David Kristofferson kristoff at
Wed Jul 17 20:23:20 EST 1991

> Dave Kristofferson has said that releasing the Authorin format to the general
> programming public in order to write a VMS-version of Authorin is not a
> good idea. I disagree. The need is obviously there and without an Authorin-
> lookalike, VMS sites (and perhaps UNIX sites) are going to continue sending in
> the Genbank form instead of an Authorin-generated file thus causing more work
> for LANL and GenBank. While I haven't looked closely at the output from 
> Authorin, it seems to me that there must be some form of data verification 
> built into it. If Genbank released this verification scheme to a few VMS 
> programmers, I'm sure an Authorin-lookalike could be safely created.

We will do whatever NIH requests us to do in this regards. (Is the
government looking for "a few good VMS programmers" now 8-)?  I still
think that the idea of releasing it "to the public" to allow everyone
to take a stab at it is silly.  NIH should identify who they want to
have do this version and put the means in place to do it if it is
necessary.  I believe that there was some discussion in the past about
having GCG write the VAX version.  We could also obviously do it at
IG, and I am sure that there are others who could also accomplish the
task.  The final decision is not in our hands (IG's, that is), but I
strongly believe that because the program must be done in
collaboration with the database effort, a formal means should be
established for doing it (i.e., not the "thousand flowers" approach).

> Dave, if you are reading this, what do the people who use the UNIX 
> Intelligenetics package do to submit data to Genbank?

First, the IntelliGenetics Suite package is both a VAX *and* UNIX
package, so *ALL* of IG's Suite users on both platforms are in the
same boat described earlier by Tony Kyne.  I hope that you all realize
that the choice of hardware platforms for Authorin was made in
conjunction with the GenBank advisory committee and the NIH project
officers.  The platforms that IG uses for its commercial software have
absolutely no bearing on this.

The user community lately has been making more noise about a UNIX
version than a VAX version, and the opinion among the advisory groups
has favored a UNIX version as the next platform *IF* anything else is
done in the current contract on Authorin.  John Devereux was at the
last meeting at which this was discussed.  Even DEC is moving more
towards UNIX, and GCG itself is porting its software to UNIX.  It will
take more than a couple of BIONEWS postings to change this opinion
unless the NIH decides to do both versions.  We can possibly knock out
one more port of Authorin under our current contract, but I doubt that
we can do two without additional resources.  If only one more version
is produced, the current thinking is that it will be a UNIX version.
If NIH doesn't have resources for a VAX version, then some public
minded group might do it, but it would require a *high* degree of
dedication to the project if it is to be done AND maintained/supported
correctly, i.e., one which I would think hard to sustain over time
without the availability of some funding.


				Dave Kristofferson
				GenBank Manager

				kristoff at

More information about the Bioforum mailing list