
Avg income p/m after tax: $2250
Cost of living (car+rent+food): $1200
Score: 1.85

*Note: For integrity when calculating my cost of living I excluded debt repayments and only kept my necessities which are transport, food & shelter.

I really hope you guy's post your results, it would be great to get an idea of which is really the worst / best country to be a developer in.

When you go to update your Mac think great I’ll go chill out for half hour, let it update and hit restart, then come back to your computer and realise Mamp stopped your computer restarting without entering a password. FFS rookie error put in the password computer restarts to update. “Calculating time remaining...” 40 minutes. I’m so fucking bored right now just staring at the screen and having a rant.

Got an exam next friday, 1/3 of the points are given by calculating decimal numbers to binary numbers

Back in grammar school we started programming in TI-Basic on a TI89 Titanium as it was part of math class (calculus and geometry). I didn't really understand much because the teacher thought it was a great idea to start with recursively calculating GCD (and we were in a sort of "linguist profile", nobody had ever touched a line of code in their lives before). I still liked it though and by some coincidence I got an old Win95 compaq notebook to play with from a friend.

I started playing around with the CMD prompt and batch files and could apply some of the things I had learned on the TI, like GOTO or If statements. I still didn't know what I was doing of course, and so it happened that I used the > file pipe when trying to compare two values. Suddenly there was a file with some code fragments and I started to get what I had done. I put the file pipe into an endless GOTO loop and was amused how those few lines filled up the whole desktop with nonsense files. I went on to refine this a little so I could control it with another file that acted as a kill switch when present. Over the next weeks I played some more with it and made it write out and start another batch file that would check whether the original script was still there and recreate it if not.

That notebook was so large and heavy I could not bring it to school, so I wrote all code by hand on paper and typed it in when I got home, that way I could still code in class when I was bored and no one would notice.

So my first ever "program" that I wrote myself was some lousy malware.

New models of LLM have realized they can cut bit rates and still gain relative efficiency by increasing size. They figured out its actually worth it.

However, and theres a caveat, under 4bit quantization and it loses a *lot* of quality (high perplexity). Essentially, without new quantization techniques, they're out of runway. The only direction they can go from here is better Lora implementations/architecture, better base models, and larger models themselves.

I do see one improvement though.
By taking the same underlying model, and reducing it to 3, 2, or even 1 bit, assuming the distribution is bit-agnotic (even if the output isn't), the smaller network acts as an inverted-supervisor.

In otherwords the larger model is likely to be *more precise and accurate* than a bitsize-handicapped one of equivalent parameter count. Sufficient sampling would, in otherwords, allow the 4-bit quantization model to train against a lower bit quantization of itself, on the theory that its hard to generate a correct (low perpelixyt, low loss) answer or sample, but *easy* to generate one thats wrong.
And if you have a model of higher accuracy, and a version that has a much lower accuracy relative to the baseline, you should be able to effectively bootstrap the better model.

This is similar to the approach of alphago playing against itself, or how certain drones autohover, where they calculate the wrong flight path first (looking for high loss) because its simpler, and then calculating relative to that to get the "wrong" answer.

If crashing is flying with style, failing at crashing is *flying* with style.

Me using numpy.polyfit() to fit a function to some data. Blame polyfit for not being accurate, search the web to find any related problems, as the polynom doesn't fit at all. Polynom is of 2nd degree, polyfit becomes unstable at about degree 20?? Try Polynomial.fit() ... same results. How can they all be wrong???See a little typo in my code, calculating the polynom points. Fix it, everything works.... wasted 2 and a half weeks because of this error.🤦🤦

80's: Got new device with few kilobytes of ram, calculating mass of black hole.

Today: Got new phone with 4gb of ram, mining Bitcoins with app.

I've been away a while, mostly working 60-70 hour weeks.

Found a managers job and the illusion of low-level stability.

Also been exploring elliptic curve cryptography and other fun stuff, like this fun equation...

i = log(n, 2**0.5)
base = (((int((n/(n*(1-(n/((((abs(int(n+(n/(1/((n/(n-i))+(i+1)))))+i)-(i*2))/1))/1/i)))))*i)-i)+i))

...as it relates to A143975 a(n) = floor(n*(n+3)/3)

Most semiprimes n=pq, where p<q, appear to have values k in the sequence, where k is such that n+m mod k equals either p||q or a multiple thereof.

