# \[Q\] Array.new - strange behavior

**URL:** <https://rubytalk.org/t/q-array-new-strange-behavior/11468>\
**Category:** ruby-talk\
**Created:** [11 May 2004 09:18 UTC](https://rubytalk.org/t/q-array-new-strange-behavior/11468 "2004-05-11T09:18:59Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![E\_R\_Bruecklmeier2](https://avatars.discourse-cdn.com/v4/letter/e/bbe5ce/32.png) [@E\_R\_Bruecklmeier2](https://rubytalk.org/u/E_R_Bruecklmeier2)\
**Post date:** [11 May 2004 09:18 UTC](https://rubytalk.org/t/q-array-new-strange-behavior/11468/1 "2004-05-11T09:18:59Z")

</div>

dear rubyists,

today i noticed a strange bahavior in Array.new while defining multi  
dimentional arrays:

p "P1"  
foo = Array.new(3,[0,0])  
foo.each do |foo1|  
p [foo1.id](http://foo1.id)  
end

p "P2"  
foo = Array.new  
3.times {foo.push [0,0]}  
foo.each do |foo1|  
p [foo1.id](http://foo1.id)  
end

produces:

"P1"  
19645868  
19645868  
19645868  
"P2"  
19645760  
19645748  
19645736  
Completed(0)

Is it intentional that the new method fills the array with just one  
instance of [0,0]? In my sense this is a violation of the PoLS.

Thanks for any hints.

Eric.

---

<div class="post-metadata">

**Author:** ![Claus\_Spitzer](https://avatars.discourse-cdn.com/v4/letter/c/49beb7/32.png) [@Claus\_Spitzer](https://rubytalk.org/u/Claus_Spitzer)\
**Post date:** [11 May 2004 09:36 UTC](https://rubytalk.org/t/q-array-new-strange-behavior/11468/2 "2004-05-11T09:36:23Z")

</div>

> dear rubyists,
> 
> today i noticed a strange bahavior in Array.new while defining multi  
> dimentional arrays:
> 
> p “P1”  
> foo = Array.new(3,[0,0])  
> foo.each do |foo1|  
> p foo1.id  
> end

I think I had a similar problem when I started with Ruby (coming from  
the C/C++ environment)… the solution was what you had following…

> p “P2”  
> foo = Array.new  
> 3.times {foo.push [0,0]}  
> foo.each do |foo1|  
> p foo1.id  
> end
> 
> produces:
> 
> “P1”  
> 19645868  
> 19645868  
> 19645868  
> “P2”  
> 19645760  
> 19645748  
> 19645736  
> Completed(0)

Interestingly, I get different results:

irb(main):001:0\> p “P1”  
“P1”  
=\> nil  
irb(main):002:0\> foo = Array.new()  
=\>   
irb(main):003:0\> 3.times{  
irb(main):004:1\* foo \<\< Array.new( 0, 0 )  
irb(main):005:1\> }  
=\> 3  
irb(main):006:0\> foo.each{ |i|  
irb(main):007:1\* p i.id  
irb(main):008:1\> }  
20999324  
20999312  
20999300  
=\> [, ,]  
irb(main):009:0\> foo = Array.new()  
=\>   
irb(main):010:0\> 3.times{  
irb(main):011:1\* foo.push [0,0]  
irb(main):012:1\> }  
=\> 3  
irb(main):013:0\> foo.each{ |i| p i.id }  
20964584  
20964572  
20964536  
=\> [[0, 0], [0, 0], [0, 0]]  
irb(main):014:0\>

> **···**
>
> On Tue, 11 May 2004 18:18:59 +0900, E.-R. Bruecklmeier [unet01@radio-eriwan.de](mailto:unet01@radio-eriwan.de) wrote:
> 
> > Is it intentional that the new method fills the array with just one  
> > instance of [0,0]? In my sense this is a violation of the PoLS.
> > 
> > Thanks for any hints.
> > 
> > Eric.

---

<div class="post-metadata">

**Author:** ![Laurent\_Julliard4](https://avatars.discourse-cdn.com/v4/letter/l/ebca7d/32.png) [@Laurent\_Julliard4](https://rubytalk.org/u/Laurent_Julliard4)\
**Post date:** [11 May 2004 09:38 UTC](https://rubytalk.org/t/q-array-new-strange-behavior/11468/3 "2004-05-11T09:38:53Z")

</div>

yes, it’s the normal behaviour.  
If you want you can initialize an Array with a block, say:  
Array.new(10) { rand }  
The block gets executed for every element in the array, giving  
different obvjects (or many copies of the same, in your case)

> **···**
>
> il Tue, 11 May 2004 11:15:34 +0200, “E.-R. Bruecklmeier” [unet01@radio-eriwan.de](mailto:unet01@radio-eriwan.de) ha scritto::
> 
> > Is it intentional that the new method fills the array with just one  
> > instance of [0,0]? In my sense this is a violation of the PoLS.

---

<div class="post-metadata">

**Author:** ![Robert](https://avatars.discourse-cdn.com/v4/letter/r/e95f7d/32.png) [@Robert](https://rubytalk.org/u/Robert)\
**Post date:** [11 May 2004 11:43 UTC](https://rubytalk.org/t/q-array-new-strange-behavior/11468/4 "2004-05-11T11:43:59Z")

</div>

“E.-R. Bruecklmeier” [unet01@radio-eriwan.de](mailto:unet01@radio-eriwan.de) schrieb im Newsbeitrag  
[news:c7q5it$jc5$1@news.BelWue.DE](mailto:news:c7q5it$jc5$1@news.BelWue.DE)…

> dear rubyists,
> 
> today i noticed a strange bahavior in Array.new while defining multi  
> dimentional arrays:
> 
> p “P1”  
> foo = Array.new(3,[0,0])  
> foo.each do |foo1|  
> p foo1.id  
> end
> 
> p “P2”  
> foo = Array.new  
> 3.times {foo.push [0,0]}  
> foo.each do |foo1|  
> p foo1.id  
> end
> 
> produces:
> 
> “P1”  
> 19645868  
> 19645868  
> 19645868  
> “P2”  
> 19645760  
> 19645748  
> 19645736  
> Completed(0)
> 
> Is it intentional that the new method fills the array with just one  
> instance of [0,0]? In my sense this is a violation of the PoLS.

Well, I’d say no, if you aware of how Ruby deals with instances and  
variables. All variables hold references - there is no distinction  
between pass by value and pass by reference as it is in C++. Then you’ll  
see that a single instance is handed over to the constructor in your first  
example. That instance is used for each array element, which is fine for  
immutable classes (integers for example). If you need different  
instances, you’ll need the block form as shown by you or even better the  
other one:

foo = Array.new( 3 ) {[0,0]}

irb(main):002:0\> foo = Array.new( 3 ) {[0,0]}  
=\> [[0, 0], [0, 0], [0, 0]]  
irb(main):003:0\> foo.each {|o| puts o.id}  
135011840  
135011828  
135011816  
=\> [[0, 0], [0, 0], [0, 0]]

Regards

```
robert

```