The Whiteboard

Jimmy Hoffa

17:00

@RobertHarvey ah c'mon, just because you had to use a hole-punch to create words doesn't mean people shouldn't hear explanations of what they are these days. It's far more common than you perhaps think for people not to really know why we use hexadecimal

Robert Harvey

They can't count the number of fingers on their hands, and figure that one out?

Jimmy Hoffa

@RobertHarvey they don't know what a word is, more specifically.

counting on your hands is great but it doesn't explain what significance the number 15 has (well, had) to computing

Robert Harvey

Well, maybe that's really what he was asking.

Sure didn't look like it.

Jimmy Hoffa

@RobertHarvey that's because he probably didn't even know the term "word"

user41796

@RobertHarvey yeah, I could have seen that being a decent newbie question had the OP edited to make it clear what they were struggling with

Jimmy Hoffa

17:03

from his perspective this hex thing is just arcane and with no noticeable reason, makes it hard to ask about clearly when you don't even know what to ask

user55340

Just try to explain to him a compact representation of ipv6 Addresses

Robert Harvey

Kinda like asking why there's such a thing as vanilla pudding when you don't even know what vanilla is.

Jimmy Hoffa

@MichaelT hell to that, I can just send him some papers from hughes...

user55340

@JimmyHoffa That one deals with base85 encoding...

Jimmy Hoffa

@RobertHarvey ...what is vanilla? Have you ever seen it? Touched it? I'm pretty sure it was invented by Kraft Services Inc. and is just some synthesized chemical concoction.

Robert Harvey

17:04

Vanilla

Vanilla is a flavor derived from orchids of the genus Vanilla, primarily from the Mexican species, flat-leaved vanilla (V. planifolia). The word vanilla, derived from the diminutive of the Spanish word (vaina itself meaning sheath or pod), simply translates as little pod. Pre-Columbian Mesoamerican people cultivated the vine of the vanilla orchid, called tlilxochitl by the Aztecs, and Spanish conquistador Hernán Cortés is credited with introducing both vanilla and chocolate to Europe in the 1520s. Initial attempts to cultivate vanilla outside Mexico and Central America proved futile b...

Eeek. Maybe I don't know what vanilla is either.

user41796

@RobertHarvey Just skimming the one box there and it doesn't appear to be quite right

user41796

There are 3 types of vanilla, IIRC

Jimmy Hoffa

do I need to add vanilla to the room now?

user41796

@JimmyHoffa, @RobertHarvey - here you go: huffingtonpost.com/2012/04/09/types-of-vanilla_n_1408217.html

Robert Harvey

@JimmyHoffa Only if it's vanilla-flavored vodka.

Jimmy Hoffa

17:09

@RobertHarvey vanilla vodka + coke -> taste like a cream soda. Great girly drink.

@RobertHarvey SEE! SEE! I told you! Some concoction from Kraft....

Robert Harvey

@JimmyHoffa Look at the fine print along the lower edge. :P

Also, y'know, Heather Lauren made it.

Jimmy Hoffa

@RobertHarvey I see. Ralph Lauren's wife is apparently an exotic booze magnate...

user41796

@JimmyHoffa All it takes is time and money

Robert Harvey

@GlenH7 That's what I tell everyone who asks "is it possible" in a programming question on Stack Overflow.

user41796

And it's true!

user41796

17:13

Wanna recreate the Xcode compiler on a system that's not a mac? You can do it! It's just a simple matter of time and money....

Jimmy Hoffa

oo, 10,142. Time to drop another 100 rep bounty on someone

user41796

ooh ooh! pick me! pick me!

user55340

@GlenH7 Glen needs to catch up...

user41796

I'm 202 away from another delete vote

user41796

@MichaelT True dat

user55340

17:24

(hmm, I'm a bit low on the month's board... still have #2 on the quarter though)

Jimmy Hoffa

I've been low on all the boards for quite a while now...

user41796

And I'm about 140 away from passing DeadMG on the all time user rep ranking

user55340

And then there's that @Ampt guy who has a bit more free time and only about 400 rep away from close...

user55340

I'm about to overtake Yannis..

user41796

@MichaelT Between his two jobs and whatever else, he's been pretty busy as of late.

Jimmy Hoffa

17:35

15y/o trying to maintain a 30kloc app he wrote. Good for him.

1

Q: Restructuring a large Chrome Extension/WebApp

First of all, I apologize if this is the wrong format for a question here, I'm new to Programmers.SE (but not Stack Exchange). I have a very complex Chrome Extension that has gotten too large to maintain in its current format. I'd like to restructure it, but I'm 15 and this is the first webapp ...

user41796

@JimmyHoffa I see that as "OMG I have this massive pile of ... on my hands now!" :-)

Jimmy Hoffa

@GlenH7 oh yeah, and the Q should probably be VTC'd, but it's still cool seeing young kids pull off full-sized code projects like that

read his ideas for his restructure, it sounds like he's learned a lot of lessons and has some good ideas about what solutions may mitigate the problems he's having with it

user41796

@JimmyHoffa The 2nd half of his question about general intent of the refactoring is on-topic. The performance considerations, not so much.

