import System
struct s:
        pass
class c():
        public l as (s)
c1=c()
c1.l=(s(),)

/*
# Output @ Sat Oct 17 20:58:18 BST 2009
System.OutOfMemoryException: Out of memory
  at (wrapper managed-to-native)
object:__icall_wrapper_mono_array_new_specific (intptr,int)
  at Intersect3Module.Main (System.String[] argv) [0x00000]
  at (wrapper managed-to-native)
System.Reflection.MonoMethod:InternalInvoke (object,object
[],System.Exception&)
  at System.Reflection.MonoMethod.Invoke (System.Object obj,
BindingFlags invokeAttr, System.Reflection.Binder binder, System.Object
[] parameters, System.Globalization.CultureInfo culture) [0x00000]
Command exited with non-zero status 127
# Used (interp) :  0:00.61 real,0.53 user,0.02 sys
*/

Hey!

Works on mono trunk.
Does it break too if the struct has at least one field?

Anyways I guess we should at least issue a warning about zero-size
structs, if not an error directly.

Cheers,

yes, see below;

import System
import System.Collections
import System.Collections.Generic

class vec3f(ValueType):
        X as double
        Y as double
        Z as double

class points():
        [Property(locations)]
        _locs as (vec3f)

p2=points(locations:(vec3f(),))
print p2

/*
# Output @ Mon Oct 19 14:12:12 BST 2009
System.OutOfMemoryException: Out of memory
  at (wrapper managed-to-native)
object:__icall_wrapper_mono_array_new_specific (intptr,int)
  at Intersect2Module.Main (System.String[] argv) [0x00000]
  at (wrapper managed-to-native)
System.Reflection.MonoMethod:InternalInvoke (object,object
[],System.Exception&)
  at System.Reflection.MonoMethod.Invoke (System.Object obj,
BindingFlags invokeAttr, System.Reflection.Binder binder, System.Object
[] parameters, System.Globalization.CultureInfo culture) [0x00000]
Command exited with non-zero status 127
# Used (interp) :  0:01.04 real,0.78 user,0.03 sys
*/

im now guessing there's something wrong with my Mono install

it works compiled but range is returning the wrong numbers!!

import System
import System.Collections
import System.Collections.Generic

class vec3f(ValueType):
        X as double
        Y as double
        Z as double

class points():
        [Property(locations)]
        _locs as (vec3f)

p2=points(locations:(vec3f(),vec3f(),vec3f()))
print p2

for i in range(0,len(p2.locations)):
        print i

/*
# Output @ Mon Oct 19 14:16:25 BST 2009
points
1
2
3
# Used (compiled) :  0:00.03 real,0.03 user,0.00 sys
*/

OK i've now tried a lot of old code, that all still works, and i've
found the problem is reproducible on another Linux machine with a
fresh install of Mono 2.4.2.3 ( the newest i can find in repositories,
i really don't want to have t build it, it sounds like it can be
tricky.) so its probably a real mono bug.

might have something to do with this bug, fixed last month, but not
really quite right; https://bugzilla.novell.com/show_bug.cgi?id=485446

can anyone let me if its working for them on Mono 2.4.x?, i'm a bit
stuck at the moment.

On Oct 19, 9:21 pm, psi <simon.pl...@googlemail.com> wrote:

Hmm... I might overlook something but the results look fine to me??

range(0, ...) should start at 0 but it outputs

that's really odd indeed.

On Oct 20, 7:35 pm, "Rodrigo B. de Oliveira" <rodrigobam...@gmail.com>
wrote:

Hmm yeah didnt catch that heh :)
Well FWIW I confirm testcase above also works on Mono trunk.

Cheers,

On Oct 21, 4:01 pm, Cedric Vivier <neo...@gmail.com> wrote:

By that, I mean it correctly outputs 0,1,2

thanks for your efforts, when i understand this i'll post.

simon