The 2nd Monitor

t3chb0t

18:00

linq doesn't do it either

static IEnumerable<TResult> SelectIterator<TSource, TResult>(IEnumerable<TSource> source, Func<TSource, int, TResult> selector) {
    int index = -1;
    foreach (TSource element in source) {
        checked { index++; }
        yield return selector(element, index);
    }
}

Hosch250

Yes it does.

skiwi

But doesn't LINQ optimize multiple actions into a single need to create a new list? You create a new list at every action

t3chb0t

here's the implementation from referencesource.microsoft.com/#System.Core/System/Linq/…

Hosch250

@skiwi No, I only create one for each method call, and add items to it.

@t3chb0t Yep, that yield keyword does it.

Dan Pantry

@Hosch250 LINQ would start to out-perform yours pretty quickly

Duga

18:01

Working code in need of review should be posted at Code Review — Johnny Mopp 51 secs ago

Dan Pantry

Simply due to the fact that your extensions create a new List every single time

skiwi

@Hosch250 Yes, but cannot LINQ make sure that as least new Lists are created as possible? Copying items over is a performance hit

Dan Pantry

whereas LINQ only does it once upon enumeration

Hosch250

@DanPantry Same for mine. Mine isn't lazy like Linq.

t3chb0t

@Hosch250 did you test it? I think we'll only know if @EBrown runs a few benchmarks :)

Dan Pantry

18:02

@Hosch250 That was my point. LINQ is lazily evaluated, so if you used .Where().Select on yours it would (probably) be slower than it would be if you used LINQ

t3chb0t

it's not lazy, yield return makes it lazy

Dan Pantry

(Which would remove any benefit not using the yield state machine would have)

Hosch250

@DanPantry Slower for the first call, but faster in the long run.

It is basically an abstraction so I don't have to write that for loop every time I want to select items from a list.

Dan Pantry

@Hosch250 Hmm, I might be conflating memory consumption with performance here, but yours would definitely consume more memory eventually, and the effect would be exacerbated on larger lists

t3chb0t

@Hosch250 how big are your lists?

Hosch250

18:04

@t3chb0t Probably no more than 10-20 items.

t3chb0t

ok, then you can never even measure it

Hosch250

But I do the same thing in RD's parser over several hundred items, and it sped it up by almost half.

t3chb0t

I had hoped you'll say 20mln or so

skiwi

@DanPantry But memory gets pretty close to equaling performance in something running in a VM...

Hosch250

Linq (and the yield keyword) is more expensive than you realize.

Duga

18:05

possible answer invalidation by Kittoes0124 on question by Kittoes0124: codereview.stackexchange.com/posts/151167/revisions

Hosch250

There is a reason you don't use it in compilers/parsers, and other performance-critical places.

t3chb0t

I've never had any issues with link and if I have, I run a profiler and optimize the critical part

skiwi

@Hosch250 That's because compilers always lag behind in features because they have been developed using older versions of the language

Dan Pantry

@Hosch250 Sure, it is, but you will consume more memory on larger lists using your approach

t3chb0t

because compilers need to run fast, no body will ever notice if a collection of 20 items has been enumerated in 20microseconds or 22microseconds

Dan Pantry

18:06

However, in a compiler, I'd agree that yeah, memory isn't really an issue

CommitStrip

CommitStrip: Coder Dilemma #10 – The Timesheet

posted on December 29, 2016 by CommitStrip

skiwi

If you're so worried about performance then ignore any type of LINQ functionality and just write down your full loop

Hosch250

@t3chb0t This is being used in the Roslyn compiler.

It is being triggered by an event, or something similar, whenever a certain node type is compiled.

