The Whiteboard

our Javascript code is in a github repo with a webhook that automatically runs all the tests on every PR, so we learn about issues like that (along with many more a traditional compiler could not catch) before anyone even hits merge

</holy war>

you do everything in remote branches?

well, local branches which we push to our forks of the main remote

I'm not sure if that's a yes or a no

so your workflow is

make a change

push to remote branch

all tests run

you find out if basic sanity checks have passed

then you do merge

both of my changes were correct, the problem is that i just didn't push one of them

yeah, in your example the more direct solution is have the code reviewing tool do a merge on acceptance, or use the Merge button as if it were the Accept button

user55340

12:14 AM

C++ advice? or Life advice?

user55340

@durron597 so ... push it. No reason it can't be pushed later after it's already been approved is there? Just notify the folks who were aware of the code fix that you mistakenly didn't push it on time but have corrected that by pushing it now; and ask if there's further tasks you need to do regarding it

you keep saying button

@JimmyHoffa i did push it, but i only noticed because i was doing a manual test

what i want is to get an email

@durron597 I suppose you could have a command line code review tool, but...I'm not sure anyone wants that

an e-mail about what? did automated tests fail and you didn't get notified?

12:16 AM

you don't have a "build the software" test?

a compilation error isn't a test - that's an outright build failure. Do you not have automated build or testing?

Ok so I think I understand completely

here's what I'm understanding

@durron597 broke the build.

2

worth it

we have three different services that serve javascript content

You need a CI suite, talk to @Ampt, he's still searching for his, maybe you two can help eachother out. No really though, do you guys have zero CI setup that you're aware of?

we have a CI suite but the problem is that package A is served to the user by one content provider and package B is served to the user by a different content provider

we have internal Jenkins and Github instances, PRs on the Github trigger Jenkins jobs which do builds and/or tests and when they pass a little green check mark appears on the PR

no experience with other CI tools though

@durron597 can't you have an integration test with both packages?

12:19 AM

@Ixrec they all basically function around the same simple set of techniques; Jenkins is somewhat good in how simplistic it is - good example of the basic idea of what's needed.

@Ixrec that would make sense, wouldn't it?

@durron597 so you have code that wouldn't "compile" -> What does that even mean in your terms? Was it JavaScript? What was "compiling" entailing? Just initial page-load execution?

@JimmyHoffa all the issues we've had with Jenkins boiled down to getting its user account the permissions to do all the things we wanted it to test, which I assume would be just as fiddly with any other system; I agree it's quite simple once it has the permissions

@JimmyHoffa compile is the wrong word in a JS context obviously

I meant TypeError

object.name is supposed to be a function () but it's actually undefined

user55340

That involves that thorny problem of "how do you actually tickle that code in Javascript as part of automated tests"

12:21 AM

@JimmyHoffa our JS code does have what I sometimes call "load-time errors" when all the modules are being evaluated, such as one module depending on a module that does not exist, or a syntax error in the initialization code

@durron597 so why don't you guys just use some JS test framework? There's phantom and other such DOM-faking approaches

our QA person is focusing all her energy on selenium based tests in Java right now

@Ixrec you just mean initial interpretation -> Not actual execution beyond the in-order execution of a given JS file

user55340

Btw, fun reddit thread on php 7 imminent release: reddit.com/r/programming/comments/3v4l98/php_7_released

i guess when you create objects and methods dynamically it's actually quite difficult to know if an object's method exists

12:22 AM

@durron597 and basically there was no coverage for this because you guys only have coverage over Package A?

once again my hatred for javascript burns with the rage of a thousand suns

@JimmyHoffa yes, though our framework enforces some constraints on module dependencies so some of those errors feel a little like a compile error

@JimmyHoffa the selenium tests are end to end

@durron597 assert(object.name !== undefined); is a fairly easy test

@durron597 welcome to why TDD was really largely constructed in the world of dynamic languages. Unit testing benefits them far more than static languages as often they have basic unit tests to just verify type-correctness since there's no compiler to do so.

