On Wed, Dec 14, 2016 at 6:12 PM, Erik Bray <erik.m.b...@gmail.com> wrote:
> On Wed, Dec 14, 2016 at 5:58 PM, Peter Luschny <peter.lusc...@gmail.com> 
> wrote:
>>> Please give this new build a try:
>>>
>>> https://github.com/embray/sage-windows/releases/download/0.1a3-7.4/SageMath-7.4.exe
>>>
>>> This was built incorporating the fixes pointed out by Dima for Givaro
>>> and FFLAS-FFPACK.  I don't see any binaries built without
>>> -mtune=generic anymore but there could still be some other
>>> "misbehaved" packages w.r.t. SAGE_FAT_BINARY--I don't know.
>>
>>
>> Yay, it works! None of the errors observed so far.
>> Well, it works almost. Some new problems appeared while
>> executing the worksheet:
>>
>> SagePtyProcess.spawn, pid = os.fork() and such alike ...
>
> Yay!  And boo!
>
> It occurs to me that I may not have done a DLL rebase after the last
> rebuild (if you're not sure what that is don't worry about it), so
> that would explain any problems with os.fork().  In this case it was a
> shortcoming of not having a fully automated build process yet.  I'll
> rebase the last build and re-upload it, hopefully at the same filename
> if GitHub will let me.

I replaced the existing binary with a new one at
https://github.com/embray/sage-windows/releases/download/0.1a3-7.4/SageMath-7.4.exe
that was built "correctly" (as far as my current procedure is
concerned) and should not have (as many) problems with os.fork()
(unless you've revealed a new issue I haven't seen before which is
entirely possible when it comes to Cygwin and fork, though the vast
majority of fork() issue's I've seen have been resolved by rebasing).

Thanks,
Erik

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Reply via email to