Re: High Sierra

Alexander Hansen <[email protected]> Wed, 13 Sep 2017 08:22:38 -0700
Newsgroups gmane.os.apple.fink.devel
Message-ID <[email protected]>
I don’t disagree.  

However, my question was specifically targeted to the fink package, because that has a more involved release procedure which either I have to do or train someone else to do.  

Most other packages simply need to be updated in the distribution by someone who can commit them, though base packages need a bit more oversight.

—akh

> On Sep 13, 2017, at 06:25, Jack Howarth <[email protected]> wrote:
> 
> Alexander,
>    FYI, I wouldn't hold my breath waiting for upstream FSF gcc to
> rewrite their time-stamping scheme just for darwin users. It is
> interesting that the APFS filesystem seems to be the only one out
> there to tickle this latent race condition in c++ header
> installations. Guess every other file system is still in seconds
> regarding its time granularity.
>          Jack
> 
> On Wed, Sep 13, 2017 at 9:22 AM, Jack Howarth <[email protected]> wrote:
>> Alexander,
>>      We definitely should commit the changes to the gcc5, gcc6 and
>> gcc7 packages that I proposed in...
>> 
>> https://sourceforge.net/p/fink/package-submissions/4990/
>> https://sourceforge.net/p/fink/package-submissions/4989/
>> https://sourceforge.net/p/fink/package-submissions/4988/
>> 
>> that patches libstdc++-v3/include/Makefile.in to add
>> 
>> .NOTPARALLEL: install-headers
>> 
>> in order to avoid https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81797
>> on machines using the APFS file system.
>>     Their scheme of time stamps to track the installation of the c++
>> headers during the build only checks for the existence of the
>> sub-directories created to install these headers in rather than their
>> population state. The nanosecond granularity of the APFS file system
>> breaks that scheme and requires the c++ header installation to be
>> single threaded to avoid race conditions under the new file system.
>>                      Jack
>> 
>> On Tue, Sep 12, 2017 at 6:15 PM, Alexander Hansen
>> <[email protected]> wrote:
>>> Since we’ve got the official High Sierra release date now, we should think about a fink-0.42.0 release.
>>> 
>>> 1) Currently, we will be sharing the 10.9-libcxx distribution for .info/patch files.  If anybody knows of a compelling reason that we should utilize a new distribution, now’s the time to mention it.
>>> 2) The high-sierra-branch branch on Github is the place for any updates.
>>> 3) It’s probably worth ending 10.9 support at this time, unless there are strong objections to this.
>>> 
>>> --
>>> Alexander Hansen, Ph.D.
>>> Fink User Liaison
>>> 
>>> 
>>> ------------------------------------------------------------------------------
>>> Check out the vibrant tech community on one of the world's most
>>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>>> _______________________________________________
>>> Fink-devel mailing list
>>> [email protected]
>>> List archive:
>>> https://sourceforge.net/p/fink/mailman/fink-devel
>>> Subscription management:
>>> https://lists.sourceforge.net/lists/listinfo/fink-devel


------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Fink-devel mailing list
[email protected]
List archive:
https://sourceforge.net/p/fink/mailman/fink-devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel