I am a researcher/trader and I spend a large portion of my time programming, but I am ultimately paid off the change in profitability of the strategies I've worked on and not on the code I've written. It's a means to an end.
When someone is working on a purely technical project, how can they successfully argue for the value of their work? How do you measure the value of a purely technical project? Are there other metrics that matter?
You must not be considering the total return of all your wealth just a limited subset. If you started with 10k and earned 300% (138.6% continuously compounded) then after 5 years you will have over $10mm. Wait another 3.3 years and you're a billionaire...
I was simply stating I have a substantially higher return than the pros. I've used that money for other purposes. for instance buying American airlines the day they went bankrupt allowed me to pay off a medical bill for a surprise medical emergency. But in that case it was literally 1000 percent return. I bought a bit of their stock the morning they went bankrupt. AMD is also an incredibly easy one to profit one. Wall Street might like to short sell on them to my benefit. From 3 dollars a share to 28 with incredibly predictable highs and lows for two decades. If I had the cash in college I could have been a millionaire by making ten trades a week on that one stock. The pros suck. I beat them year over year vs my managed retirement fund. It's pathetic.
Be sarcastic, but quite a few pros claimed it was a terrible investment. Same thing has been happening with amd. It's incredibly low hanging fruit. I just wasn't able to or decided not to reinvent the profits. My original point stands, the pros are terrible at their job. If you can't beat the rate of inflation in a multi billion dollar fund get a new job, despite all the fancy pointless tech.
>My original point stands, the pros are terrible at their job.
They are. Because no one can be good at that job. I think that's the part you're missing. You're deluding yourself that you've figured out a strategy for making money in a way that you beat the relevant market index consistently.
It was a terrible investment until they filed for bankruptcy and the stock went to 10 cents. Just because it went back up to a dollar because of everybody needed to close out their short positions doesn't mean the professionals were somehow wrong. They were making a killing, too.
That substitution was actually a result of printing text, and limiting the number of characters used. In the 13th century Scotland this was all going to be written by hand, so y was still an “i” or a “u” and thorn would have been its own character; uppercase þ, lowercase ð.
For an earlier example of y as a u, and thorns all over the place, try some of this intermediate Saxon circa 9th century
Feala ic on þam beorge gebiden hæbbe
wraðra wyrda. Geseah ic weruda god
þearle þenian. þystro hæfdon
bewrigen mid wolcnum wealdendes hræw,
scirne sciman, sceadu forðeode,
wann under wolcnum. Weop eal gesceaft,
cwiðdon cyninges fyll. Crist wæs on rood.
Amusingly, G’s at the start of a word are used like Y’s are used today, Y’s are U’s, and small furry creatures from Alpha Centauri were REAL small furry creatures from Alpha Centauri.
Julia lover here. I love having math at my fingertips and the ease of writing x’x\x’y. I like writing numerical code, knowing I can change types later and it will likely work. Still, I miss tab-tab from ipython for method completion from time to time and I work all day in mostly C++.
I get upset when they call it unlimited data and then cap the rate you can download. if you can only download 1 unit an hour, you are effectively capped at 744 Units per month.
It would also be great to see how these scale with terminal size. I personally use iterm2 but after switching to a new Macbook pro this year with two 5k displays it's noticeably slower. I'm assuming some O(n^2) scaling behind the scenes but I haven't measured anything myself. Still, @gnachman I love your term especially with terminalplots.jl and drawing Julia repl images inline.
`date 1>&2 | sleep 1 | date 1>&2 | sleep 1 | date 1>&2`
you'll actually see the three date commands output immediately so the three julia processes would actually launch in parallel if they were in a pipe