Error when making fur

[salientrob]

Hello, thanks for making and sharing Fur!

Please could you help me with an error that occurs when I try to install fur?

I have installed make, gloang, noweb and git, and cloned fur (git clone https://github.com/EvolBioInf/fur.git). Then when I use the make command most of the packages are installed successfully but there is an error for makeFurDb: divsufsort64.h is not found.

Thanks for your help!

Rob

[Acantha-Slayer]

Hey, same here. And nothing I've tried has worked. Please send help.

[haubold]

Thank you for pointing out this error. The package libdivsufsort-dev was missing from the list of dependencies. This package calculates the suffix array underlying fur. I have now added libdivsufsort-dev to the list of dependencies in the README. So in your case, the programs of fur should build if you do

$ apt install libdivsufsort-dev
$ make

[salientrob]

Thanks!

For other novice users, to work through the tutorial I also needed to install:
ncbi-blast+
phylonium
primer3
These are in the documentation but might not be obvious.

[haubold]

Thanks for pointing out these additional dependencies. I've added just them to the README, too.

[jtclaypool]

I have also found when some of these are installed via conda, you need to add a flag to the Makefile like such:
https://github.com/jtclaypool/fur/blob/work-with-conda/makeFurDb/Makefile#L11

CGO_CPPFLAGS="-I${CONDA_PREFIX}/include -L${CONDA_PREFIX}/lib" go build -ldflags "$(vf) $(df)" $(exe).go

Could be added as variable like vf and df as well for a cleaner build line

[haubold]

Thanks for this hint. Unfortunately, I am not able to reproduce the
problem. I start from an empty docker container,

https://hub.docker.com/r/haubold/mix

Inside that container I install conda; under conda I clone fur, set it
up, and make it. All that works, so I must be missing something.

Could you please show me how to reproduce an environment that requires
the make option you propose?

[jtclaypool]

I feel a little silly because I have replicated this on two on my cloud instances but using two separate base images (what you supplied and a micromamba) I have absolutely no issues during the installation. I also tried named versus the base environment and could not replicate. I'm truly at a loss as to what I'm experiencing on my cloud instance.

I guess you can ignore my comment but I greatly appreciate you looking into it.