Jimmy Hoffa

I don't know if it is or not. I'm debating voting to close because I really don't see an explicit problem for us to help you solve. Sounds like you're soliciting opinions. Perhaps you should check out CodeReview.StackExchange with snippets of your code? We do definitely talk about design stuff here and would love to help you come up with solutions to design problems, but I don't see an explicit design problem here, just sounds like you have a big messy project (as many of us do) and want some advice on how to correct it. SE isn't so good at advice, better at direct answers to direct questions — Jimmy Hoffa 10 secs ago

user41796

I'll just clear out that trace of bitterness... <sigh>

Jimmy Hoffa

17:46

@GlenH7 even more surprising: Look at the dude's JS in github, I've seen far uglier hard to follow code from people with years in industry. It's all decidedly procedural looking but it's very well broken up with very small SRP looking functions.

and from his description on github, sounds like the thing does a ton of shit.

user41796

Props to him

user41796

17:59

@RobertHarvey - No, I don't have an hour or two. But I'm not sure it needs that much either. I didn't think we would accept design reviews, but Thomas' answer kind of explained where they should be able to fit into the site.

Jimmy Hoffa

0

A: How to analyze a scenario where a bug didn't get caught and adjust development workflow to prevent similar errors

@RobertHarvey's comment is dead right: Process will not produce bug-free software, but good process can at least reduce the recurrence of classes of bugs. Sounds like you're attempting to accomplish the latter, so I will make my suggestion of where I think your mistake came in: You failed to acc...

for whatever reason, I'm actually compelled to answer questions today....

user41796

@JimmyHoffa more repz for the inevitable bounty

Jimmy Hoffa

@GlenH7 it'll be a 150 rep bounty now. Maybe I should hold out for a while so I can push @Ampt over 3k in one single bounty...

Ampt

@JimmyHoffa quit talking like that or I'm gonna have to give out some repz

user41796

48 mins ago, by GlenH7

ooh ooh! pick me! pick me!

Jimmy Hoffa

18:13

@Ampt I will not stand for trickle down stackonomics. Especially if it all goes back to @GlenH7, that's far too analogous to how real trickle down theory works in practice...

user41796

I only wish...

Jimmy Hoffa

Ampt

huh... I thought I only took 18 credits this past semester. Looking at my grades it was actually 20

Jimmy Hoffa

123 rep

@Ampt there are starving children in china...

user41796

@JimmyHoffa no, he would be a smart graduate. Especially since he took more 'learnin

Ampt

18:17

@GlenH7 technically I took the minimum amount of learnin required to get my bachelors but tomato tomato

Jimmy Hoffa

@GlenH7 so wasteful and inefficient... no wonder he still doesn't have 3k rep

Ampt

@JimmyHoffa I'm saving it for those starving kids in africa...

user41796

@Ampt Professional Engineers are defined as (at least) minimally competent in their field.

Jimmy Hoffa

@GlenH7 Medical Doctors are defined as (at least) minimally competent in their field.

that's a scary thought...

user41796

This is true

Ampt

18:19

@JimmyHoffa just like most government contractors. You bridges are built by the people who can do it the cheapest

user41796

As for the ones that didn't make the cut...

Ampt

safer and reliable have no dollar amount in that equation

Jimmy Hoffa

@GlenH7 dentists. har har har.

user41796

@Ampt Technically that's because they are presumed to be safe and reliable.

Ampt

then you end up with stuff like this: en.wikipedia.org/wiki/Hoan_Bridge

user41796

18:20

But it's no fun arguing over the semantics of "I have 4 equally sound designs for bridges, which one should I pick?"

Jimmy Hoffa

> The Hoan Bridge was temporarily closed on December 13, 2000, after two of the three support beams of the lakefront span failed, causing the north-bound lanes to buckle and sag by several feet and leaving the span in a near collapsed state. No motorists were injured when the bridge failed. [emphasis mine]

Sounds like good engineering to me!

user41796

Bad design:

user41796

> Experts believe that improperly designed welds between the lower lateral bracing and floorbeams

user41796

Of course, that's 20 - 30 years after the initial design

Jimmy Hoffa

@GlenH7 that's a pretty tiny detail to blame a larger failure on. I'd call the larger scope a success because the overall system staid solid through a failure in one of the smaller bits

user41796

18:23

But I'd be willing to bet it would have failed sooner had it actually been opened as planned

user41796

And I'm pretty sure I've never been on that bridge. I would have always taken 43 / 94

Ampt

Chunks of that bridge fall in the river below on the weekly

it is a terrible, terrible, TERRIBLE bridge

and I believe no one was injured only due to the fact that the issue happened during very low hours

user41796

Doesn't sound like I missed much by never having been on it...

Ampt

maybe death if you really wanted to experience that one

heard its better in fall though

user41796

@Ampt Yeah, the lake should be a little warmer by that point.

Ampt

18:33

@JimmyHoffa I feel like your definition of success may need to go in for some calibration

user41796

18:46

@MichaelT - I'm calling it. programmers.stackexchange.com/q/241254/53019

user41796

19:10

@psr the flag I threw on it was marked as helpful and the user has been removed. So I'm guessing that was a "yes"

Ampt

wait, what happened?

user41796

@Ampt "Too late." :-)

psr

Quick, delete everything, Ampt is onto us!

user41796

My link up above is now a 10k only link...

Ampt

@GlenH7 ain't nobody got time fo dat

user41796

19:20

@Ampt Highly likely our not-so-favorite troll was up to his usual hijinx again.

Ampt

why come back?

it's just bizzare

it's probably jimmy getting back at us for all the monad jokes

ratchet freak

trolling can be enjoyable

Ampt

@ratchetfreak trolling a small subbranch of the stack exchange network, where they just delete the question so you can't even get any feedback?

that's like some next level trolling...

in my experience you troll to see their reaction

user41796

@Ampt In the past he's manage to troll folk through the comments

user41796

But no one took the bait with this one and we got it nuked quickly enough

ratchet freak

19:25

yeah but a skilled troll can let his question stay open long enough for it to be reacted to

user41796

@ratchetfreak That one isn't that skilled.

Ampt

quit deleting stuff so damn fast!

durron597

@JimmyHoffa Thanks for a great response earlier

user41796

@Ampt Sorry. But crude comments from me get nuked by me fairly quickly.

durron597

All, I had a follow up question but I don't think I'm going to post it because it's too broad, but maybe it isn't, so I ask for the chatroom to vet it first

user41796

19:30

@durron597 We don't down vote as heavily in here...

durron597

(This occurs to me, codegolf.stackexchange.com has a sandbox question in meta, I wonder if that would be valuable for other sites)

@GlenH7 :)

Anyway, I wonder if there are any tips to debugging a problem that only appears in production but not in development

ratchet freak

logging

durron597

@ratchetfreak I had a lot of logging, I was just logging the wrong things

user41796

Log the right things.

ratchet freak

well now you know to log the right things

user41796

19:32

jynx

durron597

but that is eventually how i fixed it, i deployed a new version with EVEN MORE LOGGING™ and then found it

"How much logging should I do?" "More"

ratchet freak

well there is your answer

durron597

@ratchetfreak sometimes it's not so easy to deploy to production :)

user41796

When you get amazingly good you setup selective logging

user41796

so you can go to an admin panel of the app and tell it to log from component Foo but not component Bar

durron597

19:35

well logback has <configuration scan="true"> but it says that's very performance intensive

user41796

I like on-the-fly adjustments to logging. You avoid issues involved with constantly checking a config file.

ratchet freak

you could have the logging framework not check EACH log call but once a minute

durron597

> Given that ReconfigureOnChangeFilter is invoked every time any logger is invoked, regardless of logger level, ReconfigureOnChangeFilter is absolutely performance critical. So much so that in fact, the check whether the scan period has elapsed or not, is too costly in itself. In order to improve performance, ReconfigureOnChangeFilter is in reality "alive" only once every N logging operations.

> Depending on how often your application logs, the value of N can be modified on the fly by logback. By default N is 16, although it can go as high as 2^16 (= 65536) for CPU-intensive applications.

logback.qos.ch/manual/configuration.html

ratchet freak

it''s a just time() call and a compare to a static...

durron597

consider an app that has thousands or even millions of debug and trace calls

the overhead on one call if written properly is 20 nanos

ratchet freak

19:40

or just a load of a bool and have a separate thread do the checking

durron597

(20 nanos if the logger is off)

It already uses a separate thread

But hey, I'm just quoting the documentation, I don't actually know

ratchet freak

or they borked the config reading and it needs to be synced with the log calls

there are plenty of ways to derp a logging framework

Jimmy Hoffa

@Ampt Come now. Succesful engineering is obvious when you have a systemic failure with not but a simple service outage. The system ensured harmful faults did not befall it's users, and that is after all, the whole point of planning for failure, now isn't it? Many engineered pieces of software fail far more harmfully than that bridge did, causing the exposure or loss of critical data. A critical event if you will, that bridge did not undergo any such critical event, just an outage.

Ampt

@JimmyHoffa the bridge got lucky nothing more, nothing less

if an office building in the chicago skyline falls at night while no one was in it, is that a "Success" of it's engineering?

Jimmy Hoffa

@Ampt Did it? Do you know of it's design to be certain it was lucky and not a redundant fail safe system that ensured it?

Ampt

19:51

while(bridge):
    if(empty()):
        collapse()
    if else:
        stand()

Jimmy Hoffa

@Ampt No, but if it genuinely falls several feet, and then remains without having harmed anyone - that is pretty succesful is it not?

