Re: Digressions (Re: Bio-Linux 9)
On 09/07/18 16:12, Michael Crusoe wrote:
>
>
> 2018-07-06 5:16 GMT+03:00 Charles Plessy <plessy@debian.org
> <mailto:plessy@debian.org>>:
> [...]
> the problem with system-wide installations is important enough to be
> mentionned multiple times in the Bioconda paper just published in
> Nature Methods:
> [...]
Hi, Charles and Michael.
Bio-Linux is an APT-managed system, and Bio-Linux 9 will be based on
"med-bio" and "med-bio-dev". Although I'm running an APT-managed system,
I do sometimes find it difficult to live under the constraints imposed
by a single dependency tree with incompatible or out-of-date versions of
upstream software that I want to use...
I 'overload' APT-managed versions by giving precedence to software
installed in /usr/local or /opt to avoid breaking the APT dependency
tree by introducing package conflicts that can't be resolved. In my
view, Bioconda 'environments' are a generalisation of this approach. The
core of the system is still APT-managed, but other environments can be
superimposed on top of it without breaking the core APT-managed system.
What I think would be *good* is to package "Miniconda" for Debian-Med
and make it easier to bootstrap "Bioconda": The commonest problem I come
across with Bio-Linux is people using CRAN, CPAN or PIP to install files
manually as an admin that conflict with packages in the APT-managed part
of the filesystem, instead of putting them in /usr/local, /opt or their
own home directories - Bioconda provides a solution to that problem.
I would like to see us to get the best out of both APT and Bioconda!
Bye,
Tony.
--
Minke Informatics Limited, Registered in Scotland - Company No. SC419028
Registered Office: 3 Donview, Bridge of Alford, AB33 8QJ, Scotland (UK)
tel. +44(0)19755 63548 http://minke-informatics.co.uk
mob. +44(0)7985 078324 mailto:tony.travis@minke-informatics.co.uk
Reply to: