The regression bug 4506 (the optimizer generates bad code if you use x
* = -1; inside a loop) is one of the worst DMD bugs ever. We should to
get a release out ASAP.
But we're currently blocked by Phobos:
Unit tests are still failing in std.container. And there's no error
message at all for failing unit tests.
These two issues need to become top priority.
Does anyone know of any other blockers?

A release would be great. I'll complete work on std.container this (Monday) evening.

Andrei

On 07/26/2010 03:09 AM, Don Clugston wrote:
> The regression bug 4506 (the optimizer generates bad code if you use x
> * = -1; inside a loop) is one of the worst DMD bugs ever. We should to
> get a release out ASAP.
> But we're currently blocked by Phobos:
> Unit tests are still failing in std.container. And there's no error
> message at all for failing unit tests.
> These two issues need to become top priority.
> Does anyone know of any other blockers?
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos

I can't think of any.

On Jul 26, 2010, at 1:09 AM, Don Clugston wrote:

> The regression bug 4506 (the optimizer generates bad code if you use x
> * = -1; inside a loop) is one of the worst DMD bugs ever. We should to
> get a release out ASAP.
> But we're currently blocked by Phobos:
> Unit tests are still failing in std.container. And there's no error
> message at all for failing unit tests.
> These two issues need to become top priority.
> Does anyone know of any other blockers?
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos

The following druntime checkin:

http://www.dsource.org/projects/druntime/changeset/344

breaks the build:

src\core\dll_helper.d(372): Error: undefined identifier thread_detach
src\core\thread.d(1856): Error: undefined identifier obj

Sean, can you please fix?

I just checked in a fix for that.

On 26 July 2010 22:05, Walter Bright <walter at digitalmars.com> wrote:
> The following druntime checkin:
>
> http://www.dsource.org/projects/druntime/changeset/344
>
> breaks the build:
>
> src\core\dll_helper.d(372): Error: undefined identifier thread_detach
> src\core\thread.d(1856): Error: undefined identifier obj
>
> Sean, can you please fix?
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos
>

Thanks!

Don Clugston wrote:
> I just checked in a fix for that.
>
> On 26 July 2010 22:05, Walter Bright <walter at digitalmars.com> wrote:
> 
>> The following druntime checkin:
>>
>> http://www.dsource.org/projects/druntime/changeset/344
>>
>> breaks the build:
>>
>> src\core\dll_helper.d(372): Error: undefined identifier thread_detach
>> src\core\thread.d(1856): Error: undefined identifier obj
>>
>> Sean, can you please fix?
>> _______________________________________________
>> phobos mailing list
>> phobos at puremagic.com
>> http://lists.puremagic.com/mailman/listinfo/phobos
>>
>> 
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos
>
>
> 

Is this release going to have at least somewhat-working 64 support?  I understand it won't be fully debugged, but will it be good enough for testing/playing around with/filing bug reports purposes?

On Mon, Jul 26, 2010 at 4:10 PM, Walter Bright <walter at digitalmars.com>wrote:

> Thanks!
>
>
> Don Clugston wrote:
>
>> I just checked in a fix for that.
>>
>> On 26 July 2010 22:05, Walter Bright <walter at digitalmars.com> wrote:
>>
>>
>>> The following druntime checkin:
>>>
>>> http://www.dsource.org/projects/druntime/changeset/344
>>>
>>> breaks the build:
>>>
>>> src\core\dll_helper.d(372): Error: undefined identifier thread_detach
>>> src\core\thread.d(1856): Error: undefined identifier obj
>>>
>>> Sean, can you please fix?
>>> _______________________________________________
>>> phobos mailing list
>>> phobos at puremagic.com
>>> http://lists.puremagic.com/mailman/listinfo/phobos
>>>
>>>
>>>
>> _______________________________________________
>> phobos mailing list
>> phobos at puremagic.com
>> http://lists.puremagic.com/mailman/listinfo/phobos
>>
>>
>>
>>
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.puremagic.com/pipermail/phobos/attachments/20100726/0a571737/attachment.html>

No, only very simple programs work.

David Simcha wrote:
> Is this release going to have at least somewhat-working 64 support?  I understand it won't be fully debugged, but will it be good enough for testing/playing around with/filing bug reports purposes?
>

Sigh, there are more problems with 344. Now the dll test suite fails to compile:

..\dmd -g -d -ofmydll2.dll -version=use_patch mydll2.d dll2.d mydll2.def
std.contracts has been scheduled for deprecation. Please use
std.exception instead.
dll2.d(49): Error: undefined identifier thread_setNeedLock
dll2.d(53): Error: function
core.thread_helper.thread_helper_aux.enumProcessThreads (uint procid, bo
ol function(uint id, void* context) dg, void* context) is not callable
using argument types (_error_
 function(uint id, void* context),void*)
dll2.d(53): Error: cannot implicitly convert expression (__funcliteral1)
of type _error_ function(ui
nt id, void* context) to bool function(uint id, void* context)
dll2.d(92): Error: function
core.thread_helper.thread_helper_aux.enumProcessThreads (uint procid, bo
ol function(uint id, void* context) dg, void* context) is not callable
using argument types (_error_
 function(uint id, void* context),void*)
dll2.d(92): Error: cannot implicitly convert expression (__funcliteral3)
of type _error_ function(ui
nt id, void* context) to bool function(uint id, void* context)
dll2.d(138): Error: no property 'findThread' for type 'core.thread.Thread'

It makes me very nervous that wholesale changes were made to the dll interface, and none of it was even checked to see if it compiled, and now the test suite for it is completely broken.

Don Clugston wrote:
> I just checked in a fix for that.
>
> On 26 July 2010 22:05, Walter Bright <walter at digitalmars.com> wrote:
> 
>> The following druntime checkin:
>>
>> http://www.dsource.org/projects/druntime/changeset/344
>>
>> breaks the build:
>>
>> src\core\dll_helper.d(372): Error: undefined identifier thread_detach
>> src\core\thread.d(1856): Error: undefined identifier obj
>>
>> Sean, can you please fix?
>> _______________________________________________
>> phobos mailing list
>> phobos at puremagic.com
>> http://lists.puremagic.com/mailman/listinfo/phobos
>>
>> 
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos
>
>
> 

Most of it was refactoring.  The problem is really that I don't have a working Windows environment to test with any more.  I'll see about rectifying that though.  It really is inexcusable that I keep checking in changes that I can't verify.

On Jul 26, 2010, at 2:32 PM, Walter Bright wrote:

> Sigh, there are more problems with 344. Now the dll test suite fails to compile:
> 
> ..\dmd -g -d -ofmydll2.dll -version=use_patch mydll2.d dll2.d mydll2.def
> std.contracts has been scheduled for deprecation. Please use std.exception instead.
> dll2.d(49): Error: undefined identifier thread_setNeedLock
> dll2.d(53): Error: function core.thread_helper.thread_helper_aux.enumProcessThreads (uint procid, bo
> ol function(uint id, void* context) dg, void* context) is not callable using argument types (_error_
> function(uint id, void* context),void*)
> dll2.d(53): Error: cannot implicitly convert expression (__funcliteral1) of type _error_ function(ui
> nt id, void* context) to bool function(uint id, void* context)
> dll2.d(92): Error: function core.thread_helper.thread_helper_aux.enumProcessThreads (uint procid, bo
> ol function(uint id, void* context) dg, void* context) is not callable using argument types (_error_
> function(uint id, void* context),void*)
> dll2.d(92): Error: cannot implicitly convert expression (__funcliteral3) of type _error_ function(ui
> nt id, void* context) to bool function(uint id, void* context)
> dll2.d(138): Error: no property 'findThread' for type 'core.thread.Thread'
> 
> It makes me very nervous that wholesale changes were made to the dll interface, and none of it was even checked to see if it compiled, and now the test suite for it is completely broken.
> 
> Don Clugston wrote:
>> I just checked in a fix for that.
>> 
>> On 26 July 2010 22:05, Walter Bright <walter at digitalmars.com> wrote:
>> 
>>> The following druntime checkin:
>>> 
>>> http://www.dsource.org/projects/druntime/changeset/344
>>> 
>>> breaks the build:
>>> 
>>> src\core\dll_helper.d(372): Error: undefined identifier thread_detach
>>> src\core\thread.d(1856): Error: undefined identifier obj
>>> 
>>> Sean, can you please fix?
>>> _______________________________________________
>>> phobos mailing list
>>> phobos at puremagic.com
>>> http://lists.puremagic.com/mailman/listinfo/phobos
>>> 
>>> 
>> _______________________________________________
>> phobos mailing list
>> phobos at puremagic.com
>> http://lists.puremagic.com/mailman/listinfo/phobos
>> 
>> 
>> 
> _______________________________________________
> phobos mailing list
> phobos at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/phobos