[m-dev.] mercury installation

[Julien Fischer]

Hi,

On Thu, 27 Jun 2019, Dimitar Kenanov wrote:

> came across mercury and read the intro book, i liked what i saw so decided to 
> give it a go :)
>
> Downloaded the source and tried to compile it. So i encountered the following 
> problem.
>
> I like to install my compiled programs in /opt. So i used prefix option in 
> configure. But then the Makefile complained that DESTDIR is undefined. So the 
> easy solution was to change the line in Makefile:
>
> from: $(MMAKE) MMAKEFLAGS=$(PARALLEL) DESTDIR=$(DESTDIR) install
> to  : $(MMAKE) MMAKEFLAGS=$(PARALLEL) install
>
> This way i get my /opt/mercury directory and bin,info etc are in it. Else if 
> i set up DESTDIR in Makefile some weird things are happening in INSTALL_DIR 
> :)

You can ignore the warning about DESTDIR not being set; setting the
destination directory via configure's --prefix option is sufficient.
(You shouldn't need to set DESTDIR at all.)

> Also figured out that setting up PARALLEL=-jN directly in the Makefile works 
> pretty well and speeds up the installation.
>
> I write to you guys because today i was reading about a user who was 
> complaining that the installation is a bit strange and slow. So i thought 
> this might help other users to install mercury a bit easier. So explaining 
> these things a bit better to the users might persuade them to give it go :)
>
>
> On the side, one practical question. Do you think mercury is suitable for 
> bioinformatics? I still cant decide.

I have no idea; I've crossposted this to the users list in case someone
there does.

Julien.