12:23 AM

if a user does X with a page, does Y happen or not?

@Ixrec that will only be true after a certain point in code execution

especially if you are adding methods to objects on the fly

besides, assert only does so much if you are not actually listening for it.

well, obviously the real test would be more like addAllTheMethods(); assert(object.name !== undefined); assert(object.name() === 'test');

@durron597 so start using mocha and writing your own unit tests for just your own code if you really want guarantees on it. Also, for gods sake do not ever think of JS things as "object"s as @Ixrec just encouraged

only do that if you like hating dealing with JS

@Ixrec this assumes that addAllTheMethods() is a writeable function

cue obligatory lecture on how testable code === good code

cue obligatory lecture on how being able to define interfaces on the fly adds nothing useful and makes code impossible to test

12:28 AM

which is why we usually don't do it

usually?

I'm trying to think of a place where we do it and I'm not coming up with anything

unless you mean building new objects that fulfill the interfaces, but we always use the same constructors for that

@durron597 cue obligatory lecture on how referentially transparent methods can do any-damned-thing they want with their local state - such as using runtime-decided-structures while remaining highly unit testable

user55340

@Ixrec javascriptiy... assert(goodCode instanceof ITestable)

@durron597 selenium is neat

12:31 AM

it is, but when I asked the QA person how I would use it to look for TypeErrors her answer was "we can parse the console log"

this is why end-to-end tests that include the UI layer should not be the sole testing strategy

@Ixrec as I said above: TDD is far more beneficial to dynamic languages.

I'm not sure if I agree with that, we have just as many stupid bugs in C++ code when we forego tests as we do in JS

@Ixrec C++ is a dynamic language...

user55340

@durron597 thats "fun" code!

12:34 AM

strong is the last word I would use to refer to it's "type system"

@JimmyHoffa oh, well that changes everything

@durron597 I once worked on a product which was created in that way; except it was over 20 years by a rotation of fresh grads... :D You want to talk about bafflingly bad code...like whoa..

by your definition, are there any static languages other than purely functional ones like Haskell?

@Ixrec C# - I believe Java requires you to walk through hot coals to step out of the type system as well

many others I'm sure

@JimmyHoffa I think the entire java.util.reflect package should be scrapped

okay that might be pushing it a little, but you see my point

12:37 AM

@durron597 eh; reflection isn't an out for a type system so much as "Here's my memory space, do as you like" is as C/C++ (yes, that thing which isn't a thing! :D )

dumb question from a non-java user: do JVMs have a special "you're allowed to do stupid things like reflection" mode you can turn on and off?

you can do the same in C# handing around memory spaces, but you need to put all code working anywhere near that in unsafe blocks and turn "unsafe" on during compilation to allow it

user55340

@Ixrec Nope. Always on.

security!

user55340

And sun.misc.unsafe is there too... at least in most jvms.

12:39 AM

@MichaelT so you can do direct memory access throughout Java without any special compiler settings?

user55340

A cow-orker at Employer^^ wanted to write a language where there were two "levels" to it. There was "framework language" and "business language". Framework language let you play with reflection and all of those dirty bits... while the business language was kept rather clean.

man 185 question personality profiles wat

@enderland O_O for hire?

JVMLS 2015 - The Secret History and Tragic Fate of sun.misc.Unsafe

@MichaelT aren't they talking about phasing it out?

user55340

►

12:40 AM

@JimmyHoffa yeah

well maybe, maybe just for interview :P

this personality profile is terribad

I'm really curious what my "results" will be franklyl

user55340

At Employer^^ I had to take a 'dumb test' prior to the interview. It was to weed out people who would answer 'yes' to "Its ok to get into fights if you are on break" or "Its ok to shop lift less than $10 if you update the inventories." Apparently some people actually failed it.

yeah there are a few questions in here like that too

@MichaelT reminds me of thedailywtf.com/articles/What-Should-He-Do,-Indeed

some of these questions are at least interesting though

> Which number should come next at the end of this series: 1/12, 1/6, 1/3, 2/3?