@skiwi That's it. I abstracted my loop to make my code a little cleaner, so it looks like Linq, but acts like a normal loop.

skiwi

@Hosch250 But it doesn't, because you're allocating a new list multiple times (assuming you use Where and Select on the same target), you should only create a new list once if you're so worried about performance

Hosch250

Oh, these aren't meant to be used chained.

t3chb0t

18:09

@Hosch250 you should name it StatelessSelect or FastSelect or SelectToList

Dan Pantry

Whatever you are doing I would make it very clear that it is not a LINQ extension method

Hosch250

If I need to use them chained, I'll make a SelectWhere.

Dan Pantry

Cus it looks like one

t3chb0t

it won't be possible to use the other extension because this will have precedence, unless it's desired

RoboSanta

Naruto answer; accepted non-selfie answer with 0 score: Instantiating icons, labels, and text fields for a contact form python comparative-review form

skiwi

18:10

@Hosch250 Well, okay then, that's a major thing actually

Dan Pantry

Unrelated: bbc.com/news/technology-38458867

t3chb0t

@Hosch250 do you happen to have a link to this roslyn select? I look at github and could not find it

Hosch250

@DanPantry Can you imagine the result from a terrorist attack crashing this thing?

@t3chb0t That's because I haven't committed it yet.

Dan Pantry

@Hosch250 The media would be... droning on.... about it for weeks

t3chb0t

@Hosch250 so you're contributing this to roslyn?

Hosch250

18:14

It's currently in VSD's extensions, although I'll probably take it out.

No, VSD ties into Roslyn's API, but the recommendations are to make it as fast as possible.

The ones in Roslyn aren't allowed to use Linq.

They can't control third-party ones, but we are following their recommendations as close as possible.

t3chb0t

ok, now I get it

skiwi

@DanPantry That makes me think about one of those new Battlefield 1 maps

It's one where you need to shoot down a Blimp/Zeppelin

Dan Pantry

@Hosch250 Do you have any info on the implementation of LINQ? can't seem to find it. I'm interested in learning as to why LINQ is so bad. My understanding based on preliminary googling is that it makes a LOT of heap allocations.

(Thanks, Rust, for helping me understand the difference :D)

Hosch250

@DanPantry It uses the yield keyword and lazily returns items. Jon Skeet has quite a large section on how it works in C# In Depth.

Basically, it builds a state machine and only returns items when they are asked for.

Dan Pantry

I'm guessing that's a book?

skiwi

18:21

What's wrong with lazy?

Hosch250

Yup.

Dan Pantry

MORE BOOKS FOR THE BOOK GOD

skiwi

I can only see the state machine allocating more memory mostly

Hosch250

That creates a lot of overhead.

Dan Pantry

it has to all be stored on the heap as well, @skiwi

inline twiddling can all be done on a stack for one

The specific note about LINQ in Roslyn is that it cannot be used in hot paths

It's not "You can't use at all", but "don't use this in things that run often" to avoid lots of heap allocations, which can trigger a large GC collection = DOOOOOOM

(or at least heavily degraded performance)

EBrown

18:23

@Hosch250 I'ma benchmark the snot out of this when I take the dog to go potty.

Hosch250

davepeck.org/linq-collection-performance

That seems to be a good explanation.

Dan Pantry

That's my understanding

Hosch250

@EBrown That would be nice.

skiwi

I like the DOOOOOOOOM explanation more

t3chb0t

this one sounds resonable too stackoverflow.com/a/22894932/235671

Dan Pantry

18:24

LINQ composes well and reads nicer than for loops (some stuff you can do with LINQ is seriously hairy code in normal C#) but that's about the only benefit it has going for it

Well, that and the lazy evaluation

skiwi

Lazy evaluation makes up for most of the performance loss in most cases

Dan Pantry

Except when you don't actually need the lazy evaluation :P

Hosch250

It's about 2.5 times slower than a normal loop, according to that guy.

skiwi

@Hosch250 No way

Hosch250

Yep.

I told you that we cut the parse time by almost half by simply killing the Linq in RD.