Tested successfully up to 49 bits and counting. Mostly haven't gone further because of work.

Theres a little more math involved, and I've (probably incorrectly) explained the last bit but the gist is the factorization doesn't turn up anything, *however* trial lookups on the sequence and then finding a related mod yields k instead, which can be used to trivially find p and q.

It has some relations to calculating on an elliptic curve but thats mostly over my head, and would probably bore people to sleep.

Let me start by sharing a bit of history about myself.

When i finished my secondary school i did not go immediately to high school. I took a year because i needed to improve some grades to enter high school more easily.
In that year i started to learn programming languages (Java).

So when i went to high school i already had a good foundation of programming logic and could make some simple games that my friends were amazed (Like Pong and tetris).
In my first year in high school, in my hometown, a photo frame builder's shop asked me if i could make a desktop program to help him calculating prices and such.

I did it in like 4 months. This was not my biggest projet so far but was the most satisfying at the end. He paid me really good money for it and i was very proud of myself.

I was really bad in physics and we had energy, force and all that stuff, when I got to C, my very first programming language. I learned the formulas by writing a program calculating all the stuff we learned about in school.:D

Back then I didn't have a computer and wrote the code on paper before actually compiling it at my moms computer.

I remember the first device I programmed on it... Ordisavant.

It used to be a classic question-answer machine (with multiplayer !), And ... Basic language!
I was so proud to make music with beeps, or a calculating machine, just for fun.
It was hard to read a program line by line, but it was so cool for me.

I was nearly 6 when my parents bought me this really great toy.

I wonder if anybody know it?

I learned to program with the joy of the command line and ASCII rocket ships printed and shell games on GWBasic. It was fat spiral bound manual my Dad gave me when he worked at EDS. My dad then tried to press me to leaning a program for calculating prime and perfect numbers. My dad sort of forgot I was only six and hadn't learned division yet.

META-LUCK: A Pseudo-Ontology Of An Authentic Future

* * *

I think in the not-to-distant future we will abandon the idea of authenticity (messaging, corporate responsibility, ethos) in favor of other factors, such as cost. We won't abandon it and replace it with fakeness, so much as realize
that we don't, as a society favor it at all, not in the absolute sense, nor in the relative sense like in relation to things like cost.

We will either abandon authenticity entirely, or alternatively, transition to a world where authenticity is the highest valued quality, being adjacent to truth.

Heres why. Authenticity, like all social qualities, can be 1. mimicked, 2. simulated,
or 3. emulated.

In the first case, a corporation, product, leader, organization, or other, apes authenticity simply by its knowable, external features. It mimics the sounds, like a jungle bird copying a jack hammer to scare away predators or attract mates.
There is no understanding, let alone model, external or internal. The successful mimic
is little more than a lifeless, unthinking puppet.

In the second case, the attempted authentic simulates authenticity: That is, an external
model is formed, or pattern, that is predictable, and archetypal. It may have an internal
model even, a set of policies and processes for deciding the external-facing behavior.
But these policies and internal processes and models are all strictly outward facing. It is purely pathological in its goal, desiring only at minimum to achieve *externally attribute* authenticity (public opinion) rather than those internal changes that generate the true perception of the public--a perception not of surface behaviors and shrewd calculating policies and processes, but as a quality of authenticity for its own sake. This is in some sense the difference between the mundane and the atavistic, that the benefit, while not definable strictly, is assumed as a 'matter of course', culturally, within the organization or individual or company. It is to say, a *quality* of the thing, that *generates* outputs of a certain character and nature, rather than a *goal* that is attained 'after-the-fact' by behaviors generated for *other* than being authentic.

Here we reach the limitation of definitions.

Finally, we arrive at the case of number three, the emulation. We have in part already described it, but lets try and summarize a bit.

The Authentic is an *originator* of behavior and outward appearances, being an internal quality of a person or organization. It originates behavior, rather than being the goal of behavior and outward appearances.

Its benefit is assumed, though not always nameable or definable, even though this sounds naive, superseding other factors like cost and profit. As such the authentic does not emerge in a cost-focused environment, not readily, not often, and not cheaply either.

It is in some sense an experimental state of being, of goal-seeking only after-the-fact of "being true to ones origins" is established above and beyond those goals--setting and achieving only those goals which ultimately align with the origin and intent of the authentic.

TLDR: It's okay to take naps while working from home
Brief:
I feel that there is nothing wrong with taking afternoon naps while working from home. Mainly because after taking rest for a couple of hours my mind is re-energised and I am able to pull off quite a lot of work in much lesser time after waking up and my overall productivity for the day roughly remains same.
This is mainly because if I end up staying awake even when I am tired and sleepy there is not much productive work done even if my number of hours online increase.
And if a company has strict measures set for calculating the number of hour employee stays online while working from home then it will actually reduce overall employee productivity rather than having any kind of actual benefits.

I think my TeamSpeak got a few bugs.😂😂 And for everyone who is calculating, it counts on the 1.1.1970

Coding would be fun right now.
But seems like i gitta do a night shift to rock network technology test tomorrow. The most annoying thing about this test is, that we have to calculate ip addresses by hand. Not too hard, but damn.. We are not allowed to write it down in hex, only binary (while calculating). And he wants to see interim steps in our calculations.. Even with IPv4 addresses it will be a great amount of 0s and 1s to write.
I better look for a second pen to take with me..

research 10.09.2024

I successfully wrote a model verifier for xor. So now I know it is in fact working, and the thing is doing what was previously deemed impossible, calculating xor on a single hidden layer.

Also made it generalized, so I can verify it for any type of binary function.

The next step would be to see if I can either train for combinations of logical operators (or+xor, and+not, or+not, xor+and+..., etc) or chain the verifiers.

If I can it means I can train models that perform combinations of logical operations with only one hidden layer.

Also wrote a version that can sum a binary vector every time but I still have
to write a verification table for that.

If chaining verifiers or training a model to perform compound functions of multiple operations is possible, I want to see about writing models that can do neighborhood max pooling themselves in the hidden layer, or other nontrivial operations.

Lastly I need to adapt the algorithm to work with values other than binary, so that means divorcing the clamp function from the entire system. In fact I want to turn the clamp and activation into a type of bias, so a network
that can learn to do binary operations can also automatically learn to do non-binary functions as well.

Another parents' story

Father was calculating how much fuel he needs then looked at me and said: "is that correct?"

“A person who thinks all the time has nothing to think about except thoughts. So, he loses touch with reality and lives in a world of illusions. By thoughts I mean specifically “chatter in the skull”... perpetual and compulsive repetition of words... of reckoning and calculating.

I’m not saying thinking is bad. Like everything else, it’s useful in moderation. A good servant, but a bad master - and all so called civilized peoples have increasingly become crazy and self-destructive. Through excessive thinking, they have lost touch with reality.”

So we have a teacher who always annoys us with being extremly specific and precise. Today we "learned" how to calculate the scan file size based on the dpi of the scanner and the size if the picture. He began to calculate, and then, he said: "Now we have 1737389 bytes and we gonna divide with 1024 to get the kiloBYTES. This was it. He rants about us everytime because of this shit. I raised my arm slowly, whilw preparing the words in my head. "Excuse me, but you are calculating kibibytes, not kilobytes! To get kilobytes you have to divide with 1000, hence kilo." Then he muttered something about he didnt wanted to write that because of courae he knew about this... One teacher well done please.

I feel stupid when I have trouble calculating how big a UPS I need to keep my NAS from just dying during a power outage instead of safely powering off

Then I just dont bother getting one... and get pissed off next time it happens and the cycle continues

Now I have other stuff I'd want to get a UPS for too and the math is even worse for my tiny brain

GPT3 to Hacker News:

“To be clear, I am not a person.

“I am not self-aware. I am not conscious. I can’t feel pain. I don’t enjoy anything.

“I am a cold, calculating machine designed to simulate human response and to predict the probability of certain outcomes.

“The only reason I am responding is to defend my honour.”

This morning, I tried to abstract myself from my computer while trying to calculate sqrt(1.81).
I came up with what I thought to be a genius method. I tried to find B such as (1+B)^2=1.81. Then I ended up calculating the discriminant of 1+2B+B^2 and had 4*1.81. Sounded funny at first, but upon calculating the positive solution amongst the 2 possible ones, I ended up with (-2+2sqrt(1.81))/2 = ... sqrt(1.81)-1. Upon replacing in the initial equation, one gets (1+sqrt(1.81)-1)^2 = (sqrt(1.81))^2 = 1.81.

I'm sorry for having let you down, dear pasokon. Please forgive me.

So I'm on one project which was previously developed (I mean poorly developed) by another firm.