> Which should come next at the end of this row of letters: E I L?

i can think of multiple valid answers to the second one.

hell there's probably multiple valid answers to the first one too

12:55 AM

> a. 3/4
b. 4/3
c. 3/2

user55340

@Ixrec ...

at least the first one has an obvious "most correct" answer

> a. M
b. N
c. P

the second one also has an obvious answer, too, beacuse it's

is this one of those tests where you can get an A by choosing b every time

4 E 3 I 2 L 1 (N)

@durron597 actually the last 5 are all b, which is making me double check them :P

user55340

12:56 AM

We had a c. 1960 SAT test as part of the interview process too. When I described it to my father he recognized it - especially the estimation questions which were essential with a slide ruler. It did have an interesting aspect to it - the verbal section weeded out many people who were less skilled at communication.

@enderland I'm missing something, where does any of this come from?

@Ixrec abcd E fgh I jk L m N

I think he understood the puzzle, what he didn't understand is where the question came from

(i'm wondering the same)

oh

17 mins ago, by enderland

man 185 question personality profiles wat

user55340

Another piece of advice for the test was "before you start, write down on the scratch paper all the letters of the alphabet and the numbers 1 - 26 under or next to them"

12:58 AM

I understand the solution now, not sure if I ever would've noticed that on my own though

@Ixrec I figured it out instantly, but I chalk it up to experience with these kinds of puzzles

@Ixrec sounds kinda like the classically proven-worthless "puzzle tests prove teh smartz!" interview approach which is...bleh.

yeah, most puzzles are mostly a test of whether you've seen similar puzzles before

more than anything else

@durron597 yeah these are really dumb for that reason, at least the number ones are more "obvious if you are smart" than the stupid alphabet ones

I'm pretty sure I got 100% on them

12:59 AM

@JimmyHoffa pretty much, yeah

what comes next in this series:

@JimmyHoffa it was about 15/185 questions or so, lots of them were like:

> I greatly enjoy inviting guests over and amusing them.
a. true
b. ?
c. false

A B D O P

more standard personality profile stuff

@enderland ... what?

1:00 AM

@enderland what Jimmy said

@JimmyHoffa "amusing" is an... amusing word lol

"true! I love having people over and tormenting them and amusing them with horrible things"

@enderland b) because it's missing a comma

@enderland ? what

a. G b. Q c. R d. Y e. nothing, the sequence is over

@durron597 could be R

seems reasonable at least

@enderland i see where you're going with this but that particular pattern is pretty arbitrary, and i wouldn't do the same puzzle as the last one again

(R is not correct. but a better, less ambiguous puzzle would not have had it as an answer choice)

1:05 AM

I'm curious now :P

i'm giving people time to think about it :)

I was thinking that without another character it's not totally clear, but another two would make it quite clear if R was right :) ;)

enderland won't say what the answer is

okay here's a kind of big hint

enderland knows what it is :o

the first seven characters are

what's the next character? no multiple choice now

1:12 AM

actually that'd be an interesting question, given that sequence, what is the analogous sequence for lower case letters

oh, i screwed up anyway

hahahha

oh no i didn't

26 secs ago, by enderland

actually that'd be an interesting question, given that sequence, what is the analogous sequence for lower case letters

;)

yeah i confused myself about E for a second.

or G

right

anyway, to those of us not named enderland or durron597, the answer is:

here's a similar yet different sequence

A E F H I K L M N

what comes next?

here? no i didn't.

correct

1:15 AM

though this is why quizes about this sort of thing are dumb, lol

the key to puzzles like this is just interpret the input data differently than the typical form

both puzzles are about the shape of the letter, not about the letter itself

yeah. I was "primed" since all the ones on the quiz I took were number/index based

here's a different one:

which would be an interesting questino for puzzling, related to research about how significant priming is on people's ability to do subsequent puzzles in a mix of "index vs character shape" ones

B C D E G P

what comes next?

