On Thursday, 2 November 2017 at 08:43:33 UTC, Basile B. wrote:
> On Monday, 30 October 2017 at 11:12:29 UTC, Basile B. wrote:
>> On Saturday, 28 October 2017 at 23:21:56 UTC, Martin Nowak wrote:
>>> First release candidate for 2.077.0.
>>>
>>> The OMF Windows API import libraries were updated, a bug in std.bigint was fixed, and the version identifier for `-betterC` is now `D_BetterC`.
>>>
>>> - -Martin
>>
>> Thanks, i have no problem here with the RC, tested it well, though looking at the log there would be no reason why i would have.
>
> Actually there's a REG  :/
>
> a 192 bytes leak is created by the GC, just with an empty main.
> Also i don't know how this is possible but travisCI runs with DMD 2.077 by default (even if not released).

You can ignore this REG i guess. After reflection it appears that it's only a minor problem that only happens w/ -unittest and that's even not a true leak, despite of being reported by valgrind:

https://issues.dlang.org/show_bug.cgi?id=17960

On 11/02/2017 09:43 AM, Basile B. wrote:
> On Monday, 30 October 2017 at 11:12:29 UTC, Basile B. wrote:
>> On Saturday, 28 October 2017 at 23:21:56 UTC, Martin Nowak wrote:
>>> First release candidate for 2.077.0.
>>>
>>> The OMF Windows API import libraries were updated, a bug in std.bigint was fixed, and the version identifier for `-betterC` is now `D_BetterC`.
>>>
>>> - -Martin
>>
>> Thanks, i have no problem here with the RC, tested it well, though looking at the log there would be no reason why i would have.
> 
> Actually there's a REG  :/
> 
> a 192 bytes leak is created by the GC, just with an empty main.
> Also i don't know how this is possible but travisCI runs with DMD 2.077
> by default (even if not released).

For the install.sh script (and Travis-CI) it's released when we update downloads.dlang.org/releases/LATEST, this happens slightly before the release announcement.