[Neuroimaging] Nibabel release imminent

[Samuel St-Jean]

Add (bool) option to do so but not enforce it, so everyone can be happy? 
It also won't produce undesired side effects, I remember when I bench 
stuff I always have a tons of symlink in the same folder (tractometer 
style stuff actually) coming from everywhere, and replacing those with 
new original files might break people stuff in hard to find out way.


Le 2016-07-30 à 03:35, Matthew Brett a écrit :
> On Fri, Jul 29, 2016 at 2:35 PM, Yaroslav Halchenko
> <lists at onerussian.com> wrote:
>> I would really appreciate feedback on my "selfish" idea:
>> https://github.com/nipy/nibabel/pull/466
> I'd like feedback too.
>
> In summary - at the moment, if you ask nibabel to write to a filename,
> that is symlink, it will write to the file pointed to by the symlink.
>
> Yarik would like nibabel to overwrite the symlink file, with the written image.
>
> Imagine 'afile.nii' is a symlink, pointing to a file 'somefile'.
> Current behavior with `nib.save(img, 'afile.nii') will be to write to
> `somefile`, leaving the symlink intact.   New behavior would be to
> replace the symlink `afile.nii` with the written image (and therefore,
> to orphan `somefile`.
>
> This seems like a pretty radical change to me - what do others think?
>
> Cheers,
>
> Matthew
> _______________________________________________
> Neuroimaging mailing list
> Neuroimaging at python.org
> https://mail.python.org/mailman/listinfo/neuroimaging