1:23 AM

that one seems trickier

remember all the puzzles are the same, the key is that the letters don't have their traditional meaning

the first few were about letter index-as-a-number

the next few were about letter shape

this is one is yet another context

(the easiest puzzles are ones where the sequence DOES have it's traditional meaning, such as A E I O what comes next?)

1:49 AM

& obviously comes next

AEIO&

PROVE ME WRONG

ok you guys have been over this already

@LightnessRacesinOrbit this is obviously correct because it's the only reasonable way to dereference the result effectively returning ⊥

@JimmyHoffa uh huh

nn

(I see the broken reveal-deleted-messages JavaScriptlet has now been pin-starred. Why not pin-star my fixed version?)

@LightnessRacesinOrbit do give to me, as the broken version is ... not working

@JimmyHoffa I wrote it on the StackApp several weeks ago

along with a detailed explanation

@LightnessRacesinOrbit link

All you complaining about deleted messages who can still read them should use this

1:54 AM

Nov 14 at 21:37, by enderland

Also, change http://chat.stackexchange.com/messages to /messages So that it works with (a) HTTPS, and (b) SO chat. — Lightness Races in Orbit Nov 14 at 22:12

etc

user55340

4:55 AM

From 2001: "I will eat a week's pay if OOP is still in vogue in 2015." http://web.archive.org/web/20010310074457/http://www.geocities.com/tablizer/oopbad.htm https://t.co/gvpeNHVC2j

Tweeted by fernozzle on December 2, 2015 at 7:19 PM

user55340

That's why we shouldn't make any estimates for longer than a release cycle.

user55340

The only people allowed to talk of 2030 are plan 9 developers.

user114359

6:16 AM

@MichaelT I hope I am not alive in 2030 either because of zombie apocalypse or because of epic story starting with "we were arrested..."

A: Design Review: on-topic or not?

6:34 AM

Hey there

This was just posted to CR, but is too conceptual/whiteboard to be on-topic, was just wondering if you guys would think it would be on-topic on Programmers

SRP (Single Response Principle), going too far?

Damn thing won't one-box...

user114359

6:58 AM

@Phrancis question is deleted, can't one-box regardless of anyone's rep. Also, I don't have 10k on CR so I cannot see deleted questions there.

user114359

You can see your own deleted questions regardless of rep. I recommend posting on our meta and asking for feedback. Say "I want a design review: is this acceptable, or if not, what should I improve?"

user114359

Read this first: Design Review: on-topic or not?

user114359

12

I think that, generally, design review type questions are on-topic. However, the problem is how broad they are. My concern for this type of question is that most of them may be more suited to a discussion environment. I do think that there are good design review questions, but they need to be cle...

user114359

24

A: Migration paths to other sites

If you come across a question asking for help with a design as opposed to the code, that is technically on-topic at Programmers and may be a candidate for migration there. Programmers welcomes design review questions, but they have to be reasonably scoped with clear design goals. "Critique my de...

A: In this day and age why are web and database separate systems?

OK, didn't realize it was deleted, page hadn't refreshed. I'll check out the links though, thanks @Snowman

user114359

7:05 AM

@Phrancis thank you for asking for help first. We want good questions, and being constructive like you are helps us all

user114359

I am about to collapse from lack of ~~sleep~~ coffee but will look forward to you what you have to say in the morning.

Change

8:25 AM

hi

would disposing class X will dispose it's base class too ?