Ampt

@JimmyHoffa there was a several foot difference between two parts of the road. had someone been doing 65mph when that happened... well

you get the idea

I guess one way it wouldn't be so bad, but the return commute would be killer

rimshot

Jimmy Hoffa

@Ampt see! Though one part of the road failed, the other didn't, they managed to create the system such that components are allowed to fail independently. Do you know how little software has such attributes?

Ampt

no, I'm saying that if you were traveling down that road in one direction, you would go over a 3 foot drop

if you were going the other direction (on either side) you would hit a 3 foot wall

both sides were affected equally by the drop

user41796

@Ampt Dukes of Hazzard style?

Jimmy Hoffa

19:54

@Ampt that strikes me as a problematic occurrence.

Ampt

@JimmyHoffa It's a "success" remember?

@GlenH7 just. like. that.

or speed

Jimmy Hoffa

@Ampt :D I still stand by what I said. Nothing we engineer is perfect, remember that, but if we do our job right, failures will affect as little as possible

Ampt

@JimmyHoffa the hoan bridge is used by our structural engineering students of how not to design a bridge, ever

@JimmyHoffa If they had engineered the bridge to only fail at night, I would agree with you, but somehow, I have a hard time believing that that was in the reqs :P

Jimmy Hoffa

@Ampt probably, the thing was likely engineered for shit and lucky to have stood. Still I stand by my point that failures of the sort observed by that bridge are the best sort we can hope for in things we engineer.

Too many people engineer things to work without failure, when people should be focussing on making things work regardless of failure. Or at the least ensuring no one failure catastrophic beyond the rare scenarios that cannot be solved for. Like asteroids and moonites.

Ampt

4 hours ago, by MichaelT

@JimmyHoffa worse is better?

ratchet freak

19:59

failsafes should be the first thing on your mind when designing

Jimmy Hoffa

@MichaelT I only now realized you have the wrong link there. @Ampt you should be clear, the correct link is here

@ratchetfreak except in the event of moonites. There's no accounting for those jerks.

@durron597 didn't even realize you asked that; hope my answer fits with what you were looking for and gives some help in the future. If you have a build/cm guy at your company, set up a meeting with him to ask him about all the details of the build configuration and deployment environment. It'll help you to know how to setup your local environment to find issues more quickly in the future.

durron597

@JimmyHoffa as you'll recall my entire development team is me and a Mort

I was just pissed because I never test things as much as I want to... and then I finally did, and it still had bugs

ratchet freak

all software worth maintaining has some bugs in them

2

Jimmy Hoffa

@durron597 well if you only test minimally, at least make that minimum include some tests for actual bugs that you've encountered. At the least it ensures they don't end up recurring later.

durron597

@JimmyHoffa Oh I did that already, even before I made the post

Jimmy Hoffa

20:05

@ratchetfreak I always liked the phrase a friend of mine used: Show me a software company without legacy code, and I'll show you a software company without software.

durron597

@JimmyHoffa I've wondered if some departments at google manage to keep their code clean as a whistle

Jimmy Hoffa

@durron597 There's departments like that in many places, though they're far less common than the average, and they are merely departments - not entire companies. Usually those departments are only working on the newest bits of the system. Many companies have a team like that while the legacy is handled by others... software companies come in many shapes and organizations.

@RobertHarvey you asked was SSE was, I was just answering your question...

May 19 at 15:56, by Jimmy Hoffa

I'm helping.

I thought microwaves were the future in technology. You can cook a whole turkey in one over the course of a mere half hour to an hour! Craziness, clearly microwaves are the future in technology. — Jimmy Hoffa 7 secs ago

durron597

@JimmyHoffa s/helping/helping/

2

Robert Harvey

Scared him away.

Jimmy Hoffa

haha

@RobertHarvey to be fair, his question was...poor...

Robert Harvey

20:18

shyeah.

user41796

@RobertHarvey - you missed your favorite friend. programmers.stackexchange.com/q/241254/53019

Jimmy Hoffa

I DV'd and VTC'd. Though I hope he actually googles "vector programming" and learns, because it is neat, even if he thinks asking strangers to spoonfeed him is somehow more efficient than just googling

Robert Harvey

@GlenH7 Yeah. By the guy who's eventually going to build Skynet. It will be incrediblycut.

2

Ampt

Officially have an end date for the current job. Officially nervous for the new job.

Robert Harvey

Seriously, though. What does SIMD have to do with Websockets?

Jimmy Hoffa

20:22

@RobertHarvey but he's going to do it using only introductory level information as that's all he's requesting.

@RobertHarvey I was joking, clearly the SSE I referenced is a different SSE than he's talking about. I have no idea what SSE he was talking about.

Robert Harvey

Ah. A joke. I'm deadpan enough that sometimes people think I'm being serious.

Ampt

@RobertHarvey Look at how it reacts in it's natural environment. It doesn't see us as a threat, so we can move closer.

