On Tuesday, 30 May 2023 at 06:09:23 UTC-7 Luis Finotti wrote:
I looked at it, but I cannot tell if it is really the same issue, or how I could check. well ... it looks like the symbol that is found undefined is indeed of the type mentioned there. So that would suggest that indeed a system givaro is picked up while sage is apparently compiled with a "too advanced" c++ ABI for that library. However, you're not getting errors when you just run sage-10.0, so it would seem there is a givaro library that can satisfy the needs (it probably built its own givaro) So in your case there seems to be something about the environment in which jupyter starts sage that confuses the library resolution process. I imagine it could be something where juoyter runs using a python that has its own venv and that the initialization of sage (which should set up its venv) somehow misses overriding something or erroneously inherits something from the jupyter venv. In that case I would expect this issue was basically always present, but has only recently become relevant because now the system givaro doesn't cut it any more. -- You received this message because you are subscribed to the Google Groups "sage-support" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-support+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/sage-support/24724ea0-edcc-4470-8b30-56cde325caf2n%40googlegroups.com.