The Whiteboard

user41796

12:59 AM

@psr - that rate you cited sounds about right for locale and experience and nature of contract.

1:49 AM

@durron597 Thanks for thinking of me but as much of a hell as I am going through right now I really want to stick it out here for at least 4 years

My resume is littered with jobs, never longer anywhere than a couple years. I dont want to be unemployable because I never stick around.

Brian Snow

Is it common for headhunters to try to place people who've never had a full time job?

I would guess no...but I figured I'd ask anyway.

user55340

@durron597 I'm once again the only white guy on the team...

user55340

At employer{2,1}, $50k was a lot... The upper end of their pay scales. But it wasn't anything near a major metro.

user15026

@MichaelT I've never made that much o.O then again i've never worked in your field.....

user55340

2:08 AM

@AshleyNunn entry there was in the $25k to $30k range. And that was still above the median for the area.

user55340

I'm now making a bit more here, with very nice benefits.

user55340

But then, this is a larger metro area.. If I worked outside the public sector it would be even more (state doesn't have much wiggle room on salary)

user55340

(More, younger in the server sitting side of the house... Few younger in the code or dba side. I think the mean dba age is low 50s)

user55340

@JimmyHoffa btw, the dbas are Architects, not Administrators. It's a joy working with them.

Robert Harvey

2:49 AM

@MichaelT So basically they pay you just enough to keep you in the black, and that is all.

user55340

3:07 AM

@RobertHarvey yep, though note that unemployed there for the better part of a year, I spent less than 10k. It is very cheap to live there. My property taxes just go paid - $2k on a 2 bedroom house and double lot.

durron597

1:28 PM

So this question made my hot questions list:

23

Q: What's the best way to avoid catastrophe caused by negligence?

I have been a programmer for almost 1 year. As an ADHD adult, naturally I don't have the same strength of attention on ordinary stuffs as my colleagues do. And I find the catastrophe made by me are usually caused by trivial negligence. Like for today, I found the cron process on the server col...

@gnat forget about hot questions having 3-4 close votes, what about ones that are actually closed

durron597

2:19 PM

@GlenH7 One of these moons someone will make me a room owner so I can read those.

Feeds

GlenH7 has added durron597 to the list of this room's owners.

durron597

@GlenH7 Sweet! Thanks

user41796

I'm back to hating Silverlight this AM

user41796

I think I'm getting tripped up by their cross domain policies, but I have one in place

user55340

That implies you love it sometimes

durron597

2:24 PM

@GlenH7 Rewrite the whole thing in flash IMO. How hard could that be

user41796

@MichaelT No, sometimes I'm just indifferent to it

durron597

I know, I was joking ;)

user41796

One of the project managers was a bit disappointed when I very clearly and concisely laid out how there wouldn't be any reuse coming out of this project.

user41796

They dislike burning hours on throw-away stuff. Opportunity cost is killer in that case.

durron597

2:51 PM

Wow dasblinkenlight is bitter

2

Q: Nearest multiple of a power of two fraction

Is there an optimized, performant way to round a double to the exact value nearest power of two fraction? In other words, round .44 to the nearest 1/16 (in other words, to a value that can be expressed as n/16 where n is an integer) would be .4375. Note: this is relevant because power of two fra...

java double rounding

enderland

TGIF...

durron597

He answered that question with an answer that used Math.pow and Math.round, so I downvoted it, and he got bitter and deleted his answer saying "That's only an illustration of the math behind the trick. Had you not downvoted the answer, I would have shown you how to do it."

durron597

3:02 PM

Ha, I wanted to flag Jimbo's question as duplicate of my question from 2013 but my question was actually on SO so I can't ;)

enderland

@durron597 can you do something like

err nevermind, was going to suggest multiplying by 16 and then rounding

which appears to be what someone said :)

which gives you the fraction

7/16 in this case

durron597

if you know it's a positive number you can just add by .5 and cast

But i'm looking for something like... convert to long bits, then erase insignificant bits with &, then convert back to double

enderland

Ah

user41796

@durron597 Can you do bitwise operations in Java?