Jimmy Hoffa

hypervisor implemented entirely using introductory level knowledge and reference materials alone: Takes any VM and boots to a static screen that says "Hello World!"

Ampt

@JimmyHoffa I don't remember that in terminator...

Jimmy Hoffa

@Ampt it was in the prequel, hasn't come out yet.

user41796

20:24

@Ampt Director's cut

Ampt

@GlenH7 I hear it'll be incredible

Robert Harvey

This is going to require a bit of setup but... We have a rental home which has two occupants listed on the lease agreement, but like five people living in the house, including a live-in boyfriend. My wife called the property management company to find out why.

Her response via text message: I talked to Trish about the lease agreement we only need to have sexy 18 year old on that lease because the other two are minors.

Pretty sure it was a text typo.

Jimmy Hoffa

@RobertHarvey yeah, surely she meant miners. Dirty bastards, can't have them around without something pretty to class the joint up.

Robert Harvey

That's not the typo I meant. :)

"sexy 18 year old" is not something my wife would normally say.

I loled.

It would be equivalent to her saying "pretty young thing." And I'm pretty sure she would consider that sexist. But maybe only coming from a man.

Ampt

@RobertHarvey she doesn't like MJ? I think it's time to consider other options

Robert Harvey

20:28

Was MJ a man? Not too sure.

Ampt

@RobertHarvey depends on the year I believe. Still made some pretty good music

Robert Harvey

Yeah, it felt weird the day he died. "So this is what a world without Michael Jackson is like."

durron597

@JimmyHoffa By the way, speaking of worse is better, it was originally written in 1991. Windows 3.1 was released in 1992. Richard Gabriel didn't realize it at the time, but he was essentially predicting Windows while seeming to predict the rise of Unix and C++, and to this day the "virus" OS is still Windows and C++, which still fails to "do the right thing" in many scenarios.

Jimmy Hoffa

> But, at OOPSLA 2000, I was scheduled to be on a panel entitled "Back to the Future: Is Worse (Still) Better?" And in preparation for this panel, the organizer, Martine Devos, asked me to write a position paper, which I did, called "Back to the Future: Is Worse (Still) Better?" In this short paper, I came out against worse is better. But a month or so later, I wrote a second one, called "Back to the Future: Worse (Still) is Better!" which was in favor of it. I still can’t decide.

Robert Harvey

I'm still firmly convinced that the only reason you don't see more mac viruses is because it's got such a smaller bullseye painted on it.

Ampt

20:36

@RobertHarvey that's changing, very quickly

user41796

@Ampt iOS yes, macOS not so much

Robert Harvey

For iPhones and iPads, sure. And the story is better there, since the gestapo runs the store.

user41796

@RobertHarvey I know of somebody who claims to be really good at assembler who would love to port the xcode process to a non-mac system if you're interested.

Robert Harvey

Will it be incrediblycut?

user41796

Would you expect anything less?!

Jimmy Hoffa

20:38

Back to the future: Is worse (still) better? - Back to the future: Worse (still) is better!

Robert Harvey

heh.

user41796

oh, and the efficiencies will stretch down to the picosecond scale.

durron597

Huh, just realized something fun. Maybe many of you already realized this. Why was 2014-05-13, 16:53:20 UTC an important date?

Jimmy Hoffa

@GlenH7 because nano is bullshit.

Ampt

@durron597 Rollovertime?

Jimmy Hoffa

20:40

@durron597 ... ? Of what year? did we start using the sign bit in a short epoch or something?

Robert Harvey

@JimmyHoffa Talk about a bunch of hand-waving, wandering generality sound bites. Hasn't the story improved with optimizing compilers and faster machines anyway?

Jimmy Hoffa

@RobertHarvey you're referring to something far different than what Richard Gabriel talks about

for context, you need to read the original Rise of worse is better

Robert Harvey

He's talking about writing high-level prototypes in high-level languages.

Jimmy Hoffa

@RobertHarvey he's not talking about performance though. He's talking about design and maintainability and end-product-robustness along with end-product-reachedness

Robert Harvey

I haven't read it in detail yet, but it still feels like an academic writing about what he only knows about in an abstract way.

Jimmy Hoffa

20:42

@RobertHarvey absolutely is not the case. Read the original and follow along. He most certainly is not an academic.

Robert Harvey

[sigh] Alright.

@JimmyHoffa This is a tough slog. For some reason, his style of writing doesn't hold my interest. Or maybe it's just too much coffee today.

Jimmy Hoffa

His basic notion is about arguing for both sides of the normal debate we all have often between the merits and validity of the get-r-done approach and the get-r-perfect approach. In his original essay he says in no uncertain terms that trying to get it perfect just means you'll never release the product when someone who's doing a get-r-done implementation will have all the market share long before you don't finish your product.

@RobertHarvey likely his style of writing. He doesn't write like many others, so surely it's a bother to some... I always thought it flowed nicely, but it definitely doesn't read like a normal article

Robert Harvey

@JimmyHoffa I agree with that. But then... "C and Unix are the ultimate computer virus." WAT?