Change

8:49 AM

it doesn't in .net

msdn.microsoft.com/en-us/library/ms182330.aspx

Yannis

9:12 AM

1

Servers are designed to do one thing, and do that one thing very well. Relational database systems have had decades to mature, and so have web servers. Web browsers perform an entirely different function for which they are uniquely suited. These systems are interconnected via systems of commun...

^^^ Flagged as offensive

10:42 AM

Doesn't this feel like nannying to anyone else? Adding a restriction to a piece of software as some kind of emotional therapy seems a step too far to me. If I've upvoted a comment by mistake, found an error in it, pointed it out, then someone edits the comment to correct it, I want to be able to re-upvote it, and I see no actual reason to be prevented from doing so other than a sort of condescending "stop wasting your time" social engineering measure. — Lightness Races in Orbit 1 min ago

11:01 AM

@Yannis lol

Because any half-decent programmer appreciates the value of modularity (both for maintenance, security and logistical reasons), and because you haven't shown that the overhead of IPC between web server and data storage backend (which, as far as I can tell, is your only true complaint) is problematic enough to warrant violating such fundamental software principles. — Lightness Races in Orbit 51 secs ago

overexchange

12:30 PM

@Ixrec Can you help me on this query? stackoverflow.com/questions/34064297/…

1:39 PM

Happy Coffee Day

@MichaelT heh, well, he is right..

@Change up to the implementer of the class to make their dispose method call the base classes.

@JimmyHoffa have you watched Nick Offerman silently drink whiskey next to a fireplace for 45 minutes yet?

@ThomasOwens no, but I have silently drank whiskey next to a fireplace for 45 minutes before, and I can absolutely say it is fantastic.

@JimmyHoffa Well, I think it's on YouTube if you're interested.

I want to watch it soon.

And if I had a fireplace, I'd reproduce it.

@ThomasOwens I have one, and a load of fire wood as well... hrmm...

@Snowman I looked through the design-review on meta and I can say that one Q that was posted to CR that I linked would almost 100% likely failed #3 on that answer by gnat "Is it actual design from a project rather than pseudo-design or example design?"

1:45 PM

@Phrancis I have concerns about 3, simply because of the license that you grant to SE when you post here. I don't think that you need an actual project design.

It was very "example-ish"

I say go for it. Ignore 3.

I'm going to try and get a screenshot, if just for learning. OP deleted the question

As long as it was close enough to actually help you and others. In fact, it may be more helpful if it's generalized a little.

Oh.

(it was not my Q)

1:47 PM

> I took the time to mark every time Nick Offerman took a drink.
0:27 - Nick Offerman takes a drink. 1:36 - Nick Offerman takes a drink. 3:10 - Nick Offerman takes a drink. 10:27 - Nick Offerman takes a drink. 20:27 - Nick Offerman takes a drink. 20:45 - Nick Offerman pours a drink. 28:22 - Nick Offerman takes a drink. 30:14 - Nick Offerman examines a drink. 30:21 - Nick Offerman takes a drink. 30:34 - Nick Offerman takes a drink. 36:20 - Nick Offerman examines a drink. 36:30 - Nick Offerman smells a drink. 36:38 - Nick Offerman takes a drink. 38:39 - Nick Offerman takes a drink. 41:06 - Ni…

wow, some people...

By my count, that's 14 drinks. In 45 minutes; one drink every 3 minutes.. that's too long. 2 would suffice, but every 3? The suspense; no bueno.

@ThomasOwens It was pretty general, in fact it looked like almost verbatim what their professor was explaining about OO, but they had concerns that perhaps some applications of those principles, in some practical use cases, may be "overkill"

the hell? Nick Offerman was in the original Sin City.

Nick Offerman's 'Yule Log'

►

@ThomasOwens the most important part of the whole thing is that A) It's Lagavulin; a genuinely great scotch, and B) How much he's drinking; which has yet been undetermined. Though apparently he get's a refill so if it's a correct pour, he's having a proper night's ration.

talking about scotch Thursday morning already???