durron597

@GlenH7 of course

dcorking

3:07 PM

@durron597 it isn't really a duplicate as your question was about MVP (I don't know MVP ;) ) and this is MVC.

durron597

MVP and MVC are basically the same thing

user41796

@dcorking Feel free to one-up my answer, btw. :-)

durron597

MVVM is the thing that's different

@enderland The tricky thing is figuring out how many bits to erase when you have 12345.44 instead of a number between 0 and 1

dcorking

OK - I got the impression they were different as you said business logic should in in the Presenter, whereas classic MVC puts business logic in the Model IIUC. (Perhaps the OP is not asking about classic MVC)

ratchet freak

@durron597 look at the exponent :)

user41796

3:09 PM

@durron597 - so it's the 13th bit position that determines if you round up or down, right?

durron597

@dcorking I wrote that question two years ago, I didn't know s**t about f!!kall then

enderland

@GlenH7 that's only for a purely decimal though, right?

user41796

Version A) shift left 12 times into new variable based upon number to be rounded. Shift the 13th bit into another variable and add the 2nd variable to the first.

user41796

@enderland Oh, that's a good point, I think.

user41796

Floating point format has a section for the number + a section for the mantissa

enderland

3:11 PM

Yeah. I forget how exactly that works (we talked about this at work a while ago ha)

durron597

77777.44:   100000011110010111111010001011100001010001111010111000010100100
77777.4375: 100000011110010111111010001011100000000000000000000000000000000

dcorking

@durron597 OK! GlenH7's answer is spot on (even though it is not about classic MVC - the points he made apply in either case.)

user41796

But you can take the fractional part of the number only, I believe

user41796

@dcorking You could make a better answer by citing SRP. That's if you were looking for questions to answer and pull down more repz...

durron597

@GlenH7 I should tag the question as C instead of Java so I get bit twiddlers looking at it, then just port it to Java.

user41796

3:22 PM

float roundMe = 777.44;
int wholeValue = (int)roundMe;
float decimalOnly = roundMe - wholeValue;

int bitShiftedDecimalOnly = 0;
int thirteenthPlace = 0;

for(int i = 0; i <= 12; i++){
	if(i==12)
		thirteenthPlace << decimalOnly;
	else
		bitShiftedDecimalOnly << decimalOnly;
}

bitShiftedDecimalOnly = bitShiftedDecimalOnly + thirteenthPlace;

float shiftedToDecimal;
//reverse shift here into shiftedToDecimal

float finalValue = wholeValue + shiftedToDecimal;

return finalValue;

user41796

The shifting part is likely wrong, and there ought to be a way to do that with a single command instead of looping

durron597

Also there should be a way to do it that works for 1/8 and 1/32 etc.

where it just takes an argument 3, 5 (respectively)

wait, are you shifting BY the decimal part?

user41796

@durron597 For an 1/8, I think I might do a simple switch statement with a bitwise comparison. So same general principal as the previous idea, just a different check for the rounding

user41796

@durron597 no, that was just a trick to get the decimal value only of the number

user41796

so you could round 777.44 as well as rounding 0.44

durron597

3:26 PM

thirteenthPlace << decimalOnly; you have the decimal part on the right side

user41796

3 mins ago, by GlenH7

The shifting part is likely wrong, and there ought to be a way to do that with a single command instead of looping

user41796

<---- Doesn't remember how to shift bits in Java

durron597

What are you trying to do, in english / pseudocode

enderland

hmmm my OS environmetn variable on Windows shows as "WINDOWS_NT"

... I think my system is b0rken

user41796

Get the decimal only value
shift 12 bits of the decimal only value into a temp location
add the 13th bit to the temp location to effect the rounding
Shift the temp location back down to a decimal value
Add the decimal value back into whole (no decimal) value.

user41796

3:30 PM

@enderland No, they switched to that as the OS string around Windows 2000

user41796

the 95, 98, me kernel died when 2000 rolled out

enderland

@GlenH7 really? that seems wtfdumb

Wasn't Windows NT an actual OS release?