Jimmy Hoffa

@RobertHarvey gotta remember the time. Windows/dos wasn't super common at the time. C/Unix followed the get-r-done approach he refers to

(surely you know more about the tech of that era than I...)

Robert Harvey

There's a difference between Architecture Astronauts and bare-metal C/Unix coders. But there's also a difference between the C/Unix coders and the script kiddies, and the difference is not insubstantial.

So while I do agree with the git-r-done approach, I don't agree with the "throw spaghetti against the wall and see what sticks" approach.

Or the "will it blend" approach.

user41796

20:50

I didn't even think skiddies and C coders were even in the same galaxy....

Jimmy Hoffa

@RobertHarvey maybe now, would you say the same of '89/90 ? C/Unix coders were writing the procedural basic non-extensible, non-abstract, code with little to no semantic value then. That's basic get-r-done. They were writing NCubes.

Robert Harvey

And yet, you couldn't get anything done without their code.

user41796

@JimmyHoffa meh, that's a bit bold of a claim

Jimmy Hoffa

@GlenH7 granted, though there is precendence I think which you will find working with any coders who were writing this procedural mess back then - you can still see it in their code today often times. They think machine control flow is the whole of programming

@RobertHarvey which he argues, is largely because people who wrote code the opposite way - well their code basically never materialized. If you look at the percentages, that's also vastly true (then, and today).

Ampt

Jimmy Hoffa right now:

Robert Harvey

20:53

@JimmyHoffa But who is writing ivory tower code of that kind anyway? Certainly not anyone I know. Even you.

Not production code, anyway.

@RobertHarvey would I be able to "create one websocket to handle a notification system, a chat system, a group system" in only ONE websocket? — user3385828 6 mins ago

What is this guy missing?

Jimmy Hoffa

@RobertHarvey which is the premise of his whole essay. Production code doesn't get written like that. People who write code like that just go on and on, but never actually reach production. So no, code like that isn't written for production (often because people recognize from the outset they won't finish it in time because the amount of design and analysis it would take to get it done - somebody else would get it done first)

user41796

@RobertHarvey he's wrapped around too many axles

Robert Harvey

@JimmyHoffa Seems self-evident to me.

Jimmy Hoffa

@RobertHarvey remember, he wrote that essay to lispers basically saying "Hey guys, chill the eff out and just make shit work rather than perfect. Sorry but your expert systems will never become self aware, at least make them pick movies well"

or whatever it is their expert systems did

Robert Harvey

Hmm, but he's comparing Common Lisp (arguably the "practical" Lisp implementation) with Scheme (arguably the ivory tower Lisp).

Jimmy Hoffa

20:57

@RobertHarvey then read the counter-argument.

like he said, he's not sure which way is right. Just like none of us are ever sure if we should make some things more abstract or less abstract at times. It's a hard decision and we must go with our gut more often than we wish.

in '00 he argued for both sides, and ended it saying he still doesn't know which is right.

@Ampt want to see a great article about the ID monad? What about the null monad?

Ampt

sure, hit me

Robert Harvey

@JimmyHoffa Read it. Still sounds like so much hand-waving to me. His original article is better.

user41796

@JimmyHoffa Rule 0 Jimmy, Rule 0.

Robert Harvey

We're back to my original argument. Compilers and fast machines have gotten good enough that precise data structures and exact algorithms no longer matter most of the time.

user41796

@RobertHarvey That's been true for quite some time

Jimmy Hoffa

21:01

@RobertHarvey from my recollection he never talked about those things. His point was about design approach and release-readiness vs. extensibility and maintainability decisions.

@Ampt learnyouahaskell.com/a-fistful-of-monads

user41796

@Ampt - see, I told you he'd ignore the rule regardless of what number it was.

Ampt

@GlenH7 maybe he's not playing golf?

Robert Harvey

"Many high-level programming languages hide detail by providing big abstractions. This means that
you are working with modular parts and not with small pieces constructed by hand. For the right
level of close detail you need a low-level language like C, assembler, or C++. With these languages
you can control data layout and design very precise and exact algorithms. With Lisp and Smalltalk,
for example, you are working with their big abstractions in a ham-fisted way—just fine for prototyping
and understanding larger issues, but not good for minute algorithm and data structure design.

Jimmy Hoffa

@GlenH7 I've stuffed myself into a monad with no rules. Your impure rules can't side effect me in here.

Ampt

rule 1,000,000: Don't talk about monads

Is assembly commonly called assembler?

I've only ever heard it called assembly

Robert Harvey

21:04

They're used interchangeably, although assembler is supposed to mean the program that reads the assembly.

