How about the std.base64 unittest failures?

Andrei Alexandrescu wrote:
> I'd forgotten to svn update druntime, now everything works. Thanks!
>
> Andrei
>

works4me

Andrei

Walter Bright wrote:
> How about the std.base64 unittest failures?
> 
> Andrei Alexandrescu wrote:
>> I'd forgotten to svn update druntime, now everything works. Thanks!
>>
>> Andrei
>>
> _______________________________________________
> dmd-beta mailing list
> dmd-beta at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/dmd-beta

On Tue, 26 Jan 2010, Steve Schveighoffer wrote:

> the included DMD change log lists this version as 2.039, but the executable reports it is 2.040.  There is already a released 2.039, so I'd guess that it's the changelog that is in error.
> 
> Also, inout still does not work correctly.  Should I report bugs on inout on bugzilla?  I'm not sure if that's appropriate, since the feature is new, I'm not sure if its behavior is considered to be a bug. I find it completely useless for object properties, which is its most prominent use-case.  Would it help if I created test cases for how I think it should work?

Report them.  That'll give a check list with tests.  If you wait until some are fixed it'll almost certainly just mean multiple rounds of fixing since chances are pretty high not everything will be fixed the first time around.

Walter Bright wrote:
> The latest patches affect both dmd and druntime, and both need to be there so it'll work, so here's a new baseline.
>
> You can download the software as a zip file from:
>
> http://download.digitalmars.com/dmd1beta.zip
> userid: customer
> passwd: download7
diffing this with the last dmd.1.*.zip shows some stuff about shared and a few others keywords in the html files. Isn't that 2.0 only? Maybe a missing macro?

Brad Roberts wrote:
> Report them. That'll give a check list with tests. If you wait until some are fixed it'll almost certainly just mean multiple rounds of fixing since chances are pretty high not everything will be fixed the first time around.


Brad's right.

Benjamin Shropshire wrote:
> diffing this with the last dmd.1.*.zip shows some stuff about shared and a few others keywords in the html files. Isn't that 2.0 only? Maybe a missing macro?
>

Can you be more specific about which files?

Walter Bright wrote:
>
>
> Benjamin Shropshire wrote:
>> diffing this with the last dmd.1.*.zip shows some stuff about shared and a few others keywords in the html files. Isn't that 2.0 only? Maybe a missing macro?
>>
>
> Can you be more specific about which files?

class.html references shared. I guess I was in error remembering other keywords being mentioned.

> _______________________________________________
> dmd-beta mailing list
> dmd-beta at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/dmd-beta
>

Benjamin Shropshire wrote:
>
> class.html references shared.

Thanks, I'll fix it.

On 01/26/2010 01:19 AM, Walter Bright wrote:
> The latest patches affect both dmd and druntime, and both need to be there so it'll work, so here's a new baseline.
>
> You can download the software as a zip file from:
>
> http://download.digitalmars.com/dmd1beta.zip
> userid: customer
> passwd: download7
>
>
> http://download.digitalmars.com/dmd2beta.zip
> userid: customer
> passwd: download7
> _______________________________________________
> dmd-beta mailing list
> dmd-beta at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/dmd-beta

There seems to be a regression in the way overridden functions are
resolved.
Compiling GtkD with the latest beta i get the following error:

src/gtk/Window.d(623): Error: function gtk.Window.Window.getScreen does not override any function

The fix seems simple, just remove the override but then when compiling with warnings on (-w) the compiler tels me the following:

warning - src/gtk/Window.d(623): Error: overrides base class function gtk.Widget.Widget.getScreen, but is not marked with 'override'

So dmd seems to disagree with itself, this is with dmd2beta/2.040. and the errors are coming from files that don't directly import gtk.window. Compiling gtk.window directly doesn't produce the first error.

I haven't been able to reduce it to a small test case yet.

Mike Wey wrote:
>
> There seems to be a regression in the way overridden functions are
> resolved.
> Compiling GtkD with the latest beta i get the following error:
>
> src/gtk/Window.d(623): Error: function gtk.Window.Window.getScreen does not override any function
>
> 

I think the checked in patch 356 fixes this.