user41796

No, the 95, 98, me kernel was wretched, wretched bad

user41796

Yes, NT was a much more robust version of winders

enderland

oh, so its the kernel. not the "operating system"

user41796

3:31 PM

NT was targeted for "small servers"

user41796

Maintaining 2 kernels (9* & NT) was killing M$ internally, so they tried to merge them but stuck with most of the NT kernel

user41796

2000, XP, Vista, 7, 8 all use the merged kernel that is primarily NT

enderland

Ah

user41796

@durron597 apparently you can't shift on a float in java

user41796

So I'm trying to think through another way to tackle that

user55340

3:35 PM

Sure?

user41796

@MichaelT No, not certain about that. I don't have any practical experience with twiddling in Java

user55340

Look at Float methods

Robert Harvey

What happens when you shift the bits in a float? I'm guessing nonsense.

user55340

floatToIntBits

user55340

intBitsToFloat

user41796

3:37 PM

I think this would possibly work too:

Robert Harvey

Don't you have a mantissa, exponent and some sign bits stuffed in there?

user41796

long bitShiftedDecimalOnly = decimalOnly * 1,000,000,000,000;
long thirteenthPlace = decimalOnly * 10,000,000,000,000;
decimalOnly >> 12;

user55340

There is also a raw bits method

user41796

because when every picosecond counts, we need to use raw bits.... :-D

durron597

HOLY CRAP

user55340

3:38 PM

Flag as offensive?

Robert Harvey

NO SHIT?

user55340

Or star... Decisions

user41796

@durron597 Yeah, @MichaelT is da bomb with the internals of Java.

durron597

Did you know that -1 << 23 == -1 << 55 in java

even in 64 bit jvm

oh wait, maybe it needs to be -1l

public static double round(double d, int precision) {
  int intPart = (int) d;
  double decimalOnly = d - intPart;
  long bits = Double.doubleToLongBits(decimalOnly);

  System.out.println(System.getProperty("sun.arch.data.model"));

  long mask = -1l << (54 - precision);

  return Double.longBitsToDouble((bits & mask))  + intPart;
}

user55340

@GlenH7 I dabbled in C# in one float answer too.

user41796

3:43 PM

I'll just rescind a little bit of that statement to leave it at "you da bomb with internals"

user55340

►

user55340

@GlenH7 I like bits. Every femtosecond counts.

user55340

And car talk rocks

durron597

@ratchetfreak did you see my answer

enderland

also.

being sick SUCKS.

durron597

3:50 PM

@ratchetfreak Math.round is not TERRIBLE, it's Math.pow that's really awful.

user41796

I always smile a little bit when I can downvote a question and push the required delete vote count down too.

ratchet freak

@durron597 and I avoided it by using bit shift to get the double value for a power of 2 ;)

durron597

@ratchetfreak I know, I was just responding to you here instead of there because it was a bit of a chatty comment

Robert Harvey

4:07 PM

@RobertHarvey But does the 'program' picture makes any sense? Like what would it do? Or is it just random code? (And if yes how wrote this?!) — Rizier123 2 mins ago

[sigh]

In one Thunderbirds episode, Brain's computer/robot recited some magic formulas for turning around a spaceship. As an adult watching the episode again, I was disappointed to find out that the formulas were just gibberish.

user41796

4:19 PM

@RobertHarvey Can't you just suspend them with a message of "go do something productive"

Robert Harvey

Turns out there's a meta post for everyone who asks about the code.

user41796

Just redirect them to never-never land and move on...

enderland

Always nice for your team lead to basically say he wishes he could hire you fulltime

Robert Harvey

Yeah, wishes are always nice. :/

user41796

@enderland "So hey, how do we make that happen?"

Robert Harvey

4:22 PM

The more I work in computers, the more I realize that code is just... code. There's nothing magical about it, never was. It only seems magical sometimes because a lot of people have worked on it for a long time, or a lot of money or processing power has been thrown at it.

@enderland Show me the money.

enderland

yeah... lol

durron597

Wow. I ran a comparison of the bit shifting answer and the multiplication answer... and the multiplication answer is FASTER.

enderland

It's nice though to have that perspective when you work with a lot of people who are ridiculously smart and experienced though

user55340

Multiplication can use math libs and hardware. Bits can't.