Dan Pantry

18:26

Though, again, if you're doing a single loop, it's cheaper.

skiwi

Java 8 Streams aren't even that bad as far as I'm aware

Dan Pantry

If you're doing multiple transformations over a query, and you want it to be lazily evaluated, LINQ comes out on top

I would assume

Hosch250

It doesn't just do the looping. It would have come well before C# 3 if that was the only thing involved.

Dan Pantry

but it would still be faster to use a regular for loop with multiple transformations if you don't need lazily evaluation

skiwi

And I won't believe C# LINQ is worse than Java 8 Streams as they are essentially the same concept

Dan Pantry

18:27

Skiwi, when drinking the LINQ kool-aid then finding out about the performance hit

Hosch250

@skiwi It also does stuff with the types, and stuff.

Dan Pantry

Write everything in Rust. Problem solved. /s

2

RubberDuck

@DanPantry I support this opinion.

Hosch250

Wait, the reflection-like stuff might have been the dynamic stuff that came in C# 4. I do remember that it does stuff to figure out the types, and stuff too.

skiwi

@DanPantry We must go deeper. Assembly.

Hosch250

18:29

Meh, just read C# In Depth. He'll explain it.

skiwi

Where I'm going we don't need no books :o

I'm probably doing a ton of things wrong here, but it works and is maintainable: pastebin.com/ptS9Y7wg

RubberDuck

@Hosch250 I also support this opinion.

Hosch250

@skiwi You might be surprised about that last point.

EBrown

@Hosch250 I'm about to start benchmarking.

Any predictions on yours vs. builtin LINQ?

RubberDuck

@skiwi not bad, but what's up with all the casting?

skiwi

18:33

@Hosch250 I can maintain it :P

@RubberDuck WPF doesn't do generics, mostly, except when it does

RubberDuck

Ahhhh yeah.... WPF needs updated. It predates generics and that's unfortunate.

UWP is slightly better about it, but MS reused a lot of WPF, so.....

skiwi

I'm not sure if UWP was a serious attempt? Because it cannot interact with the core of the OS

Dan Lyons

@RubberDuck eh? generics was .NET 2.0, while WPF was .NET 3.0

EBrown

@Hosch250 @t3chb0t Benchmarking.

3

I have three comparisons: Where, Select, and Where.Select.

And there are three methods in each: a baseline (a regular loop programmed to do that operation), the method from @Hosch250, and the LINQ method.

So, in about 2 minutes we'll see.

Hosch250

Mine will be slower than an inline loop because of the member invocation.

RubberDuck

18:44

@skiwi you've got to remember that UWP was designed to deploy to any of Microsoft's operating systems.

Hosch250

I moved mine to be inline, anyway.

RubberDuck

When it was first designed, they had a different OS for mobile.

EBrown

The mission is to take 1,000,000 integers (0 to 999,999) and then for just filtering return the integers that are even, and for the select return double each integer.

RubberDuck

@DanLyons my bad. In that case, screw microsoft. Jerks.

EBrown

               Method |       Mean |    StdDev |    Gen 0 |    Gen 1 |    Gen 2 | Allocated |
--------------------- |----------- |---------- |--------- |--------- |--------- |---------- |
       BaselineFilter |  4.9656 ms | 0.0262 ms | 798.9583 | 783.3333 | 783.3333 |    4.2 MB |
           HoschWhere |  9.9907 ms | 0.0452 ms | 583.3333 | 583.3333 | 583.3333 |    4.2 MB |
      LinqWhereToList |  9.3006 ms | 0.0402 ms | 566.6667 | 566.6667 | 566.6667 |    4.2 MB |
       BaselineSelect |  7.5805 ms | 0.0654 ms | 218.7500 | 218.7500 | 218.7500 |   8.39 MB |

@Hosch250 Only your .Select alternative is faster than the LINQ one.

