more on Entrez interfaces and CDs and Networks

[jasper at molbiol.ox.ac.uk]

In article <1993May12.070047.7211 at mcclb0.med.nyu.edu>, smith at mcclb0.med.nyu.edu writes:
> In article <1993May11.052759.29180 at nic.funet.fi>, harper at convex.csc.FI (Rob Harper) writes:
>> 
>>>However, the big problem, as Bill points out above, is
>>>networking!  Should every individual who wants to run Entrez be forced 
>>>to buy a CD reader?  

No!  they should get a network connection. (Sorry Bill I know that was 
rhetorical).

>> 
>> 	No! It is quite simple to give lots of people access to a
>> 	CD reader.

Given that Rob's SGI is *on* the network I don't see the logic of this (maybe 
it's the flashing lights on the box :)  why not run the networked version?

Going back to two issues here though: 

1. GUI's - sadly I think it would be useful to have a character cell interface 
(no necssarily command line, unless I understand something else by that than 
what is implied), for all the people who haven't got X on their desk, but want 
*sequence* data on their desk in their unix/VMS account filestore.  This is 
unix and VMS, but not Mac's or PCs.

For Macs and PC's it's neat if you have a sequence analysis package on them.
But most of us don't (have, need or want).  What we *do* have there is papers,
reference databases, and the need to get Medline onto the desk. Entrez does
this wonderfully, but I probably don't need this part on my unix/VMS account. 

So there are two distinct uses here for Entrez. We want them both in the
appropriate enviroment. And that actually means if both are there in both
environments that is great. So a character cell interface fills the lst gap in
that picture.  

But keep it as one interface at this level, I'm not suggesting splitting the 
functionality. Just get a character cell one that looks close to the 
X/Mac/Windows one.

2. Networked or CD versions.  This is a no contest one for networked machines. 
The network version is about 10 times faster in response to the fastest 
network CD servers, even its running through the transatlantic link on a bad 
day. And you have the equivalent of the SEQDATA and REFDATA CD's and they 
updated everynight. So unless you are off a network, and can't get to one, the 
CD's are not the best way.

Anyway, I think its the most useful tool I've gained in the last 5 years 
(maybe more!) and I'm seeing a very rapid rate of take up in use here, even 
without getting documentation to people, or even telling them we installed it 
on the machines! So I rate that a very big sucess.

Jasper Rees
Oxford University Molecular Biology Data Centre
and The Sir William Dunn School of Pathology

jasper at molbiol.ox.ac.uk