Gamified software dev. On your nintendo, you slay bigger & bigger bugs until you get to the boss, which is a multithreaded race & dead-lock, 1-off, unicode, infinite recursion Heisenberg bug & the doc is in traditional Chinese & you have to go to Taiwan to find the original dev. https://x.com/msavichtweet/status/1006495150883909632
Home
Public posts preserved on my own site.
Lotto get's triple taxed, typically:
1st cut = house edge
2nd cut = winnings taxed as income
3rd cut = unless you keep it for yourself, a gift tax on anything you give to friends & family over the exempt limit.
4th cut = sales tax, but that is small
export PYTHONPATH = $PYTHONPATH:../mylib/src:../:../mylib:.:..:../mylib
This is the "I don't care where you find the package, it is 5:32 & time to go home pattern"
Also rebranding- the Abble iBone
def reprand(name):
return name.replace("p", "b").replace("P", "B")
Wait, what about PB&J, should that be BP&J?
def reprand(name):
return name.replace("p", "XXX").replace("b", "YYY").replace("YYY", "p").replace("XXX", "b")
For your next bress release.
This is a interesting new way to keep API keys out of your repos. https://x.com/joshboyd/status/1006271682665943040
We got a transit system that is also metaphor for how well we fund it. https://x.com/srepetsk/status/1006205164523610112
Ihob has a code & can't say p right today.
As per the Apple Store Jira Cloud is for ages 4 and up. After he gets thru HR my five year old will be joining the team and taking over the issue tracker.
IDEA: Pacific Rim redubbed with David Attenborough as a Kaiju nature special
IDEA: a federal lotto. You pick a "candidate" and if pick the right "candidate", you win health care, college education and a retirement. If you don't play, everyone except you wins a tax subsidy.
A book I read on self control said the poor use pick 3 as a sort of savings account for those with no self control &/or no access to banks.
W/my python simulation it checks out. After 32 years even with 50% payouts, house ends up with only ~20% of total played. Stats are weird.
(if you run the simulation for ever, (a 1000 years), then the state ends up with all the players money)
♡ 0 ↻ 0
Trying to wrap my head around this:
1 in 10^100 bet -expected value $0.50
1 in 10^10 bet -expected value $0.30
1 in 10^5 bet -expected value $0.25
I'd prefer the 3rd, despite low EV. First two feel like the range of impossibility, so EV doesn't matter. EV could be $1.05, same.
My genetic algo can't cope with highly infrequent events. They never happen in the number of simulated events I can simulate. A mega millions to powerball simulator would just say, gee, you never win. (Unless the simulation ran for hours or days, I guess)
Another prob q: What has better odds of winning a million: a single lotto ticket, or a "red paper clip" strategy of playing a game higher probability of winning x $ & then betting x $ on the low probability, million dollar game? Both bad bets, but 1 is worse.
Probability question: If all you have left is $3, which drawing game maximizes your chance of winning $2500? (Jackpot games especially have varying expected values)
Something wrong with cyclomatic complexity scores-- a class with many methods, each with a few if-blocks gets penalized like a single method with many if-blocks. Also looks like you get penalized for argument guards (if arg1 is None, raise, if arg 2 is None, raise) Feels wrong.
# zen of python, update to reflect reality.
["There should be 100s-- and preferably not more than 10,000s -- esoteric ways to do way to {0}.".format(x) for x in ["lint", "check types", "make docs", "build packages", "write build scripts", "parse options"]]
# this spins up all the cpus but has no concept of how hot it is making this machine run.
with Pool(processes=cpus) as pool:
results = http://pool.map(work, playable, chunksize)
# maybe if it had a param,while CPU temp<75
This is actually not that new of an idea, "Starbucks Green Tea Latte with espresso" == blended green stuff with coffee. https://x.com/mental_floss/status/1004473761972850688
Cool, a chart of where tourists visit and where there are traffic circles. https://x.com/randal_olson/status/1001863736972328960
# Keno Number Picker
p = set()
while len(p) < 20:
value = http://date.today() - timedelta(days=d) + timedelta(days=random.randint(0, 365 * 80))
p.update({int(str(value.year)[0:2]), int(str(value.year)[2:]), http://value.day, value.month}-{0})
print(p)
Shoot, has a bug, allows numbers over 80. Anyhow, it is keno number picker that only uses parts of dates, the year, century, day or month of something that plausibly could be a birthday.
