Thread overview
[D-runtime] [D-Programming-Language/druntime] 871de1: update MANIFEST
Jan 25, 2011
Brad Roberts
Jan 25, 2011
Walter Bright
Jan 25, 2011
Walter Bright
Jan 25, 2011
Brad Roberts
Jan 25, 2011
Walter Bright
Jan 25, 2011
Walter Bright
Jan 25, 2011
Brad Roberts
January 24, 2011
Branch: refs/heads/master
Home:   https://github.com/D-Programming-Language/druntime

Commit: 871de1a7665f766fbcdf2071187be9a1d4c23eb7
    https://github.com/D-Programming-Language/druntime/commit/871de1a7665f766fbcdf2071187be9a1d4c23eb7
Author: Walter Bright <walter at walterbright.com>
Date:   2011-01-24 (Mon, 24 Jan 2011)

Changed paths:
  M posix.mak

Log Message:
-----------
update MANIFEST


January 24, 2011
On 1/24/2011 9:26 PM, noreply at github.com wrote:
> Branch: refs/heads/master
> Home:   https://github.com/D-Programming-Language/druntime
> 
> Commit: 871de1a7665f766fbcdf2071187be9a1d4c23eb7
>     https://github.com/D-Programming-Language/druntime/commit/871de1a7665f766fbcdf2071187be9a1d4c23eb7
> Author: Walter Bright <walter at walterbright.com>
> Date:   2011-01-24 (Mon, 24 Jan 2011)
> 
> Changed paths:
>   M posix.mak
> 
> Log Message:
> -----------
> update MANIFEST

IMHO, this is the wrong direction for makefiles in general.  Duplication of long lists is just asking for them to grow
out of sync.  Why did you unroll the $(addprefix src,$(addsuffix .d,$(SRC_D_MODULES))) line?
January 24, 2011

Brad Roberts wrote:
> On 1/24/2011 9:26 PM, noreply at github.com wrote:
> 
>> Branch: refs/heads/master
>> Home:   https://github.com/D-Programming-Language/druntime
>>
>> Commit: 871de1a7665f766fbcdf2071187be9a1d4c23eb7
>>     https://github.com/D-Programming-Language/druntime/commit/871de1a7665f766fbcdf2071187be9a1d4c23eb7
>> Author: Walter Bright <walter at walterbright.com>
>> Date:   2011-01-24 (Mon, 24 Jan 2011)
>>
>> Changed paths:
>>   M posix.mak
>>
>> Log Message:
>> -----------
>> update MANIFEST
>> 
>
> IMHO, this is the wrong direction for makefiles in general.  Duplication of long lists is just asking for them to grow
> out of sync.  Why did you unroll the $(addprefix src,$(addsuffix .d,$(SRC_D_MODULES))) line?
>
> 

I didn't. Did I? github is down :-( can't check
January 24, 2011

Brad Roberts wrote:
> IMHO, this is the wrong direction for makefiles in general.  Duplication of long lists is just asking for them to grow out of sync.

I like and use manifests because:

1. Otherwise, it's an error prone pain to figure out what files are
source and what are generated
2. I use the manifest to copy druntime sources between systems
3. I use the manifest to copy the files to the release directory
January 24, 2011
On 1/24/2011 10:13 PM, Walter Bright wrote:
> 
> 
> Brad Roberts wrote:
>> IMHO, this is the wrong direction for makefiles in general.  Duplication of long lists is just asking for them to grow out of sync.
> 
> I like and use manifests because:
> 
> 1. Otherwise, it's an error prone pain to figure out what files are source and what are generated
> 2. I use the manifest to copy druntime sources between systems
> 3. I use the manifest to copy the files to the release directory

What I'm objecting to is the removal of by reference inclusion of SRC_D_MODULES into MANIFEST and instead duplicating the whole list.  MANIFEST should use references where possible and list only what's unique to the MANIFEST directly.

Makefiles are just as important as source files.. good factoring goes a long way towards maintainability.
January 25, 2011

Brad Roberts wrote:
>
> What I'm objecting to is the removal of by reference inclusion of SRC_D_MODULES into MANIFEST and instead duplicating the whole list.  MANIFEST should use references where possible and list only what's unique to the MANIFEST directly.
>
> Makefiles are just as important as source files.. good factoring goes a long way towards maintainability.
>
> 

Fair 'nuf
January 25, 2011

Brad Roberts wrote:
>
> Why did you unroll the $(addprefix src,$(addsuffix .d,$(SRC_D_MODULES))) line?
>
> 

It didn't work. The std/gc/*.d files were not included, which is why I was having the gc failures.
January 25, 2011
On 1/25/2011 1:11 AM, Walter Bright wrote:
> 
> 
> Brad Roberts wrote:
>>
>> Why did you unroll the $(addprefix src,$(addsuffix .d,$(SRC_D_MODULES))) line?
>>
>> 
> 
> It didn't work. The std/gc/*.d files were not included, which is why I was having the gc failures.

Yup.. slightly buggy.  The proper fix:

        import/object.di \
        import/std/intrinsic.di \
        \
-       $(addprefix src,$(addsuffix .d,$(SRC_D_MODULES))) \
+       $(addprefix src/,$(addsuffix .d,$(SRC_D_MODULES))) \
        \
        src/core/dll_helper.d \
        src/core/thread_helper.d \