I'd like to make a lightweight install-less version of iRAP available via Bioconda. It would just clone the repo and make the scripts and Makefiles available, allowing management of software externally via Conda itself. In systems like Nextflow this is handy, where only the dependencies necessary for a given workflow step get installed, on the fly.
The main obstacle for this right now is the downloads folder. Even without installation, just cloning the repo with this folder is too heavy on storage. The best solution would be if this folder became a sub-module, so that it could be avoided while cloning.
Could you see your way clear to changing the repo structure in this way? I can do it on our fork, but then the sub-module referenced would be on our org, when it really should be on your account close to the main iRAP repo.
Hi Nuno,
I'd like to make a lightweight install-less version of iRAP available via Bioconda. It would just clone the repo and make the scripts and Makefiles available, allowing management of software externally via Conda itself. In systems like Nextflow this is handy, where only the dependencies necessary for a given workflow step get installed, on the fly.
The main obstacle for this right now is the downloads folder. Even without installation, just cloning the repo with this folder is too heavy on storage. The best solution would be if this folder became a sub-module, so that it could be avoided while cloning.
Could you see your way clear to changing the repo structure in this way? I can do it on our fork, but then the sub-module referenced would be on our org, when it really should be on your account close to the main iRAP repo.
Thanks,
Jon