Mast

18:45

@RubberDuck They fixed that though, not sure whether you see that in the development but the deployment goes more universal now.

skiwi

@RubberDuck Well yeah, but other frameworks do that too and then allow OS-specific code

RubberDuck

@skiwi MS allows that too. They just expect you to design your code base properly for it.

Which is probably a mistake.....

skiwi

@EBrown Careful you might hit branch prediction there, it might matter, but it might also not

@RubberDuck You mean that I can interact with the Win32 API on Windows (10) in UWP code?

RubberDuck

@skiwi I'm almost certain you can, but you have to inject the behavior and code to an interface.

EBrown

@skiwi If that's the case then it should have a different result if I use a random list of 1,000,000 integers, which I have just done.

skiwi

18:49

@EBrown Cool

Branch prediction definitely does its job on a sorted list, but I suspect it's able to look through even/uneven too

EBrown

SourceIntegerList = Enumerable.Range(0, 1000000).Select(x => random.Next(int.MinValue / 2 + 1, int.MaxValue / 2 - 1)).ToList();

skiwi

There's no .Shuffle?

EBrown

@skiwi Don't need it, I just use a Random to initiate it.

skiwi

Oh.. right

lol, guess what I found when I looked for a shuffle method on Java 8 streams to compare

5

Q: How to shuffle a stream using the Stream API?

I decided to take the functional approach in generating a string or random characters, so far I came up with this, it should perform better than boxing and then using a StringJoiner as collector: Random random = new Random(); String randomString = IntStream.concat( random.ints(8, 'a', 'z...

EBrown

                Method |       Mean |    StdDev |    Gen 0 |    Gen 1 |    Gen 2 | Allocated |
---------------------- |----------- |---------- |--------- |--------- |--------- |---------- |
        BaselineFilter | 14.7134 ms | 0.1176 ms | 150.0000 | 150.0000 | 150.0000 |    4.2 MB |
            HoschWhere | 17.0308 ms | 0.0937 ms | 183.3333 | 183.3333 | 183.3333 |    4.2 MB |
       LinqWhereToList | 18.4492 ms | 0.0966 ms | 200.0000 | 200.0000 | 200.0000 |    4.2 MB |
        BaselineSelect |  7.5587 ms | 0.0700 ms | 152.0833 | 152.0833 | 152.0833 |   8.39 MB |

Branch prediction eliminated..

skiwi

18:51

I kind of want to code that in Clojure again to see how easy it could be

EBrown

Doing the Where or Select in a raw loop body is the fastest every time. The Where provided by @Hosch250 is barely faster than the LINQ Where, and the Select provided is somewhat faster than the LINQ Select. When the provided are slower than the LINQ alternatives.

skiwi

And lol on my question I could just box the ints, map them to a string and then string join, /sillyme

EBrown

And if we make the list an int?, I suppose we might get different results.

Let's see how it does with non-value types.

Captain Obvious

0

Q: Advent of Code 2016 Day 2

I'm working through Advent of Code 2016 in Rust. The prizes go to the swift, but I'm aiming for well-written code. This is Day 2. To summarize the problem statement, given a newline-separated list of directions (e.g. "ULL\nRRDDD\nLURDL\nUUUUD\n"), move about on a 9-key keypad starting at 5, free...

programming-challenge rust

t3chb0t

I thought there would be a larger difference in memory allocation and those times, the differ only by a few milliseconds, I can imagine this matters mainly for compilers

EBrown

18:57

One of these methods just got really slow for reference types.

                Method |       Mean |    StdDev |    Gen 0 |    Gen 1 |    Gen 2 | Allocated |
---------------------- |----------- |---------- |--------- |--------- |--------- |---------- |
        BaselineFilter | 18.9028 ms | 0.1191 ms |        - |        - |        - |   8.39 MB |
            HoschWhere | 23.9599 ms | 0.0988 ms |        - |        - |        - |   8.39 MB |
       LinqWhereToList | 28.3955 ms | 0.1128 ms |        - |        - |        - |   8.39 MB |
        BaselineSelect | 12.6064 ms | 0.0951 ms | 175.0000 | 175.0000 | 175.0000 |  16.78 MB |

Dan Lyons

x64?

EBrown

@DanLyons Any CPU

Mast

Opinions?

@Mast Not in case of serious bugs .(An exception to that would be if a user wrote a comment saying that your code is completely broken, and needs to be fixed before it can be reviewed.) This is exactly what happened. — user1095108 2 hours ago

t3chb0t

wow, the baseline beats everything else by a lot

skiwi

Assembly would beat the baseline

Dan Lyons

19:01

relatively speaking, yes

EBrown

@skiwi That's not a given.

skiwi

But I'm afraid you lose a bit of maintainability

Dan Lyons

though if it's not hot path code, it's not that big a deal

Hosch250

Whew, I have Field Can Be Readonly done!

Time for a live checkout.

skiwi

@EBrown Hand-crafted assembly will always beat generated code... as long as the person writing it is skilled enough ;)

EBrown

19:02

That's still not a guarantee.

For very simple things like this, it's possible that the compiler is capable of generating the most optimal machine code.

skiwi

No, you can always beat the compiler

(Or be as least as fast)

Hint: You can even write it in C, take the assembly output from the compiler and then start working based on that

Mast

To what end?

Pre-mature optimization is the root of all evil.

EBrown

@skiwi Prove it.

I want to see evidence that every single compiler output can be beat by hand-written assembly.

Because until that happens, there will always be edge cases.

skiwi

@EBrown I'm not talking about beating, I'm talking to be as least as good, it's okay to be the same

@Mast No end, just saying that assembly would likely beat the C# baseline

Dan Lyons

but if it's the same, then you've wasted (no small amount of) your time

skiwi

19:07

That's true, but it won't be worse

Mast

Who cares about development time anyway.

Dan Lyons

or maintainability

EBrown

@Mast Only companies spending lost of money on building things.

                Method |       Mean |    StdDev |    Gen 0 |    Gen 1 |    Gen 2 | Allocated |
---------------------- |----------- |---------- |--------- |--------- |--------- |---------- |
        BaselineFilter | 14.7097 ms | 0.0753 ms | 250.0000 | 250.0000 | 250.0000 |    4.2 MB |
            HoschWhere | 17.3036 ms | 0.0738 ms | 216.6667 | 216.6667 | 216.6667 |    4.2 MB |
       LinqWhereToList | 18.4912 ms | 0.1034 ms | 216.6667 | 216.6667 | 216.6667 |    4.2 MB |
        BaselineSelect |  7.6007 ms | 0.0375 ms | 177.0833 | 177.0833 | 177.0833 |   8.39 MB |

Value-type benchmarks ^^

skiwi

There shouldn't be a difference between primitives and value-types, should there?

EBrown

@skiwi Correct.

Doing true reference-type benchmark now.

Custom class, with a string property. Randomly generated string objects, filtering out string.Empty strings, and then for Select replacing all characters with different type of character.

One of these methods just got really slow...holy carp.

skiwi

19:16

Try overclocking your CPU :P

EBrown

WOW

This still isn't done.

That was only group 2 that it ran.

It still has to do group 3.

Mast

slow

Dan Lyons

attack of the garbage collector?

EBrown

@DanLyons It's probable.

Hosch250

Which group is slow?

The Linq one, or mine?

EBrown

19:22

The Select and WhereSelect groups were slow.

Definitely garbage collector.

Hosch250

Mine, or Linq?

EBrown

Both

So was baseline.

The great thing about benchmarking is that it wastes a great deal of time.

6

t3chb0t

to save it later

EBrown

Don't forget the time you spend finding the chart to look up what you save. And the time spent reading this reminder about the time spent. And the time trying to figure out if either of those actually make sense. Remember, every second counts toward your life total, including these right now.

Duga

possible answer invalidation by Jamal on question by Samuel Cambridge: codereview.stackexchange.com/posts/108985/revisions

EBrown

19:31

BE DONE ALREADY

Omg it still has more to do.

Mast

lol

EBrown

Just. Finish. The. Friggin. Benchmark.

So much slow.

Working with strings is terrible.

#TIL

                Method |       Mean |    StdDev |     Gen 0 |     Gen 1 |   Gen 2 | Allocated |
---------------------- |----------- |---------- |---------- |---------- |-------- |---------- |
        BaselineFilter | 18.9009 ms | 0.2465 ms |         - |         - |       - |   8.39 MB |
            HoschWhere | 24.3390 ms | 0.3323 ms |         - |         - |       - |   8.39 MB |
       LinqWhereToList | 28.5717 ms | 0.3118 ms |         - |         - |       - |   8.39 MB |
        BaselineSelect | 68.9511 ms | 0.2759 ms | 7354.1667 | 7041.6667 |       - |  48.39 MB |

Reference-type benchmarks ^^

Mast

EBrown

@Hosch250 Your WhereSelect lost again.

Hosch250

That's not surprising. I told you they shouldn't be used chained.

EBrown

19:35

It wouldn't be hard to make them chain able.

But then you get into the same thing LINQ does (sorta).

Hosch250

That's exactly it--we aren't supposed to be using the yield keyword.

EBrown

Actually, you could get away from that entirely.

And still use the List idea.

Captain Obvious

0

Q: Basic Noughts & Crosses - Java

So I got bored and made a simple noughts and crosses game with colours instead of symbols. The colours are red and blue and I made the GUI in NetBeans with my own code in combination with the NetBeans GUI Editor. My Code: package com.coal; import java.util.Random; public class Noughts exte...

java

Duga

I'm voting to close this question as off-topic because it belongs on CodeReview. — Scott Hunter 27 secs ago

Hosch250

@JeroenVannevel Now I got the tests passing by updating all the NuGet packages, but I can't run it live anymore:

> Warning AD1000 An instance of analyzer VSDiagnostics.Diagnostics.General.FieldCanBeReadonly.FieldCanBeReadonlyAnalyzer cannot be created from C:\USERS\HOSCH\APPDATA\LOCAL\MICROSOFT\VISUALSTUDIO\14.0ROSLYN\EXTENSIONS\JEROEN‌\VSDIAGNOSTICS.VSIX\1.0\VSDiagnostics.dll: Could not load file or assembly 'System.Collections.Immutable, Version=1.2.1.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a' or one of its dependencies. The system cannot find the file specified.

Mast

19:46

Tell your file to stop playing hide-and-seek.

Hosch250

Once we get everything working, though, I have a Field Can Be Readonly one done.

@Mast NuGet packages.

The tests can find it :(

Stack Exchange

0

Q: I think I've answered a borderline off-topic question, is that answer still valid?

I've just written up an answer for this question Scattergories program After scanning through the code and writing up my points to improve, I noticed that OP asked how to make that working as intended at all: I would like to add a way to either check if a word is correct, or to suggest word...

discussion specific-question

Hosch250

@StackExchange Looks like a potentially problematic user. Just read that comment and their profile description.

Captain Obvious

0

Q: Counting bigrams (pair of two characters) in a file

I want to count the number of occurrences of all bigrams (pair of characters) in a file using C language. Bigram counts count the frequency of pairs of characters. Actually this is about Letter Frequencies. i implemented monogram, that was simple but i didn't find a solution related to bigrams. T...

c

Duga

Use code reviews. — Lew Bloch 14 secs ago

Mast

20:01

@Hosch250 He's been here before. Whatever got him a suspension of a month at SO, let's hope he behaves here.

skiwi

One does not simply get a suspension.

Captain Obvious

0

Q: can i modern my code?

ho can I improve my code ? Im new in swift and I want to know that I'm doing something wrong or harder way :). I make a simple war game app :D, names of variables aren't correct I know but I was doing it quickly and I didn't worry about names of var In that small project so I call them in English...

ios swift3

1

Q: Improving this TileMap

I created a TileMap-class to load tile sets, create and display tile maps. It is far from perfect so I want your opinions on how to improve it! How it works: 1. Create TileMap object with "new TileMap( int tilesize )". 2. Create a tile set with "createSet( String path )" and specify the tile typ...

java

0

Q: How might the Mapbox GL JS examples be implemented in React?

How would the Mapbox GL JS examples, such as (https://www.mapbox.com/mapbox-gl-js/example/filter-features-within-map-view/), be implemented using React?

dom react.js opengl hash-map

Hosch250

@CaptainObvious Burn with fire.

@CaptainObvious Now that would be a candidate for the Jamalizer award if it gets edited.

Vogel612

@CaptainObvious I don't even know what "wargame" they mean ..

Captain Obvious

20:20

0

Q: Composed Architecture

When refactoring and old class in PHP, I noticed that it contained a lot of business logic and thus was not allowing code reuse. So I thought about composition in order to achieve horizontal code reuse (since inheritance permits only vertical reuse) and the possibility to do Dependency Injection ...

php design-patterns

Stack Exchange

0

Q: What to do if there's not much space for improvement on a question?

I've been coming over this question unordered_multimap approach to adjacency list and graph search which specifically asks about I'm looking for suggestions on how to improve the usage of modern C++ features with this simple graph search algorithm, particularly with regard to the use of ST...

discussion

Phrancis

@EBrown All managers should bookmark that

Dan Pantry

20:37

@Phrancis They'd probably just reply with something like this.

(I have no reason to link that image other than I wanted to link it because I find it hilarious)

Hosch250

@StackExchange Already spamming our meta?

@DanPantry Haha, a Sith/Jedi cross.

Phrancis

LOL

Captain Obvious

-2

Q: i am in need of a basic Tic Tac Toe game

i really need a basic tic tac toe game. it should be very basic. just player vs player needed.i'm using Qt 5.6 and 5.7. please help me in this. i really need the game. i would be very thankful.

qt

0

Q: What is wrong with my missile trajectory model?

I'm trying to make a (simple!) trajectory model of a missile that is launched from the surface in a certain angle. The missile applies force to go forward until it's fuel has run out. I thought I had it figured out right. But when I simulated some trajectories I found out that the angle with opti...

javascript simulation physics

skiwi

@CaptainObvious No.

Phrancis

20:56

What kind of behavior would one expect from push() ing a node to the front of an empty linked list? By that I mean it is only the head pointer

Hosch250

Well, I would expect that it links to itself if it is a circularly linked list.

If it isn't a circularly linked list, then it links to null, I guess.

I posted my latest analyzer/code fix on CR.

Take that, R#.

Captain Obvious

0

Q: Field Can Be Made Readonly

Here is my Field Can Be Made Readonly analyzer and quick fix. I would appreciate tips on how to improve my code, and on missed test cases. First, the rules it works with: It will only fire on private fields to make things simpler (fields in other scopes should be properties, anyway). If the f...

c# meta-programming roslyn

Duga

21:21

I'm voting to close this question as off-topic because it belongs in CodeReview. — Scott Hunter 12 secs ago

Hosch250

21:38

OK, how do you contact someone about a copyright when the postmaster returns email sent to the posted address?

And they don't have any other contact methods listed?

Vogel612

21:51

then you double check the email adress provided and then you make intentional typos

Duga

I'm voting to close this question as off-topic because it belongs in CodeReview.StackExchange.com, rather than here. — Prune 53 secs ago

Captain Obvious

0

Q: How to optimize the following algorithm?

