BioBit No 12 (LDBASE for BIONAUTS)

"Robert Harper ", Finland HARPER at finfun.bitnet
Fri Dec 29 04:26:00 EST 1989

     1212121212                       1212121212
     1212121212                       1212121212
     1212  1212   1212   1212121212   1212  1212   1212   121212121212
     12121212            1212121212   12121212            121212121212
     12121212     1212   1212  1212   12121212     1212       1212
     1212  1212   1212   1212  1212   1212  1212   1212       1212
     1212121212   1212   1212121212   1212121212   1212       1212
     1212121212   1212   1212121212   1212121212   1212       1212

                                 No 12
                     << EDITED BY ROBERT HARPER >>

   LISTSERV at IRLEARN is much more than a programme that distributes mail.
   It can also function as a file server and a database. This edition of
   BioBit will look at how you can have an interactive session with
   LISTSERV to retrieve old messages.

   The problem that I was faced with was this... I often get requests for
   back issues of BioBit, and the folks at EMBL decided to make BioBit
   available on their server. The only trouble was that issues  No. 1 and
   No. 3 were missing... yes things do occasionally disappear over the blue
   event horizon... so I had to find the old issues from somewhere.

   I had read about LDBASE on the network but never had the inclination or
   the opportunity to try it out. So the first thing  to do is to get the
   software that allows me to have an interactive session with LISTSERV.
   Here is how to get the programmes for CMS and VAX.

   The  database can be accessed either interactively or in "batch" mode.
   In the former case, you must  obtain  the  LDBASE  user  interface  by
   sending the following commands to your nearest LISTSERV:

  o   For VM/SP CMS systems:

      The command to start the user interface is simply LDBASE to access
      your "home" server, or LDBASE nodeid to access the LISTSERV server
      at nodeid.

  o   For VAX/VMS systems:
      -   SEND LISTSERV at nodeid GET LDBASE COM

      The  command  to  start  the user interface is @LDBASE.  This will
      install the required files in  your  directory  and  display  more
      detailed instructions about the program.

  o   Other systems may not presently access the database in interactive

   I obtained the LDBASE.COM from FINHUTC and executed it to get a small
   LDB.COM.  To create the .EXE files execute the BUILD.COM file
   (@BUILD). To remove the source files you no longer need use @CLEANUP.
   Also included are some INFO files. Once you receive the LDBASE on the
   VAX and run it, it unpacks to give you all the files needed to run the
   system. For example:

$ receive *.*
from:   LDBASE.COM;1
$ dir


Total of 1 file.
$ @ldbase
%TPU-S-FILEIN, 34 lines read from file SYS$INPUT:.;
Unpacking BUILD.COM
Unpacking DBS.FOR
Unpacking LDB.COM
Unpacking LDB.INFO
Unpacking SEND.FOR
Unpacking SMG.DEF
Unpacking TRAP.CSTR
Unpacking TRAP.FOR
Unpacking TRAP.INC
Unpacking TRAP.MESS

    On VMS/VAX you can call up  the LDBASE with the command @LDB
    which then spawns a sub-process. Here is a LDBASE session from
    the VAX.

$ @ldb
%DCL-S-SPAWNED, process HARPER_1 spawned
?  short help
$  to issue dcl commands
*  for list of databases (LISTSERV 1.5n or later)
^Z to leave and/or switch listservers

   I was  presented with a Server> prompt and asked to fill in the name
   of the LISTSERV that I wanted to talk to. Since all of the BIOSCI
   messages are stored at IRLEARN then that is where to look. If you
   give the "*" command you will be given a list of ALL the databases
   that can be accessed at IRLEARN. Some of them may be confidential and
   only open to query by members, and not by the public at large.

Server>LISTSERV at irlearn
start? [Y]y

   Once you have answered Y to start the session you can begin to query
   the database. I knew that some of the BioBit newsletters went into
   Bionews so lets look there first of all.

   It is nice that the syntax for search is plain and simple english
   $BIONEWS for the word BIOBIT. (By the way you should note that all
   the BIOSCI lists have stoped using $ signs and now use + signs...)

db>search biobit in $bionews
Welcome to LISTSERV at IRLEARN - Release 1.5o (FIX15O1), backbone server.
CPU model 0580, DASD model 1280.
--> Database $BIONEWS, 11 hits.

   LISTSERV replies by saying telling you its version number (in Europe
   it is v1.5 while in USA it is v1.6... but that is an entirely
   different story) and the type of machine it is running on.  Within a
   few seconds it will also report back if there have been any HITS.

   If you obtain some hits you should then run an INDEX on them to see
   what they are. It should be noted that even though the hit is
   recorded in the main body of the text, you are only given the SUBJECT
   line by LISTSERV. So it is quite important that you give good subject
   lines that will help others when they want to do searches on the
   database. Item # 000083 does not have a subject so it is difficult to
   tell what the content of that message might be.

