[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Freeze exception for gdc-4.3





On 22 August 2010 16:20, Adam D. Barratt <adam@adam-barratt.org.uk> wrote:
On Sat, 2010-08-21 at 10:02 +0100, Iain Buclaw wrote:
> I'd like a freeze exception for gdc-4.3-1.060-4.3.5-1.
> Fixes #577598 and all D software can now build on sparc (closing
> #475857, and about a dozen others like it).

Was there no way of resolving the sparc issue without importing a new
upstream version?

Yes, the sparc issue is resolved via a local patch in the debian patches folder. The upstream version I feel is necessary as it includes a modern D frontend (1.060), which works with more emerging D projects and technologies out there - such as the tango library, xfbuild, and gtkD - making the compiler much more accessible for using in the long-term scope of development.
 

The diffstat between the testing and unstable packages is

 824 files changed, 133190 insertions(+), 141498 deletions(-)

which is practically unreviewable.  Ignoring whitespace changes in the
diff reduces it to

 767 files changed, 82109 insertions(+), 90417 deletions(-)

which is still very high.

The source tarball also houses the GDC D2 project too. And the diff will have shown all the major updates to that frontend and it's accompanying libraries, which includes the dmd2, phobos2 and druntime directory. To my count, this accounts for 560 of those files. None of which are used in the building of D1.

If you feel it be necessary to review the changes that actually _affect_ the compiler, just take note of changes to the source files in the toplevel 'd' directory, and the 'dmd' and 'phobos' subdirectories too. But as you have stated, the subdirectories are consumed by quite a number of changes to whitespace, line endings, and a few dozen file deletions. Anything else I haven't given mention to in the diff is not actually used.



(Your changelog also says "Update to SVN 20100714 from the
gcc-4_3-branch" which I assume is incorrect as the same text was
included in the changelog for 1:1.046-4.3.5-2)

Regards,

Adam

That was not my doing, sorry. :-)

Regards

--
Iain Buclaw

*(p < e ? p++ : p) = (c & 0x0f) + '0';

Reply to: