July 09, 2008
Don wrote:
> Lars Ivar Igesund wrote:
>> Lars Ivar Igesund wrote:
>>
>>> Walter Bright wrote:
>>>
>>>> Mostly bug fixing.
>>>>
>>>> http://www.digitalmars.com/d/1.0/changelog.html
>>>> http://ftp.digitalmars.com/dmd.1.032.zip
>>> Again a stable release that breaks Tango :
>>>
>>> http://d.puremagic.com/issues/show_bug.cgi?id=2207
> 
> The probably related bug #2208 also breaks Tango.
> 
>>>
>>> It would be nice if the release process could stop this from happening.
>>>
> Yup. But I would have caught these before release if I hadn't had an email problem.
> 
>>
>> There is another one triggered by the Tango unittests:
>>
>> tango/core/Variant.d(361): Error: functions cannot return static array
>> char[3u]
>>
>> I'm not able to minimize it out of the box though. Maybe this is a result of
>> the compiler becoming stricter, but it is any case a breaking change.
> 
> Not necessarily. It might have uncovered a pre-existing bug. (Any accepts-invalid bug is very likely to break broken code).
> 
> 
Bug 870 is not completely fixed. It still gives missing line numbers.
July 09, 2008
Lars Ivar Igesund wrote:

> It would be nice if the release process could stop this from happening.

I wrote a short piece on this topic : http://larsivi.net/node/114

As it is, I will recommend against using this release for Tango users.

-- 
Lars Ivar Igesund
blog at http://larsivi.net
DSource, #d.tango & #D: larsivi
Dancing the Tango
July 09, 2008
On Wed, 09 Jul 2008 15:17:29 +0200, Lars Ivar Igesund wrote:

> Lars Ivar Igesund wrote:
> 
>> It would be nice if the release process could stop this from happening.
> 
> I wrote a short piece on this topic : http://larsivi.net/node/114
> 
> As it is, I will recommend against using this release for Tango users.

I would really appreciate it if W would send release candidates (for the intended stable branch) to some bigger projects for testing.

I just found another issue with dmd that breaks Tango (looks like a bug
in dmd rather than Tango).
(bugzilla is coming..)
July 09, 2008
On Wed, 09 Jul 2008 13:33:38 +0000, Moritz Warning wrote:

> On Wed, 09 Jul 2008 15:17:29 +0200, Lars Ivar Igesund wrote:
> 
>> Lars Ivar Igesund wrote:
>> 
>>> It would be nice if the release process could stop this from happening.
>> 
>> I wrote a short piece on this topic : http://larsivi.net/node/114
>> 
>> As it is, I will recommend against using this release for Tango users.
> 
> I would really appreciate it if W would send release candidates (for the intended stable branch) to some bigger projects for testing.
> 
> I just found another issue with dmd that breaks Tango (looks like a bug
> in dmd rather than Tango).
> (bugzilla is coming..)
Hm, I can't reproduce the problem after messing up my system a bit.
Maybe someone with Tango and dmd 1.032 could test if
isStaticArrayType!(char[])
compiles.
July 09, 2008
"Walter Bright" wrote
> Mostly bug fixing.
>
> http://www.digitalmars.com/d/1.0/changelog.html http://ftp.digitalmars.com/dmd.1.032.zip
>
> http://www.digitalmars.com/d/2.0/changelog.html http://ftp.digitalmars.com/dmd.2.016.zip


BTW, Don, thanks for commenting all the bugs as 'fixed' (at least for 1.032).  Walter can you please at least delegate someone to do this for every release?  I can't always remember which bugs I am CC'd on or have reported, so it is nice to see that it is fixed in my email.

-Steve


July 09, 2008
On Wed, 09 Jul 2008 18:42:41 +0400, Moritz Warning <moritzwarning@web.de> wrote:

> On Wed, 09 Jul 2008 13:33:38 +0000, Moritz Warning wrote:
>
>> On Wed, 09 Jul 2008 15:17:29 +0200, Lars Ivar Igesund wrote:
>>
>>> Lars Ivar Igesund wrote:
>>>
>>>> It would be nice if the release process could stop this from
>>>> happening.
>>>
>>> I wrote a short piece on this topic : http://larsivi.net/node/114
>>>
>>> As it is, I will recommend against using this release for Tango users.
>>
>> I would really appreciate it if W would send release candidates (for the
>> intended stable branch) to some bigger projects for testing.
>>
>> I just found another issue with dmd that breaks Tango (looks like a bug
>> in dmd rather than Tango).
>> (bugzilla is coming..)
> Hm, I can't reproduce the problem after messing up my system a bit.
> Maybe someone with Tango and dmd 1.032 could test if
> isStaticArrayType!(char[])
> compiles.

No, it doesn't.
July 09, 2008
Steven Schveighoffer wrote:
> "Walter Bright" wrote
>> Mostly bug fixing.
>>
>> http://www.digitalmars.com/d/1.0/changelog.html
>> http://ftp.digitalmars.com/dmd.1.032.zip
>>
>> http://www.digitalmars.com/d/2.0/changelog.html
>> http://ftp.digitalmars.com/dmd.2.016.zip
> 
> 
> BTW, Don, thanks for commenting all the bugs as 'fixed' (at least for 1.032).  Walter can you please at least delegate someone to do this for every release?  I can't always remember which bugs I am CC'd on or have reported, so it is nice to see that it is fixed in my email.
> 
> -Steve 
> 
> 
I've only done the 'missing line number' bugs, and the ones I'd reported.
BTW: All known 'missing line number' bugs are now fixed, except for #870. There were 16 such bugs in bugzilla.
July 09, 2008
On Wed, 09 Jul 2008 18:53:24 +0400, Koroskin Denis wrote: [..]
> No, it doesn't.

Ok, thanks. http://d.puremagic.com/issues/show_bug.cgi?id=2210
July 09, 2008
On Wed, 09 Jul 2008 19:45:16 +0400, Moritz Warning <moritzwarning@web.de> wrote:

> On Wed, 09 Jul 2008 18:53:24 +0400, Koroskin Denis wrote:
> [..]
>> No, it doesn't.
>
> Ok, thanks.
> http://d.puremagic.com/issues/show_bug.cgi?id=2210

Arrrgh!
July 09, 2008
== Quote from Don (nospam@nospam.com.au)'s article
> Lars Ivar Igesund wrote:
> >
> > There is another one triggered by the Tango unittests:
> >
> > tango/core/Variant.d(361): Error: functions cannot return static array
> > char[3u]
> >
> > I'm not able to minimize it out of the box though. Maybe this is a result of the compiler becoming stricter, but it is any case a breaking change.
> Not necessarily. It might have uncovered a pre-existing bug. (Any accepts-invalid bug is very likely to break broken code).

While it's invalid to return a static array from a function, it does seem like this could have undesired consequences:

    T fn(T)( T val ) { return val; }

Shouldn't the above be valid for all types?


Sean