Re: BioSquid - should it be removed from the archive?
On Thu, 22 Jan 2009, Charles Plessy wrote:
Le Thu, Jan 22, 2009 at 01:01:07AM +0100, Steffen Moeller a écrit :
I just learned from the upstream author of BioScuid that this is obsolete and
no longer maintained. Should it be removed from the archive? What are the
alternatives to using it?
Thanks for informing the list about this. Did upstream gave any details
about successors? Are there any reasons why it is obsolete (does it lead
to wrong data or whatever) or is just the code unmaintained. In the later
case this would not be a special feature of BioSquid - we have other software
in the archive that is orphaned upstream.
From the description "... number of small utility programs to convert, show
statistics, manipulate and do other functions on sequence files." it sounds
like something some scientists might have become really used to and sometimes
you can not imagine how long programms that are orphaned upstream might last
in practical use. So as long as the program does not have to be considered
"harmful" or "just wrong" I would be careful before we remove it. IMHO a
reasonable measure to transport the knowledge you gained about this program
is to put the following extension to the description:
Upstream author of BioScuid regards this program obsolete and does
no longer maintain the code. Debian will support ist users with this
code until the next stable release. Please see
/usr/share/doc/biosquid/README.Debian for more information.
In README.Debian we might tell something:
If nobody will raise up his voice that he is using the biosquid
package to the maintainer team at
Debian-Med Packaging Team <firstname.lastname@example.org>
we consider the package as not needed any more and will remove
it from the Debian distribution after Debian 5.0 (Lenny).
We recommend [package_x or package_y] which provide the same
functionality as BioSquid.
IMHO this is the right behaviour to keep our users happy. You should not
force them to stop using a program they are possibly used over years and
to what they are probably used to.
our hmmer package recommends biosquid, although I do not know to what extent it
would be a problem if it were not available. How about:
- Wait for Hmmer 3.0 release,
In any case we should wait with the removal until no package lists it in its
dependencies any more.
- remove biosquid,
- file an important bug on it in the meantime.
In how far would this bug be "important". The information that upstream
does not maintain a package does not deserve an bug of severity important.
The only BTS scenario I might imagine would be:
Wishlist: Please provide a package that replaces BioSquid
If there is a package that provides the same functionality as
BioSquid please provide wrapper scripts to provide the BioSquid
API in a BioSquid transition package.
My arguing would be different if there is some knowledge about severy
security problems of BioSquid or something like that.