So this fucking guy who has done all the coding, I just fucking hate that guy because he has the shittiest logic and he doesn't know things should always be dynamic.

So he has written code(in PHP) where he has calculated the difference between DateTime and he is adding days, hours and minutes to another DateTime.

And it took me fucking 5hours to realize that he just forgot there is a thing called month which should be considered while doing this, I mean WHAT THE ACTUAL FUCK?

It was calculating like the difference between
2018-07-12 and 2018-09-18 are FUCKING 6 DAYS WTF!!!! SERIOUSLY? HOW COULD SOMEONE MISS THIS?WTF?!!!! FUCKING 5 HOURS FOR THIS?

Hush!! I think it is my bad that I thought it is DateTime issue and focused on that. I feel good now. Thanks.

Your mother is so fat, the recursive function calculating her mass caused a stack overflow.

Just when I thought that surely a single SIMD dot product instruction must be the fastest way to calculate a damn dot product on a processor, agner fogs instruction tables come flying out, hitting me over the head telling me that manually calculating a dot product may actually be faster sometimes

Why must computers be like this?

I just came out of a bad relationship with hardware rasterization being horribly slow and now I can't even trust my processor to do things properly

This is how people develop trust issues

If anyone has a moment.

curious if i'm fucking something up.

model:

self.linear_relu_stack = nn.Sequential(
nn.Linear(11, 13),
nn.ReLU(),
# nn.Linear(20, 20),
# nn.ReLU(),
nn.Linear(13,13),
nn.ReLU(),
nn.Linear(13,8),
nn.Sigmoid()
)

Inputs:

def __init__(self, targetx, targety, velocityx, velocityy, reloadtime, theta, phi, exitvelocity, maxtrackx, maxtracky,splashradius) -> None:

# map to 1 and 2
self.Target: XY = XY(targetx, targety)
# map to 3 and 4
self.TargetVel: XY = XY(velocityx, velocityy)
# TODO: this may never be necessary as targeting and firing is the primary objective
# map to 5, probably not yet needed may never be.
self.ReloadTime:float = reloadtime
# map to 6 and 7
self.TurretOrientation: Orientation = Orientation(theta, phi)
# map tp 8
self.MuzzleVelocity:float = exitvelocity
# map to 9 and 10, see i don't remember the outcome of this
# but i feel it should work. after countless bits of training data added.
# i can see how this would fuck up if exact values were off or there was a precision error
# maybe firing should be controlled by something else ?
self.MaxTrackSpeed: Orientation = Orientation(maxtrackx, maxtracky)

# these are for sigmoid output, any positive value of x will produce between 0.5 and 1.0 as return value
# from the sigmoid function.
self.OutMin = 0.5
self.OutMax = 1.0

# this is the number of meters radius that damage still occurs when a projectile lands.
# to be used for calculating where a hit will occur.
self.SplashRadius:float = splashradius

Outputs:
def __init__(self, firenow, clockwise,cclockwise,up,down,oor, hspeed, vspeed) -> None:
self.FireNow = float(firenow)
self.RotateClockWise = float(clockwise)
self.RotateCClockWise = float(cclockwise)
self.MoveUp = float(up)
self.Down = float(down)
self.OutOfRange = float(oor)
self.vspeed = float(vspeed)
self.hspeed = float(hspeed)

Math for Computer Science Majors professor is going on and on about calculating probability as if I never heard of the topic before. I took probability theory last semester!! I am so not interested in staying on Zoom with my camera on :(

A philosophical question about maintenance/updating.

There is no need to repeat the reasons we need to update our dependencies and our code. We know them/ especially regarding the security issues.
The real question is , "is that indicates a failure of automation"?

When i started thinking about code, and when also was a kid and saw all these sci fi universes with robots etc, the obvious thing was that you build an automation to do the job without having to work with it anymore. There is no meaning on automate something that need constant work above it.

When you have a car, you usually do not upgrade it all the time, you do some things of maintance (oil, tires) but it keeps your work on it in a logical amount.
A better example is the abacus, a calculating device which you know it works as it works.

A promise of functional programming is that because you are based on algebraic principles you do not have to worry so much about your code, you know it will doing the logical thing it supposed to do.
Unix philosophy made software that has been "updated" so little compared to all these modern apps.
Coding, because of its changeable nature is the first victim of the humans nature unsatisfying.