1:58 PM

@enderland We're talking about Nick Offerman sitting by a fireplace, silently sipping Scotch.

For 45 minutes.

not sure if that's better or worse... ;)

@enderland this is plainly not my fault. Seriously, I said Happy Coffee Day and @ThomasOwens went straight for the Scotch. I gotta tell ya, I'm worried about him.

HAPPY SCOTCH DAY

2

I love this chat

Duga

2:15 PM

Maybe this should be on programmers.stackexchange.com? — roryap 14 secs ago

@ThomasOwens ^^ (and @Snowman )

Q: Why do interview questions make poor Programmers.SE questions?

scotchgit.bitbucket.org/#

I'm just going to let that sit there

user55340

2:33 PM

@Phrancis interview questions can be difficult to ask right.

user55340

17

Occasionally a question that is asked at an interview comes up - "how would you design Amazon Web Services" or "how would you write Google" or some such question. These questions find their way to Programmers.StackExchange in hopes of getting an answer for the next time they are asked on an inte...

@MichaelT If this were your question, and given your experience, how would you ask it differently to be "asked right"?

Unrelated but funny

in The 2nd Monitor, 11 hours ago, by Mat's Mug

we're at the edge of a milestone. when your site hands out fizzbuzz tag badges, you know you've done something right.

@MichaelT I really like your answer on there, that makes a lot of sense

user55340

3:02 PM

Ask the question without needing to reference the interview. Or trying to justify or understand the interviewers response.

OK. Assuming that was rectified, would a question like this be "OK" for Programmers? It does seem to ask about a specific design concern, and presents a specific example use case. Would it be answerable enough? Or still kind of poorly received?

user55340

It really depends on how well the problem can be articulated.

Hm. Can you think of any question on your main site that is in a style like this, and well articulated, as an example?

user55340

In theory, any question that could be an interview question. It's much like homework questions - the good ones don't need to say they are homework or interview.

Gotcha

user55340

3:18 PM

For a previous job interview I wrote a log cleanup / warning of something broken script in perl. If I was to post it to code review, writing the question well shouldn't need to say this was for an interview. Same here.

user114359

@Phrancis I think that question would probably be a duplicate of something

user114359

But the answer to it is, quite simply, SRP is a little more nuanced than it sounds at first. If the responsibility is to "frobnicate widgets" then frobnicate it must, even if that is a large and complex operation. SRP is more about methods than classes according to Uncle Bob, although I would argue a class should not have too many concerns, either.

user114359

SRP would dictate that in that Execute() method it should use functional decomposition to break up the big operation into many smaller ones in separate methods or classes, and leverage those smaller operations.

user114359

Think about it this way: if your program is an IDE with a "compile" button, that is a huge process. No rational person would argue that compiling should require separate buttons for lexing, parsing, creating an AST, object code, linking, etc.

@Snowman So for instance you could have a class FrobnicateWidgets which would encompass the whole frobnicating operation, but you would break down the different steps that frobnicating encompasses into smaller methods/functions etc.

user114359

3:26 PM

@Phrancis correct.

Ah, learning stuff today :)

user114359

A method should only have a few lines of code, 5-10 depending on the language and what it does. This is a subjective criteria.

user114359

Making it quite obvious that a complex operation would require many simpler methods and classes to get it done. Those simpler methods and classes are much easier to test.

I can see that totally

user114359

amazon.com/Clean-Code-Handbook-Software-Craftsmanship/dp/…

user114359

3:29 PM

amazon.com/Code-Complete-Practical-Handbook-Construction/dp/…

user114359

Both books are a little bit old by now, that second one is almost 12 years old. But they are still quite relevant and timeless.

@JimmyHoffa I'm looking for something on there that I've had to compare....

after looking at price, I see why I have had 0 of these.

@JimmyHoffa I put it down somewhere and when I turned around it was gone and a note was left about "Not being in scope for this client"

Some say it's still failing builds to this day.

on a completely unrelated note, the client check cleared for the month, so we're putting development on hold til the new year.

