Re: p4p and numpy 2.0

[Florian Feldbauer via Tech-talk <tech-talk at aps.anl.gov>]

Hey Torsten,

if you look e.g. here https://urldefense.us/v3/__https://github.com/epics-base/p4p/issues/145__;!!G_uCfscf7eWS!aDkmc0x5-iSMQd5LEQkoTvYILwIY10lVKq7YSiMyDPGb6EoJ-2Lfn2iK-GOShD9djI8UBllDuf_FK_HlLN1enyQCzT6F55mSM4Q$  or 
https://urldefense.us/v3/__https://github.com/epics-base/p4p/issues/156__;!!G_uCfscf7eWS!aDkmc0x5-iSMQd5LEQkoTvYILwIY10lVKq7YSiMyDPGb6EoJ-2Lfn2iK-GOShD9djI8UBllDuf_FK_HlLN1enyQCzT6FSaTbaX0$  this issue is known.
The current release on pypi.org (p4p==4.1.12) is still working with 
setuptools < 60 and thus numpy < 2, but the team is working on a new 
relase that will work with the new versions.
I guess p4p 4.2 will be released soon.

Cheers,
Florian


Am 26.09.2024 um 11:17 schrieb Torsten Bögershausen via Tech-talk:
> Folks,
>
> I just installed a fresh Linux box.
> Trying to install p4p did work, but it failed to work :-(
> The main reason seems to be that there is now a numpy 2.something,
> which seems to be incompatible.
>
>  ValueError: numpy.dtype size changed, may indicate binary 
> incompatibility. Expected 96 from C header, got 88 from PyObject
>
> Manually un-installing numpy and running
> pip install numpy==1.22.0
>
> made a success.
>
> p4p seems to be un-aware about the numpy 2 thing,
> it states happily
>
> Requires-Dist: numpy >=1.22.0
--
Ruhr-Universität Bochum
AG der Experimentalphysik I
Dr. Florian Feldbauer
NB 2/131 / Fach 125
Universitätsstr. 150
D-44801 Bochum

Office: NB 2/134
Phone:  (+49)234 / 32-23563
Fax:    (+49)234 / 32-14170
https://urldefense.us/v3/__https://paluma.ruhr-uni-bochum.de__;!!G_uCfscf7eWS!aDkmc0x5-iSMQd5LEQkoTvYILwIY10lVKq7YSiMyDPGb6EoJ-2Lfn2iK-GOShD9djI8UBllDuf_FK_HlLN1enyQCzT6FMYp4l8Y$