mathog at seqaxp.bio.caltech.edu mathog at seqaxp.bio.caltech.edu
Thu Jun 15 18:39:17 EST 1995

In article <19950615124144.bloksber at thomashaw-at.css.msu.edu>, bloksber at pilot.msu.edu  (Leonard N. Bloksberg) writes:
>We switched from VMS to UNIX on an SGI about 2yrs ago, and
>I've been very happy with it.  Unix makes it easier for me to write batches,
>aliases, and other tricks to make it easier for the computer illiterate 
>folks in the lab to use GCG.

Ie, *you* prefer Unix to VMS and are more comfortable with its scripting 

>As archane as GCG can be, it is now (finally) 
>in general use in the lab, while under VMS the staff considered it unusable.

As far as the command line user interface goes, the only major difference
on the two platforms is the use of / for -.  Unless your screens have some
obscure polarization problem, it's hard to imagine this tiny change
resulting in a significant increase (or decrease) in usability. 

I'd guess that within WPI the only clue as to the underlying system
would be the filenames shown.

An earlier post referred to a difference in speed on Unix vs. VMS.  I'd
believe that on modern RISC unix vs. old VAX VMS.  Very hard to believe
though for the current AXP machines.  Rather than bullshitting about it, if
anybody out there has a 190 Mhz 2100 server running Digital Unix please
contact me and we can run some comparative GCG benchmarks on your machine
versus mine and get an answer to this question. That comparison would be
a straight test of GCG's speed on VMS vs. Unix, no hardware considerations. 
We could compare between some other architectures if you want, but that
will be less of an OS speed comparison than a hardware one.

Having said that, if you keep your database files on the VMS system in the
format that GCG ships them, then yes, you will have a performance penalty. 
GCG documents that, and they provide a tool to reformat the databases into
a more efficient form.  (Ie, you have the choice).  The rationale for this
common format (really, a Unix format) was that some sites could then share
a single database, presumably over NFS, between GCGs running on both Unix
and VMS machines.  Is anybody out there actually doing this?  It seems like
a spectacularly bad idea.  I wouldn't want people shoveling GENBANK back
and forth over my nets every time they ran a FASTA!!! 

In the end it doesn't really matter much if it's Unix or VMS, since these
days what the users really want is a Mac or Windows interface, not any
flavor of command line interface.  I suspect that they grouse pretty much
the same about one versus the other. We have no plans to move our site to
Unix unless forced to.  However, if one ever became available at a price we
could afford, we'd happily chuck it all for a decent client/server


David Mathog
mathog at seqvax.bio.caltech.edu
Manager, sequence analysis facility, biology division, Caltech 

More information about the Info-gcg mailing list

Send comments to us at biosci-help [At] net.bio.net