Yuushi

general recent conversations

Discussion on question by Orient: Agg…

Imported from a comment discussion on codereview.stackexchange...

Yuushi

Jan 25, 2016 18:52

Just a question - why the (std::declval< void * >()) in the noexcept? This also seems overly complex to me; since this is for emplace_back (which uses placement-new), the only thing that can throw is the actual constructor itself - thus I would expect a noexcept specification of type {std::declval<arguments>()...} (and similarly for the single argument case).

The 2nd Monitor

General discussion about codereview.stackexchange.com - Welcom...

Yuushi

Jul 11, 2015 13:24

I don't mind too much - there's lots that can be improved with it

Yuushi

Jul 11, 2015 13:24

don't have an implementation

Yuushi

Jul 11, 2015 13:24

@SimonAndréForsberg I can't test it as it is written without adding to it, as both /*Unweighted Graph */
def addedge(s: Vertex, d: Vertex): Unit
//Weighted Graph */
def addedge(s: Vertex, d: Vertex, w: Int): Unit

Yuushi

Jul 11, 2015 13:19

technically I still say it is broken code

Yuushi

Jul 11, 2015 13:18

@rolfl yes

Yuushi

Jul 11, 2015 13:17

@rolfl Your opinion on the on-hold question I linked?

Yuushi

Jul 11, 2015 13:13

or at least, is not guaranteed to

Yuushi

Jul 11, 2015 13:13

so any vertex with more than one adjacent vertex will not properly visit each neighbour

Yuushi

Jul 11, 2015 13:11

so it will only return one adjacent vertex

Yuushi

Jul 11, 2015 13:09

@SimonAndréForsberg the way getUnvisitedChildNode works means it is -still- somewhat broken

Yuushi

Jul 11, 2015 13:05

it just prints out each vertex it visits...perhaps that's not enough to qualify it as borken?

Yuushi

Jul 11, 2015 13:05

hrm, so was I too hasty in closing codereview.stackexchange.com/questions/96490/…

Yuushi

Jul 7, 2015 15:36

@SimonAndréForsberg done

Yuushi

Jul 6, 2015 18:51

I can't acronym

Yuushi

Jul 6, 2015 18:51

TAOCP

Yuushi

Jul 6, 2015 18:51

err

Yuushi

Jul 6, 2015 18:50

The Art of Computer Programming

Yuushi

Jul 6, 2015 18:49

On a separate topic, Vol 4B of TACOP is available on Knuth's site to read and find errors in (have fun with that): www-cs-faculty.stanford.edu/~uno/fasc6a.ps.gz

Yuushi

Jul 6, 2015 18:45

@Jamal but they're not quite the same

Yuushi

Jul 6, 2015 18:42

@Jamal replied, I say burn it

Yuushi

Jul 6, 2015 18:30

@Jamal What's broken in this question? codereview.stackexchange.com/questions/96003/…

Yuushi

Jun 30, 2015 07:05

how's everyone doing?

Yuushi

Jun 30, 2015 07:04

hey people

Yuushi

Jun 29, 2015 07:23

hey @QPaysTaxes

Yuushi

Jun 24, 2015 23:17

I've seen some horrendous crap in my time, I don't think I need to learn about that :)

Yuushi

Jun 24, 2015 23:16

...and then not posting an answer because it's a few hours later and I can't be bothered

Yuushi

Jun 24, 2015 23:16

@QPaysTaxes It's probably not so much learning, as going "that's a really bad way to do it", and then writing a better way

Yuushi

Jun 24, 2015 23:15

I've spent way too much time fiddling with code based on stuff I've read here in the past few days

Yuushi

Jun 24, 2015 21:06

@QPaysTaxes Python has an (unconventional) ternary operator as well

Yuushi

Jun 24, 2015 18:57

@Donald.McLean only works up to the limits of double precision

Yuushi

Jun 24, 2015 18:43

hey @QPaysTaxes

Yuushi

Jun 10, 2015 07:42

that's definitely borderline

Yuushi

Jun 8, 2015 15:22

Drunking

Yuushi

Jun 8, 2015 15:15

so hopefully I'll have some more time to answer stuff

Yuushi

Jun 8, 2015 15:15

but I'm off contract now

Yuushi

Jun 8, 2015 15:15

I was on contract last year...wrote about 20k lines of Python and 140 pages of documentation in 10 months

Yuushi

Jun 8, 2015 15:13

yeah, both flattening

Yuushi

Jun 8, 2015 15:07

@rolfl Am I a distracted regular?

Yuushi

Jun 7, 2015 19:11

@Vogel612 Not sure that is a good example of a review

Yuushi

Apr 4, 2015 16:06

function signature would be shorter

Yuushi

Apr 4, 2015 16:03

using a reference means an extra pointer dereference on each member method

Yuushi

Apr 4, 2015 16:03

@Steephen any decent compiler will use NRVO there

Yuushi

Apr 4, 2015 16:02

there's no reason to assume that at all

Yuushi

Apr 4, 2015 16:02

Yuushi

Mar 21, 2015 01:43

I can read some of it ><

Yuushi

Mar 21, 2015 01:41

indeed it does

Yuushi

Mar 21, 2015 01:40

I don't know if I would call C++ memory management "automatic"

Yuushi

Mar 3, 2015 14:20

quite detailed answers, both deserve more than the 1 upvote I've given them

Yuushi

Mar 3, 2015 14:18

Q: Type-safe cartesian co-ordinates

I've recently been fiddling around with a type safe implementation of cartesian co-ordinates (and a few operations on those co-ordinates). Often it's easy to get units mixed up: is something in metres, or in some other unit? The idea for this code is to provide a framework to catch all such erro...