Robert Harvey

@durron597 Does that really surprise you? I would imagine that the only time there's a speed benefit is when the multiplication and bit shifting are congruous, as in divide or multiply by a power of two.

Jimmy Hoffa

4:30 PM

@MichaelT that's always great

durron597

I thought bit shifting was always supposed to be faster because it's one instruction whereas multiplication is 32 instructions

maybe it isn't anymore, and I learned all the wrong things in college

Robert Harvey

But multiplication is a generalized mechanism. Bit-shifting is very specific.

If you're multiplying or dividing integers by a power of two, bit shifting is almost certainly going to be faster, because bit shifting is specialized on that operation. But only for powers of two.

user55340

Multiplication can be optimized to reuse intermediate results.

Jimmy Hoffa

not just that but the key to the libs @MichaelT mentioned above is there are low level code primitives for lots of common things like multiplication that have been obscenely optimized by the chip manufacturers themselves because they're such common operations

user55340

Bit shifts are taken as is.

durron597

4:34 PM

@RobertHarvey I am always multiplying by a power of two. But I'm multiplying a double, not a long

Robert Harvey

Yeah, only integers.

If you're just working with the mantissa, maybe.

Jimmy Hoffa

besides what language were you doing the shifts in? What the actual compiled code comes out to be may be more than a single op depending on whether or not a bit shift requires a jump to where that op code is or not for instance, no saying how the generated code looks without disassembling it

durron597

@JimmyHoffa Java

Jimmy Hoffa

perhaps it inlines multiplication but jumps to bit shifts - who knows

user55340

If you hit floats, the math from coproc and such are insanely optimized for those ops

Robert Harvey

4:35 PM

Note that modern Intel processors have native floating point operations. You won't get any faster than that doing something different in ordinary code.

durron597

But I used caliper for the test, not some janky System.currentTimeMillis() - startTime;

Robert Harvey

It's not the test.

Anyone want some coffee? I'll make some.

Jimmy Hoffa

@MichaelT so what ya doing in your new job? Java? Oracle? Web? Desktop? Enterprise? Services? LOB apps?

durron597

Unrelated, want so see some odd voting patterns? Probably a function of rep loss for downvoting questions vs answers

-2

Q: Why does -1 << 23 and -1 << 55 return the same value in Java?

Why does -1 << 23 and -1 << 55 return the same value in Java? It seems they should be very different, as I'm shifting by a lot more places! However, they result in the same value. Why does this happen? public class BitShifting { public static void main(String... args) { long foo = -1 << 23...

java long-integer bit-shift

Robert Harvey

Ah, you got hit with the golden dupe hammer.

durron597

4:39 PM

It is a dupe, I was gonna VTC it myself when I saw the suggested dupe

But I might get a Reversal badge on a question where I posted both the question AND the answer

user55340

Btw, so 404 polyglot needs Piet in the code too.

Robert Harvey

Java is such a strange language. Though if I'd learned Java first, I'd probably think C# is such a strange language.

durron597

@RobertHarvey Strings are not primitives.

Robert Harvey

They aren't in C# either. They're just treated like values for ==, unlike Java, where you must use Equals(). == will give you reference equality in Java.

user55340

But does == give reference equally for other objects? Or just strings in c#?

Robert Harvey

4:45 PM

The == operator is overloaded for strings.

It uses a custom comparator.

durron597

I also think operator overloading can only lead to madness

Robert Harvey

Yes, but it's useful in certain cases, like this one. Most C# programmers never use operator overloading.

It's always method calls.

user55340

That, to me, is a bad thing. Inconsistent behavior.

durron597

@MichaelT Agreed

Plus there's the danger some fool will do this (which I'm sure you've all seen already but I link anyway): mcfunley.com/from-the-annals-of-dubious-achievement

(Scroll to the bottom)

user55340

I don't like + and += for strings in Java either.

Robert Harvey

4:47 PM

@MichaelT Reference equality is useless on strings. Interning makes it not only useless, but dangerous.