   As you can see I did not give a very good title, but since I know
   that BioBit No 3 dealt with the TRICKLE servers it was quite easy to
   see that Item # 000061 dealt with that subject. LISTSERV also gives
   you the date and time when the message was entered, and how big the
   item is.

Item #   Date   Time  Recs   Subject
------   ----   ----  ----   -------
000015 88/10/21 11:22  497   BIOBIT No2
000061 89/01/02 15:37  473   TRICKLE (The RED server)
000079 89/01/19 16:40   59   RED-UG (New list for discussing TRICKLE)
000083 89/01/19 17:09  116
000148 89/03/21 15:12  398   BIOBIT No 4
000149 89/03/21 23:19   35   Re: BIOBIT No 4
000150 89/03/22 13:37   35   Re: BIOBIT No 4
000151 89/03/22 16:46   13   Biobit
000329 89/08/23 18:07  179   BIOBIT No 9
000344 89/09/18 12:15  235   BIOBIT No 10 (staying alive in the network jungle)
000379 89/11/06 14:37  352   BioBit No 11 (Your friendly neighbourhood server)

   But just as a check you can print out the first twenty lines or so to
   find out if you have got the right article. The PRINT command should
   only be used to recover short messages. From experiance if the  RECS
   column in the INDEX is 26 then it is possible to retreive the whole
   of the message interactively. Anyway you can look at the beginning
   of it by giving the command PRINT ITEM #

db>print 000061
>>> Item number 61, dated 89/01/02 15:37:00 -- ALL
    Mon, 2 Jan 89 15:37:00 O
    TRICKLE (The RED server)

     3333333333   333   3333333333   3333333333   333  3333333333333
     3333333333         3333333333   3333333333        3333333333333
     333    333   333   333    333   333    333   333       333
     333333333    333   333    333   333333333    333       333
     333333333    333   333    333   333333333    333       333
     333    333   333   333    333   333    333   333       333
     3333333333   333   3333333333   3333333333   333       333
     3333333333   333   3333333333   3333333333   333       333

                              No 3

db> Output too long for interactive mode, use the SENDback command.

   So now I have the proof that item 000061 is the message that I want
   to retrieve. LISTSERV is not all that happy that I tried to print the
   item interactively and politely informs me that I  should use the
   SENDBACK command to retrieve item 000061. The RECS in the INDEX for
   ITME 000061 was 473... so I should have known better not to try and
   PRINT it. The proper way to retrieve a large item is to simply adding
   SENDBACK in front of the PRINT command.

db>sendback print 000061
* File "DATABASE OUTPUT" has been sent to you in Netdata format.

   And within a couple of seconds LISTSERV tells you that a file has
   been sent to you, and depending on the network traffic you can get
   the file within half an hour or so.

   I repeated the process and found that BioBit No 1 was entered into
   the $BIOTECH list... in duplicate... seems we had mailing loops back
   in 1988, and recovered it by the same procedures. Once the files
   arrived in Helsinki I then posted them to EMBL so if anybody wants
   back issues of BioBit they can get them from there by mailing the
   server at NETSERV at EMBL.

   While I was writing this edition of BioBit it struck me that you
   could do literature searches on the BIO+JRNL list, and I had the idea
   that this database function of LISTSERV could be used to set up a
   database for "BIONAUTS" (biologists who know their way about the
   network and who want to share their skills and expertise with
   others... or for anyone who reads the BIOSCI Bboards). No-one ever
   uses NAMESERV... so something aimed specifically at bio/computing
   users might be appropriate and timely. A "BIONAUTS" database would
   provide a convenient "Who's Who" :-) And it would be a confidential
   list so only those subscribed to it could get information out of it.
   People do get paranoid about the information in databases:-)

   Way back in the days of SEQNET someone from Cambridge began to
   collect information on scientists who were using the networks. I do
   not know if this database still exists or if the information  in it
   was ever made public, however the idea was very simple. Namely any
   scientist could volunteer information about himself, concerning his
   PROFESSIONAL SKILLS, ADVISE ON, CAREER etc... if I remember correctly
   there was quite a good response from the SEQNET community.

   Well lets say a BIONAUTS list existed to which anyone on BITNET/EARN
   could SUBSCRIBE. The database could then be used as a knowledge pool
   of information... a SEARCH CALIFORNIA in BIONAUTS would find network
   contacts in California. This way you could perhaps find someone in
   your backyard who could help you on some particular matter, rather
   that E-mailing someone on the other side of the world.

   Anyway I would like to know if there is any real interest in setting
   up a BIONAUTS list, and what people would want for the different
   database fields. If there is sufficient interest I will see what I
   can do about setting up a list.

   If you have any opinions on the matter then please feel free to
   express your ideas here in BIONEWS, and if you want to indicate that
   you would like to see a BIONAUTS list established then drop me a note
   to HARPER at FINFUN simply saying YES for BIONAUTS in the  subject line.
   If I get enough votes then I will definately do something about it.

-=ROB=- %% Robert Harper %% HARPER at FINFUN %% HARPER at CC.HELSINKI.FI %%

More information about the Bioforum mailing list