Jimmy Hoffa

I've half a mind to change my nick to JimmyMonad, but then I'd be afraid someone would try to >>= me

Ampt

@JimmyHoffa You'd only be allowed to talk when needed, as per monadic rules

Jimmy Hoffa

@Ampt just because comonads have costrength while monads only have strength doesn't mean they can't be eager. It just means eagerness is unintuitive given their attributes.

if laziness were demanded, monads would only have >=> instead of >>=

Strong monad

In category theory, a strong monad over a monoidal category ({\mathcal C},\otimes,I) is a monad (T,\eta,\mu) together with a natural transformation t_{A,B} : A\otimes TB\to T(A\otimes B), called (tensorial) strength, such that the diagrams :, , :, and : commute for every object A, B and C. Commutative strong monads For every strong monad T on a symmetric monoidal category, a costrength natural transformation can be defined by :t'_{A,B}=T(\gamma_{B,A})\circ t_{B,A}\circ\gamma_{TA,B} : TA\otimes B\to T(A\otimes B). A strong monad T is said to be commutative when the diagram : commutes fo...

Robert Harvey

Do I really need to know what all those weird symbols mean to understand monads?

Jimmy Hoffa

@RobertHarvey not at all. It's like programming, you don't need to know every languages foreach syntax to know how foreach works. Those symbols are just one language for expressing the concept. Others are less arcane and more suited to your past experiences (like one written in C# I showed you and you agreed was fairly clear and simple)

Robert Harvey

21:09

OK.

Still into rather mundane concepts in LYAH.

[starting to wonder what all the fuss is about]

Jimmy Hoffa

@RobertHarvey yeah, it is slow... but it's kind of good that it is because when it starts doing weird things, you'll begin referring back to the earlier stuff to figure out how it derives the more complex stuff.

Robert Harvey

Fair enough.

Jimmy Hoffa

alternatively it might be a waste for you. Many say they learn better from RWH than LYAH.

I can't know.

Ampt

@JimmyHoffa is that knowledge stuck in a monad?

Jimmy Hoffa

@Ampt like schroedinger's car, that knowledge hasn't been observed yet so it is yet unknown.

Ampt

21:21

@JimmyHoffa It is both locked and not locked at the same time?

Jimmy Hoffa

@Ampt that's a stupid deduction, schroedinger was an idiot too arrogant to just utter a few simple words: I don't know.

Robert Harvey

It does make one wonder what the significance of an unobserved state might be. Collapse of the wave function, and all that. Some spiritualists ascribe mystical qualities to these wave functions (like consciousness), an assertion that the physicists vehemently deny.

Ampt

@RobertHarvey that still boggles my mind... observing the particle changes it's behavior

how does it even know it's being observed?

Robert Harvey

The only way you can make observations at that scale is to bounce particles off each other, and infer behavior by observing the resulting collision tracks, so the idea that observing it changes the system should be no surprise to anyone.

Ampt

@RobertHarvey I believe that the idea was that you could see where the particles ended up at, as in the slit experiment?

Double-slit experiment

The double-slit experiment is a demonstration that light and matter can display characteristics of both classically defined waves and particles; moreover, it displays the fundamentally probabilistic nature of quantum mechanical phenomena. This experiment is sometimes referred to as Young's experiment.While there is no doubt that Young's demonstration of optical interference, using sunlight, pinholes and cards, played a vital part in the acceptance of the wave theory of light, there is some question as to whether he ever actually performed a double-slit interference experiment. * The expe...

user55340

21:33

►

Robert Harvey

Yeah, the paradox of the slit experiment is that you can fire photons at the slits one at a time and still get the same result. In other words, how could the second photon interfere with the first, if it had already arrived?

user55340

Copenhagen interpetation!

user55340

Copenhagen interpretation

The Copenhagen interpretation is one of the earliest and most commonly taught interpretations of quantum mechanics. It holds that quantum mechanics does not yield a description of an objective reality but deals only with probabilities of observing, or measuring, various aspects of energy quanta, entities that fit neither the classical idea of particles nor the classical idea of waves. The act of measurement causes the set of probabilities to immediately and randomly assume only one of the possible values. This feature of mathematics is known as wavefunction collapse. The essential concep...

Robert Harvey

But that just means there's some underlying mechanism that we don't fully understand. It doesn't mean that the two particles know about each other, in some sort of consciousness way.

user55340

Why are you saying the two? One interpretation is that the particle interferes with itself from another 'universe'

Robert Harvey

21:37

Alright. Excuse me while I go order a drink from that other universe.

user55340

I'd suggest watching that video...

Ampt

@RobertHarvey doesn't evil Robert Harvey have to pay for it then?

Robert Harvey

I have a beard in that other universe. :)

user55340

Evil Robert Harvey has a gotee.

Ampt

If I have a goatee, does that make me Evil Ampt?

Robert Harvey

21:38

My go-tee is the one I wear jogging.

Ampt

or is evil ampt clean shaven?

user55340

> Long-term versus short-term facial
hair is a very important distinction as short-term facial hair, also
known as the temporary illness "goatee universitis" (which symptoms
range from full goatees to the less popular chin-goatee) is a common
affliction for university-based males.