Hello I have the following three lists, they have the same lenght and every element from a list as its corresponding value in the same position of the other lists: for instance the label of the comment1 is 23 since is the first position of the list labels and it's corresponding user would be 899...

python performance

Duga

I'm voting to close this question as off-topic because "better" is opinion based. You might be able to adapt this into a question that would be on-topic for code review. — Quentin 56 secs ago

This really seems like a question better suited for softwareengineering.stackexchange.com or codereview.stackexchange.com. You might find the people there more willing to help tackle something as complex as this. Stack Overflow is really only meant for getting help with specific code problems. — Random Davis 51 secs ago

codereview.stackexchange.com/questions/74443/… — Krease 22 secs ago

skiwi

22:14

Q: Wait, how did "cut" become "cum"? A: Cyrillic italic/cursive letter forms are sometimes very different between cases. — Rhymoid Dec 26 at 21:05

denis

Is there a way to get the member's name of an UnaryExpression ?

Duga

There is another site on StackExchange, called CodeReview, to answer questions such as you have asked. — Scott Hunter just now

Code Review is only appropriate if there's code (and the other conditions given in the Help Center there are met). This is too broad for SO as-asked, but it's not a fit for CR either. — Charles Duffy 44 secs ago

Hosch250

22:35

@t3chb0t You misunderstood what I'm doing there.

I'm not finding symbols to report, I'm finding ones that can't be reported...

I like your idea of a hashmap though.

@Vogel612 That last one should be interesting.

Click the names in the footer to get the email address: research.fit.edu/shells/index.php

Vogel612

check with the fit secretary for an email

denis

I want to post my answer as a question to be reviewed I remember @Hosch250 mentioned there was a tag for that ?

Vogel612

rags-to-riches

Hosch250

@denis Rags to riches.

@Vogel612 Thanks. I was thinking that was the next step to try.

Captain Obvious

0

Q: Simple Logger Written In Erlang

I wrote a simple logger for a server I'm writing. Wondering if anyone has any suggestions to improve it. doom_logger.erl -module(doom_logger). -author('Andrew McRobb'). -include("include/log_types.hrl"). -export([start/1, start/0, terminate/1, log/4, log/3, update_logger_state/3, get_local_ti...

logging erlang

denis

22:42

How should my question look ? Do I've explained pretty much everything in the answer should I just put a link to it or rather copy paste it into my question's body ?

Vogel612

the code needs to be present in the question. Why not read some other questions in the same tag to see how those are done?

Duga

This is probably a better fit for codereview.stackexchange. — Michael0x2a 58 secs ago

denis

Great ida @Vogel612 will do, thanks !

Duga

If the code works without any errors (hard to see, with such a huge dump), you can ask on Code Review. Make sure to read their guidelines first. — Rad Lexus 21 secs ago

Hosch250

I wouldn't be surprised if that email was just inactive for so long that it was automatically reclaimed.

Duga

22:45

I'm voting to close this question as off-topic because the question is better suited for codereview.stackexchange — Mike Müller 38 secs ago

Vogel612

@Hosch250 doubt it.

Hosch250

Well, I know online email providers do that.

Duga

Thanks! I didn't know Code Review. — pyring 56 secs ago

Hosch250

That's a custom email, though, so...

AFK, BBIAB.

denis

22:58

Good night everyone :)

Captain Obvious

0

Q: Accessing properties by name with Compile-time typesafety

I've recently answered a question here. This is my code : public class TypeAccessor<T> { private readonly Func<T, T> m_applyDefaultValues; private readonly Func<T> m_constructType; public ReadOnlyCollection<string> CloneableProperties { get; } public ReadOnlyDictionary<string...

c# generics lambda rags-to-riches type-safety

Captain Obvious

23:41

0

Q: Simplify a program that simulates a game

The following code simulates a game of signals in which four participants (playing in pairs) show (muestra) and see (observan) signals. Signals are randomly assigned in the first round. Signals in the second round are assigned depending on a probability equation that takes into account memory dic...

python python-2.7

Transcript for