So you might as well make == do something that makes sense.

durron597

> So much so that about 70% of his logic took place in them. More than once I deleted code that looked like this:

foo.x = foo.x;

Only to break entire pages, because the side effects of that assignment were doing everything. Anyway, I hope you can all see where I was coming from now.

user55340

(+ for string literals... Okish. But += is awful)

Robert Harvey

@durron597 Bah. That's a whole 'nother ball game.

durron597

@RobertHarvey I just prefer it to not even be possible in the language.

Robert Harvey

Side effects not possible? Use Haskell. Oh, wait... Monads.

Those pesky little things. Aw, geez. You mean we actually have to interact with the user?

durron597

4:50 PM

@RobertHarvey λ> let 2+2=5 in 2+2 codegolf.stackexchange.com/a/28794/17249

user55340

Side effects hidden behind what appears to be a simple operator.

Robert Harvey

@durron597 There's no need to swear.

@durron597 That's just crystallography weenies engaging in mental masturbation.

durron597

@RobertHarvey That's a stupid thread anyway, right @MichaelT?

user55340

Heh. Most of my rep from there

durron597

@MichaelT 8-)

Robert Harvey

4:53 PM

Well, there's no side effects in ==. It's just a read-only comparator; everyone with a passing knowledge of C# understands that it's a convenient exception to the rules.

user55340

It's the inconsistency that it behaves special for one class.

durron597

FINE

enderland

@durron597 gotta one up me I see

Robert Harvey

Inconsistencies. There are certainly plenty of those, in both languages.

user55340

@JimmyHoffa plumber and janitor of data. Mostly Java. Can say more when on real keyboard.

Robert Harvey

4:56 PM

Anyway, we use ReferenceEquals() for strings in C#, if we really need it (which, 99.9% of the time, we don't).

Which is clear and unambiguous. We can still use Equals() too, if we really want to.

durron597

@RobertHarvey It's not about which is used more often, it's about operator overloading, ever.

Robert Harvey

I know. Use sparingly, wear seat belts.

durron597

== on objects is always object reference. Always. When I read code I know what it means.

Robert Harvey

The best example I know of for operator overloading is Complex Numbers, which usually don't have a native implementation in the language. It's nice to be able to overload the usual math operators so that they work as expected.

@durron597 And as you've already pointed out, that strings are not already a primitive is kinda weird anyway.

user55340

Vectors too

user55340

4:59 PM

Or matrix

durron597

If a language element supports a . operator, at all, it's not a primitive.

user55340

But I've seen the rabbit hole of overloads and the like in ruby where you write a DSL inside the language

Robert Harvey

@durron597 In that sense, there aren't primitives in either language. Everything is an object.

int.Parse()

user55340

Go find two lisp shop 'standard' macro libs and try porting code from to the other.

Robert Harvey

@MichaelT [gets evil idea to write Scheme in C# using operator overloads]

user41796

5:03 PM

@RobertHarvey We need to find you a job pronto.

Robert Harvey

LOL

user41796

"Look folks, he's conspiring to rupture the space-time continuum here. Let's get him to work on something productive already, alright?"

durron597

@RobertHarvey 3.toString(); doesn't get autoboxed

Robert Harvey

That works in C#.

irony.codeplex.com

> Irony is a development kit for implementing languages on .NET platform. Unlike most existing yacc/lex-style solutions Irony does not employ any scanner or parser code generation from grammar specifications written in a specialized meta-language. In Irony the target language grammar is coded directly in c# using operator overloading to express grammar constructs. Irony's scanner and parser modules use the grammar encoded as c# class to control the parsing process.

user55340

Just realized I put a db change request in for 20140130

Robert Harvey

5:12 PM

The WhiteBoard. Where @MichaelT keeps a running log of his daily work activities.