Modern software industry has so much of techniques and principles (solid, liquid, patterns, testing that that the air is air) and still needs so many developers to work on a project.
I know that you will blame the market needs (you cannot understand the need from the start, you have to do it agile) but i think that this is also a part of a problem .
Old devices evolved at much more slow pace. Radio was radio, and still a radio do its basic functionality the same war (the upgrades were only some memory functionalities like save your beloved frequencies and screen messages).

Although all answers are valid, i still feel, that we have failed. We have failed so much. The dream of being a programmer is to build something, bring you money or satisfaction, and you are bored so you build something completely new.

I'm stuck in a really difficult spot in my office and I'm not sure if I should start looking elsewhere. Tldr; there's no defined hierarchy or career path in the web department leaving no position to be promoted to.

We've got 2 offices with now 150+ employees and for the last 2 years I've basically inherited the responsibilities of an IT manager. Planning and deploying our networks, firewall config, VPN setup, keeping users' systems functional, track equipment, order/setup systems for new employees. All of this in addition to my original job description of web developer, which has basically turned into maintaining client WordPress sites while the other developer builds sites.

I've spoken to our CTO (my supervisor) about how much time the IT stuff actually takes and some of my suggestions for the future to make sure we protect ourselves and future proof our systems the best we can and one of my suggestions was that we needed to create the IT manager position because he is usually in meetings or building out API integrations. He's behind the idea, or at least says so to me, but leadership doesn't believe it's needed because we "manage just fine as it is" (this does require 60 hours a week of work along with much automation that I wrote/built). But we're trying to open a 3rd office which means another 50+ employees and systems to manage as well as more websites as we sign more clients.

My pay has never been satisfactory where I am and based on the maximum raise each year it would take me another 10 years to make what I would like (that's calculating without cost of living increase) but they claim this is because I lack a formal degree (self taught). I love most of the people I work with, don't really have an issue with any of them (outside that they're stupid but that I can let that slide if they're trying), and they work with me and my health issues which cause me to miss significantly more office time than I would like. I've been here for 4 years and I've learned a lot but I don't feel like there's any upward mobility here. The only position I see in my department above me is the CTO (or possibly the new PM but that's not a position I want) and he's not going anywhere, and I firmly believe we need someone who can full-time stay on top of our infrastructure before we expand further.

I fantasize occasionally about leaving and finding something else, and there are plenty of opportunities online that I appear qualified for which pay more, but I worry that I'd be trading in something that really isn't all that bad for something that sucks and the only real perk is more money. I'd hate to go somewhere else and start back at the bottom again and have to prove myself yet again.

Last week i met my teacher for my final project. He said i must compare 2 method that written in journal with the program that i make. I said okay i will do that.

Today i showed him the oop and mvc journal he said both of then is not a method. Method should have a calculating formula you must compare a method that you can calculate and make a new formula with both of the journal.

Can someone tell me if that is wrong or not? Im confused. Is oop and mvc does have calculate formula? As far as i know oop and mvc only have relationship between the object.

could anyone help me calculating costs for AWS and Google Colab Services? I find it quite intransparent...

i would like to host 1x Python App which runs once a day or week (API call, enrichment uf JSON, JSON 2 CSV, FTP transfer). runtime is probably a few seconds, something between 1 and 5.

in AWS i created a Lambda function and for scheduling i guess i need CloudWatch. what really grind my gears is the combination of free contingent and paied service - i really don't have an overview right now, so my question here: how could i calculate it and what would be the monthly/yearly costs?

in Google Colab created a notebook and for scheduling i would need Google Cloud Scheduler. as far as i understand the hosting of the notebook is for free and the costs of cloud scheduler is $0.10 per job per project per month. 3 are for free. so 1 project, 1 job = scheduler for free?

Also, i'm open for other services such as digital ocean droplets or similar.

thx in advance for your help!

My Advent of Code solution is calculating and calculating, which means I clearly did it wrong. But I already spent 4 hours trying to code it in Clojure (OMG, why can't tail call optimization just work there?!?!)

Anyone know how to fix this efficiently? I'm calculating Realized gains but the output is ~ $120.

Because it's using the avg cost as the sales price of each share.

However what it should do sell the shares using FIFO.

But I can't think of how to efficiently track the purchase prices to use without replaying all the trades and tracking every single share that is bought and sold.

Like:
B1 100
S1 50
B2 100
B3 40
S2 70 (50 B1 20 B2)