♡ 0 ↻ 0
Needs to be restated, people keep wanting to set me straight on 10 y/o posts https://x.com/mistersql/status/859513284705361923
Mac just black screened, asked to send a panic report to Apple, & Sophos can stop attempting to restart & popping up windows.
Wonder if they'll realize they could be popping open the "buy a replacement mac" page on the Apple store instead.
"can stop" - "can't stop" - of all the typos I make, dropping the negative marker is the worst. Is it?
♡ 0 ↻ 0
Many/most states offer a download of the winning lotto numbers, but not for numbers played. That require a FOIA request?
The winning numbers are unbiased & boring. Played numbers are highly biased & fascinating.
Luck or because Ds have a better intuitive sense for game theory? https://x.com/Adam_Unger/status/1004313488230113281
Ad tech is in an impossible position here. The best customers for alcohol, opium, and gambling are going to be people with dependency & addiction problems & targeting is going to end up inferring people's health status in order to show them ads. Twitter API ToS 'prohibits' that. https://x.com/larimer/status/1004086414626062336
Sophos AV: "Oh! I see you are on low bandwith connection on a train. Let me spend the whole 3 hours trying to download a $#!+bytes of updates"
Sophos AV: "Oh! I see the screen is closed & you're on battery, let me unzip all the files (over & over) & check for viruses."
# python lets you quietly redefine member functions
class Wtf(object):
def __init__(self):
http://self.wtf = None
def wtf(self):
http://self.wtf = "Hey!"
return http://self.wtf
wtf = Wtf()
http://wtf.wtf() # boom https://x.com/mistersql/status/1004126241862152193
Python has gone all Buddhist on me.
(No self...ah, never mind) https://x.com/mistersql/status/1004126241862152193/photo/1
I hear Amtrak is shopping for new trains. Everyone write your congressman & ask for either the one in Snowpiercer or the one in Solo. Or given how high the rivers were to NYC, maybe one that runs under water.
Kinect for Github:
Raise arm - commit -m
"Xbox, fix bug"
commit -m "fix bug"
Swipe - Oh $#!t, repo has been reverted to 3 mo ago!
Swipe back - Oh $#!+, repo has merged into prod!
<cat walks in, steals your skeleton> "You want to delete? Really? OK, Repo successfully deleted."
Swatting is the police turning themselves into free hit men.
Anonymous phone calls should be illegal & come with a huge penalty, like $1 mil per call paid by phone companies.
I suspect phone spam will trigger the move before swatting does.
Who'd of thunk ... a Keno simulator (essentially a Gambler's ruin simulator) is isomorphic with modeling sediment moving down a river
https://www.researchgate.net/publication/260013838_Application_of_gambler's_ruin_model_to_sediment_transport_problems
Xerox's f'ing printers - who thought it was a good idea to ask users to type in their windows credentials on a twichy touch pad everytime they want to print? The help docs say "this is a one time thing" - it is a lie.
This whole Xerox thing is just industrial sabotage.
Relevant, since I turn 45 today. At least I know I'm not old for another nine years. https://x.com/TomerUllman/status/1003632069933785088
I have implemented AI!
while True:
input("What is your name?")
Alzheimer's Intelligence
Last nights Keno simulator coding. Combing my passion for programming & paying optional taxes.
Whoa nelly that last build was buggy. New optimal ticket (for goal of win/lose $2500):
DC 6 spot, 2 games, $3 bets, no bonus (no spin)
https://github.com/matthewdeanmartin/keno
I ascribe to the homeopathic theory of productivity and where productivity really kicks in is just a nano-week more than 0 days a week. https://x.com/jamisonfoser/status/1003730279939301376
KATA: write a biased RNG that picks numbers just like people, not empirically, but heuristically (i.e. relying on dates & avoiding runs & patterns)