3:45 PM

@Ampt They paid you, so you stop working for a month?

I really like that deal.

@ThomasOwens #justconsultingthings

Shouldn't you getting money enable you to continue working? I don't get it. :\

@ThomasOwens (I'm playing on the joke that consultants get paid to do nothing, we aren't actually stopping dev)

@Ampt Oh. I was worried I was in the wrong field. I'd take money to do nothing.

@ThomasOwens Nah, I'm just dreaming unfortunately.

user114359

4:06 PM

@ThomasOwens perhaps you would like being a CEO or CTO then?

@Snowman Do you have a job opening?

user114359

@ThomasOwens Nope. If I did, I would be taking it.

I see.

user55340

@Snowman the requirement for a cto to read the trade magazines is too onerous for me to consider it.

user55340

cio.com/magazine

user55340

4:17 PM

Try reading that for more than 10 minutes.

@enderland The last one I took, years ago, they told me (after years of my working as a developer) that I did not have the personality for a developer because I was too friendly, and refused to even do an interview.

@durron597 Z because it's the solutions to (X-2)(x-3)(x-4)(x-5)(X-7)(x-16)(x-26)=0 in ascending order.

@Ampt ours do :|

@psr lol. performance needs improvement, not cynical/bitter enough? :)

@Phrancis funny, they asked me if I was familiar with Fizzbuzz today in my interview and I was like... "yes..." and so they on the fly changed the question. LOL

2

people must really fail that if that's their goto interview question...

user114359

4:33 PM

@MichaelT apparently you are not good at pretending to work. Even if "working" is "reading boring magazines full of buzzwords." Sounds like nap time to me.

@Snowman you can become CEO easily, just start your own company!

@enderland That's hilarious

@Phrancis "you mean where you write fizz for mod 3, I mean multiples of 3, and buzz for mod 5?"

@whiteboard. My company just got me a subscription to safari books online. What should I read (I've started The Pragmatic Programmer)? I don't have anything in particular I need to read (though I may force myself to read some Angular books).

user41796

@psr The room's canonical list is maintained in MichaelT's profile

4:42 PM

@GlenH7 Thanks. I couldn't find it.

also, interview went much better this time imo if anyone cases

user41796

Props. Didn't realize today was the nth round

user41796

Although you'd think I would have picked up on that based upon your fizzbuzz comment

they on-the-fly changed it to a far more interesting thing, taking a dictionary/array and doing some "return true if any value in dictionary is in array multiple times" and some really good followup discussion there

user41796

That's cool. I'll do similar things based upon how well they handle the first few questions

4:47 PM

@enderland haha

@enderland Thank goodness. Buzz Fizz is brutal.

"Have you ever heard of Fizzbuzz?" "No." "{describes the problem}" makes note - no hire.

@enderland I hate interviews and this is why

@LightnessRacesinOrbit why?

that was a great discussion

meh

boring

I enjoy programming but I don't think I care about it. Perhaps that's the problem.

4:54 PM

@enderland How come? Is there anything better than sort array or build new dictionary so you can check dups?

user55340

@Snowman it's more a desire not to have a lobotomy to keep reading after minute two.

@psr I'd build a new dictionary with an element -> count mapping.

I can think of little more tedious than having a discussion about dictionaries and arrays and algorithms and optimising for space versus optimising for time

@psr those sorts of questions (that I asked) showed a lot more technical depth than fizzbuzz does

And compare the key set of one dictionary to the things in the other.

4:55 PM

despite the fact that I really enjoyed composing my binary data log format complete with ten-min-resolution index block and all the algorithms that power it

"depends on characteristics of array/dictionary, is one huge compared to other? etc"

@ThomasOwens Same thing. Depends on if you want count or true/false.

@psr I think it would be easier. Look for keys where the value is > 1.