You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
src directory contains C++ and IDL code. IDL code is not currently maintained. C++ code set up to build via CMake using Sparse2D backend. Additional C++ code can easily be added with current set-up.
Not much new C++ code expected to be developed, so only really useful for preserving older codes
pycs directory contains Python code. At present this has been taken from multiple sources and some redundancy likely exists.
Plenty of work needed to clean up the code already included
Alternatively can start building this from scratch
Discussion
Should everything be centralised into a single package?
This will be difficult to maintain without proper Git practices (pull requests, etc.)
Not 100% obvious who benefits from this
May be difficult to motivate people to contribute without clear objectives
Should we instead have a several more focused packages (e.g. lenspack, something for theory, something for signal processing)
If so, how many packages?
Individually easier to maintain
Easier to have targeted focus
We can easily create a unified environment for separate packages
Should we consider contributing tools to bigger open source packages (e.g. AstroPy)
Significant work to meet standards
No package maintenance
What should the package name be?
Several places in the repo have different names (cosmostat, pycs, python-cs)
One name should be consistently used
The text was updated successfully, but these errors were encountered:
During a maintainers meeting (26/10/2020) @sfarrens, @EiffL, @martinkilbinger and @santiagocasas discussed the current content and scope of the package.
Current Status
src
directory contains C++ and IDL code. IDL code is not currently maintained. C++ code set up to build via CMake using Sparse2D backend. Additional C++ code can easily be added with current set-up.pycs
directory contains Python code. At present this has been taken from multiple sources and some redundancy likely exists.Discussion
Should everything be centralised into a single package?
Should we instead have a several more focused packages (e.g. lenspack, something for theory, something for signal processing)
Should we consider contributing tools to bigger open source packages (e.g. AstroPy)
What should the package name be?
The text was updated successfully, but these errors were encountered: