[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: vienna-rna is almost ready



On Sat, Feb 05, 2011 at 12:05:17PM +0100, Steffen Möller wrote:
> 
> I still think we were too nice back then. We should have a policy
> that scientific binaries shall not be renamed if not in conflict with
> essential packages.

I do not consider this as a valid suggestion because there is no
guarantee at all that the "scientific namespace" is not polluted.  If we
assume for a moment it would be clean there is a brain dead easy
workaround:

Create a /etc/profile.d/scientific-workstation with the content

   export PATH="/usr/share/sciencebin:$PATH"

and hava a common science policy which links any scientific binary
with the (original) name to this place.

> Our focus should be on collaboration. Not with some other remote part
> with Debian, but with the very close scientists all over the world.

IMHO there ar no different parts in Debian but it is the nature of "the
universal operating system" that you have to find reasonable technical
trade-offs and the solution I suggested above is simple enought that Joe
Scientist will probably not even notice this issue (except if he wants
to use for instance the now hidden /usr/bin/plink from putty - but at
least he has the chance whic he would not with your suggestion).
 
> My stance: no renaming.

Back to the original issue: The binary in question just has a generic
name and has probably by far a way lower user base than it might be
the case for plink.  The accompanying README says explicitely:

  This directory contains a few scripts, you might find useful.
  They're not installed by default.

So *if* we even install these scripts we do more for the user than
upstream suggests.  Thinking about a wrapper is in *this* specific
case a completely valid thing to do (even if I'm personally not sure
what might be the correct advise here).

> And we should fight for it more.

I refuse to spend my time in fights were I see no chance to win and
an obvios workaround for the conflict is avialable (see above).  The
time could better been spended in writing a

   dh_installsciencebin

wrapper which considers files like

   debian/<packagename>.sciencebins

which mentions necessary renamings or something like this.

Kind regards

     Andreas.

-- 
http://fam-tille.de


Reply to: