IUBio

GCG Justification

Tim Cutts tjrc1 at mole.bio.cam.ac.uk
Mon May 12 03:56:26 EST 1997


In article <5l6gdj$e3 at mserv1.dl.ac.uk>,
Adrian Jenkins  <ajenkins at nibsc.ac.uk> wrote:

>ADvantage of local GCG
>----------------------
>
>
>1. depending on your hardware, generallyyt much quicker and more
>convieniant than internet resources...

Yes and no.  Internet resources are generally much larger machines,
and as a result it takes more to saturate them.  Our machine which we
replaced over the weekend was fine if only a couple of jobs were
running, but once the load passed 10 it was a real snail.

>2.  If you local network is up and running, you have gcg access

This is a major plus.  With a local service (a) you have more direct
control over the facilities that are available and (b) you are
protected against (some) network problems.

>3.  Some very useful programs and utility tools 

Having said that, a lot of GCG programs are available as free software
as well, and in much more recent versions.  GCG's blast algorithm is
very old, and uses far more memory than it should.  This can really
cripple a small machine.

>Disadvantage:
>
>1 Dbases are only updated every 3 months...if you require updated dbases
>as in daily or weekly, you need to go through the process of updating
>yourself, though there are various scripts to help you do it as far as i
>am aware.

database updating is a doddle once you have done it once.  We do not
use *any* GCG-supplied databases, and I hardly need to touch anything;
once a week a cron job fetches the latest updates to EMBL, formats
them, puts them in the appropriate place and rebuilds the lookup
indexes.  I don't need to lift a finger.

>2. each release requires more hard disc to store the dbases...

True, but that would be the case regardless of what local software you
use.  Disk space is cheap these days.

>3. becoming pricey (as in licence etc)

This is very true, and something that GCG really ought to re-think.
They are charging us through the nose for software which is actually
not very current.

>4. Becoming X-window dependant very helpful and good but we havwe had
>problems with x-window software installed on PC's...

If you really want to do X on PCs seriously, you should install Linux
on the machines and use a proper UNIX X server that way.  This is what
I use in my office.  I reboot into Windows NT just occasionally when I
need to do some word-processing.

Besides, there are some good WWW interfaces to GCG in development.
W2H in particular is very impressive.

>Internet resources...
>------------------ 

>2.  Dbase updating is done automaitically as far as the end user is
>concerned

see above.. :-)

>Disadvantages:
>
>1.  Connections can be rather slow at times, from the uk trying to use
>an American fcility can be rather time com=nsuming

Not to say totally impossible.  :-)

>2.  keeping upto date with servers can be quite time cons=uming though
>various sites/personalities will have list of newer useful mol bio sites

Speaking as someone who has to try to maintain a list of such sites, I
agree, it's almost impossible.

>Currently where I work, most people use GCG, Staden and the internet
>resources side by side.

This is what happens here as well, although Staden is used almost
exclusively for sequencing projects, and GCG tends to be used for the
subsequent analysis.

Tim.






More information about the Info-gcg mailing list

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