On 5/9/2011 6:36 PM, Michel Fortin wrote:
> Le 2011-05-09 ? 20:17, Walter Bright a ?crit :
>
>> Done.
>>
>> On 5/9/2011 2:36 PM, Stephan Dilly wrote:
>>> Thanks, can you bundle a new beta ? This issue blocks dwt2 from being build or testing if there are other issues with this release.
> Thanks for the new beta. I see that the osx/lib32 directory has been renamed to osx/lib, which is good. But now there's something strange: I see a osx/lib32 *file* about the same size as libphobos2.a, and its change date is today while libphobos2.a is two days ago... Likely there is something wrong with your packaging script, I suspect it copies and rename the newer libphobos2.a to osx/lib32.
>

That's right, I forgot to fix the install script.

From: "Walter Bright" <walter at digitalmars.com>
> At long last! Thanks for all the hard work, everyone!
>
> http://ftp.digitalmars.com/dmd2beta.zip

Should std.iterator be missing? If so, maybe it should be mentioned in the changelog?

And speaking of the beta's changelog, 2.051, 2.052 are missing from it.

The failing case mentioned in my last pull request is
http://d.puremagic.com/issues/show_bug.cgi?id=5972
It used to fail with an "index assignment is not supported in CTFE";
now it fails with
"CTFE internal error" (but gives line number).

There are three options, all of which are reasonable IMHO:
(1) ship as-is
(2) change the error message (interpret.c, around line 2884)
(3) fix the bug so that the code works.

I am working on (3), but it's a little bit complicated. If the other
things (especially the druntime issues) take a while to fix, then (3)
will be an option, but I don't think it's worth delaying the release
for (it isn't a regression, though it will probably be mistaken for
one).

Also, in the changelog, I think it would be good to separate the optlink bugs from the dmd bugs, since there are so many of them this time. They are all at the top of the list of DMD bugs.


On 10 May 2011 08:21, Nick Sabalausky <bus_dmdbeta at semitwist.com> wrote:
> From: "Walter Bright" <walter at digitalmars.com>
>>
>> At long last! Thanks for all the hard work, everyone!
>>
>> http://ftp.digitalmars.com/dmd2beta.zip
>
> Should std.iterator be missing? If so, maybe it should be mentioned in the changelog?
>
> And speaking of the beta's changelog, 2.051, 2.052 are missing from it.
>
> _______________________________________________
> dmd-beta mailing list
> dmd-beta at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/dmd-beta
>

Another one:

[2.053 beta][CTFE] ICE: 'global.errors' on line 1416 in file 'constfold.c' http://d.puremagic.com/issues/show_bug.cgi?id=5975

int foo(wstring replace)
{
    wstring value = "";
    value ~= replace;
    return 1;
}
enum X = foo("X"w);

> dmd test.d
Assertion failure: 'global.errors' on line 1416 in file 'constfold.c'

abnormal program termination

I got that with this version of DMD:

https://github.com/D-Programming-Language/dmd/commit/4055820a360ae56c5469678f280e374151c18c11

Compiled in release mode.

PATCH:
interpret.c, copyLiteral(), line 2163:

        se2->committed = se->committed;
        se2->postfix = se->postfix;
        se2->type = se->type;
+        se2->sz = se->sz;
        return se2;


On 10 May 2011 11:56, Nick Sabalausky <bus_dmdbeta at semitwist.com> wrote:
> Another one:
>
> [2.053 beta][CTFE] ICE: 'global.errors' on line 1416 in file 'constfold.c' http://d.puremagic.com/issues/show_bug.cgi?id=5975
>
> int foo(wstring replace)
> {
> ? wstring value = "";
> ? value ~= replace;
> ? return 1;
> }
> enum X = foo("X"w);
>
>> dmd test.d
>
> Assertion failure: 'global.errors' on line 1416 in file 'constfold.c'
>
> abnormal program termination
>
> I got that with this version of DMD:
>
> https://github.com/D-Programming-Language/dmd/commit/4055820a360ae56c5469678f280e374151c18c11
>
> Compiled in release mode.
> _______________________________________________
> dmd-beta mailing list
> dmd-beta at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/dmd-beta
>

I just grabbed the beta and am not sure if these are in fact regressions or just fixes, but they idiom is used in LuaD and DFL.

Can't pass __traits value as const(char)*
Can't used post defined constant at compile-time

http://d.puremagic.com/issues/show_bug.cgi?id=5980 http://d.puremagic.com/issues/show_bug.cgi?id=5981

On Fri, May 6, 2011 at 9:47 PM, Walter Bright <walter at digitalmars.com> wrote:
> At long last! Thanks for all the hard work, everyone!
>
> http://ftp.digitalmars.com/dmd2beta.zip
> _______________________________________________
> dmd-beta mailing list
> dmd-beta at puremagic.com
> http://lists.puremagic.com/mailman/listinfo/dmd-beta
>