(and his random thoughts)

user55340

It's that or outlook journal

user20683

9:10 PM

@GlenH7 I cooked up a fiendish fizzbuzz after question: How would you determine which of the divisors has the majority?

user20683

it's stupidly easy but it will catch people

durron597

@WorldEngineer If you're talking about this isn't it always the smaller one?

user20683

@durron597 correct

durron597

@WorldEngineer Why is that fiendish? Actually asking.

user20683

@durron597 I suppose it isn't that much but I've been dealing with grotesque levels of stupidity lately

durron597

9:18 PM

@WorldEngineer You and me both brotha

>++++++++++[<++++++++++>-]<[>+>[-]>++++++++++[<++++++++++>-]<+<<[->>->+<<<]>>>
[-<<<+>>>]<>>+++<<[->+>-[>+>>]>[+[-<+>]>+>>]<<<<<<]>[-<+>]+>[-]>[<<->>[-]]>[-]
<<<[[-]++++++++++[>++++++++++<-]>++.+++.[-]<[-]+++++++++++[>+++++++++++<-]>+..
[-]<[-]<<[-]>>]<>>+++++<<[->+>-[>+>>]>[+[-<+>]>+>>]<<<<<<]>[-<+>]+>[-]>[<<->>[
-]]>[-]<<<[[-]+++++++++[>+++++++++++<-]>-.[-]<[-]+++++++++[>+++++++++++++<-]>.
+++++..[-]<[-]<<[-]>>]<<[[-]>>++++++++++<[->-[>+>>]>[+[-<+>]>+>>]<<<<<]>[-]>>[
>++++++++[<++++++>-]<.[-]]<>++++++++[<++++++>-]<.[-]<<<]>[-]++++++++++.[-]<[-]

(esoteric.sange.fi/brainfuck/impl/interp/i.html)

user20683

@durron597 okay? what's with the wall of brainfuck?

durron597

@WorldEngineer Fizzbuzz :)

user20683

@durron597 I figured

user20683

@gnat \o

gnat

9:49 PM

@durron597 closed questions sometimes leave hot list with a bit of delay, up to 10-15 minutes if memory serves. Not really a big deal (otherwise you'd see my loud complaints at MSE... and not only me)

durron597

@gnat That one had been closed for 9 hours.

gnat

@durron597 wow. 9 hours, never seen anything even remotely close to this. This is probably related to today issues wit hot questions, these were reported at MSE today. But anyway thanks, I will keep an eye if this continues...

durron597

@gnat I mean, it was a dupe-close with 23 upvotes

So I wasn't totally surprised.

gnat

2

A: Why would a closed question be hot?

Bikeshed-style questions could be both hot, because lots of people want to participate, and closed, because the question may not be suitable for a Stack Exchange site.

durron597

@gnat It was the hot sidebar, not the hot tab. Don't know if there's a difference

gnat

9:57 PM

@durron597 sidebar is the place where genuine network hot questions are shuffled to...

33

A: What changed in the Hot Questions sidebar algorithm?

The usual algorithm calculates a score, ranks the questions and then just pulls off the top few results. We're testing an alternative method that calculates a score, ranks the questions and then pulls a somewhat larger number off the top, shuffles them, and stuff the top of the deck into the si...

durron597

26

A: How do the "arbitrary hotness points" work on the new Stack Exchange home page?

Basically what's documented here: What formula should be used to determine "hot" questions? We have a few tweaks: Succeeding questions from the same site are penalized by increasing amounts. So, the first question from SO in the list gets multiplied by 1.0, the second by 0.98, the third ...

That formula doesn't mention whether a question is closed.

gnat

@durron597 yeah. I could not find the reference on that. I swear I saw it once upon a time

durron597

@gnat Of course that question is 4 years old so they may have changed it since then, but who knows. Maybe a closed question isn't a boolean but rather a score multiplier? 23 votes is a lot.

gnat

@durron597 no, it doesn't multiply, close drops from list no matter what. That's what I remember. By the way, found an old request that feels close and says status-completed, but it's rather fuzzy...

15

Q: StackExchange's question selection is horrible

Update Thiings are looking so much better right now. Thank you. Right now, by looking at the tin (my connection sucks and I can't open every one) Hidden features of Google -- list of X What is your most-used web application? -- poll Returning from a method with implicit or explicit "else" o...

discussion status-completed stackexchange hot-questions question-lists

^^^ "why are closed questions listed at all?!"

durron597

> Update: though I see things improving, it still seems closed questions aren't filtered out!

gnat

10:08 PM

@durron597 yes I also have read this. There's no clear indication of what it ended at

Transcript for