Robert Harvey

Anyone have a calculator? My debugger is saying that (int)834 / (double)1050000 is (double).002 ... WAT?

Ampt

.0008 here

user55340

bc 1.06
Copyright 1991-1994, 1997, 1998, 2000 Free Software Foundation, Inc.
This is free software with ABSOLUTELY NO WARRANTY.
For details type `warranty'.
scale = 20
834 / 1050000
.00079428571428571428

Robert Harvey

21:40

[sigh]

psr

A lot of the quantum stuff is almost like conditional probability - if you don't see whether the runner got thrown out at second the game is in a "superposition" of states in which there is one out and nobody on base and in which there are no outs and a guy on second.

A lot of things about it make sense that way - but quantum "probability" is this weird 2 dimensional probability in which squares of things that have no real world name add up to 1, rather than 1 dimensional probabilities summing to 1.

Robert Harvey

Yeah, that's what I'm getting on my calculator as well. What the hell.

Ampt

@RobertHarvey did you check your quantum field generator? Sounds like the cat is dead

Jimmy Hoffa

@RobertHarvey like I said, all Schroedinger had to say was "I don't know". Instead he derived a formula that proved he didn't know. Talk about taking CYA to a whole new level.

user55340

(I'd still watch that video that I linked... its the Feynman explanation of the double slit paradox... and its not mathy at all)

Robert Harvey

21:42

@JimmyHoffa Except that his formula is borne out by observations. Wave function collapse is a real thing.

Jimmy Hoffa

@MichaelT Feynman is great at that.

Robert Harvey

@MichaelT I'm getting to it. First things first. Why does .002 equal .008, for example.

Jimmy Hoffa

@RobertHarvey just because he deduced he didn't know something, and proved it by observation, doesn't change the fact that he still doesn't know.

user55340

Thats too mathy. Watch the video instead.

Ampt

@RobertHarvey sounds like your double precision sucks

Robert Harvey

21:43

@JimmyHoffa Well, you kinda have to have some insight to come up with a formula that actually works. Otherwise... Well, y'know. Guessing.

@Ampt I think there's a cast somewhere that's breaking the math.

Jimmy Hoffa

@RobertHarvey if he just guessed it would have been really much less interesting. "If you put a cat in a thing with a thing, it will be ded." see, hardly has the mystique

Robert Harvey

Usually when I do these kinds of calculations, I use decimal if the range isn't too extreme. Solves a lot of otherwise thorny problems.

I hate floats.

psr

Currently working with code that uses Hungarian notation: sh, lng, and i all mean short, long, or integer. And I mean "all", not "each (respectively)" - which of the 3 each means in a particular case is entirely random.

Ampt

why is 1050000 a double anyway?

lmao what?

short == long == integer?

Jimmy Hoffa

@psr haha, proof of the value of mumps: It doesn't support hungarian notation!

Robert Harvey

21:46

@Ampt Because the original source is a text field that allows scientific notation.

It's a bit rate.

Ampt

@RobertHarvey text field now only allows integers. problem solved!

Jimmy Hoffa

@RobertHarvey bi-trate.

Robert Harvey

Bitter tater.

Ampt

@JimmyHoffa why would you do that?

I'll never say bi-trate properly again

Robert Harvey

At least he didn't Rick Roll you.

Oh, wait.

Ampt

21:47

[not clicking on links for another 24 hours]

Jimmy Hoffa

@Ampt The answer to life, the universe, and everything

Robert Harvey

This is me not clicking on that.

Ampt

@JimmyHoffa to live? what is it, PSN?

Jimmy Hoffa

@Ampt PowerShell Network? You're really taking this Embrace-Microsoft-DreamSpark thing a bit far, aren't you?

Ampt

@JimmyHoffa Playstation network. You know, the answer to Xbox Live?

Jimmy Hoffa

21:54

@Ampt just think ti-trate whenever you're trying to figure out something's transfer rate. Especially if you're speaking publicly.

Robert Harvey

Cast everything to decimal, still get .002. Write literal expression in immediate window, get the correct result.

What. The. Hell.

Jimmy Hoffa

@RobertHarvey ? What's the code you're fiddling with?

Ampt

@RobertHarvey And they say that software isn't magic

Jimmy Hoffa

@Ampt magic is way cooler than software, haven't you ever seen someone get out of a straight jacket? If somebody thinks we just dunk ourselves in a tank, shake and shimmy for a minute or two, then emerge with an enterprise product, they're clearly having too much fun.

user55340

@RobertHarvey ideone.com/fjPgcf

user55340

21:59

I assume you're dealing with C# there.

Jimmy Hoffa

@MichaelT what about that code is supposed to be surprising?

user55340

@JimmyHoffa Nothing... just that its not 0.002.

Jimmy Hoffa

@MichaelT Why would it be?

user55340

4 mins ago, by Robert Harvey

Cast everything to decimal, still get .002. Write literal expression in immediate window, get the correct result.

Transcript for