[212] in ad-lib
B.Index changes needed
daemon@ATHENA.MIT.EDU (sbyrd@MIT.EDU)
Thu Mar 23 13:47:50 1995
From: sbyrd@MIT.EDU
To: ad-cat@MIT.EDU, tat@MIT.EDU
Date: Thu, 23 Mar 1995 13:47:32 EST
Changes to B.Index
--Sam Byrd 950323
Here is a very much in-flux list of changes to B.Index for the 2nd
GMA, in no particular order. Please comment, question, add,
dispute, etc. [For purposes of this list, "/" = subfield delimiter,
so "/a" means "subfield a."]
[If the change has been made to Advance in the OPAC set-up, in
OSI (Online Search Setup, Search Indexes), I will say so.]
1. 600 /t should NOT be separately indexed as a title.
Delete 2nd layer of /n,p,s,t from 600
Change 600 dcode to 1MS (X no longer
neccesary if 600 only indexed once, correct?)
In other words, treatment of 600 should be
consistent with treatment of 610 and 611.
2. 130 should index as title.
Change dcode P to T
OR add dcode T
*Does P need to be there for auth. control
purposes? See discussion of this issue in my
B.Index questions e-mail of 950322.
[Fixed in OSI, but with attendant problem of displaying as
"Series" in search display]
3. 730 should index as title.
Change dcode P to T
OR add dcode T
*Does P need to be there for auth. control
purposes? See discussion of this issue in my
B.Index questions e-mail of 950322.
[Fixed in OSI, but with attendant problem of displaying as
"Series" in search display]
4. Generally speaking, all series should be searchable both as
series alone AND as titles. Most users don't know that their
citation to, say, "Lecture notes in chemistry" should be searched
as a series. They will likely search it as a series.
The most obvious way to fix this in B.Index would be to
add dcode value T to every dcode occurence of P. But because
there are other problems with specific series fields, I'm listing each
one separately here.
5. 440 should index as title as well as series.
Add dcode T?
[Fixed in OSI]
6. 830 should index as title as well as series.
Add dcode T?
[Fixed in OSI]
7. 800 /t should index as title as well as series.
Add dcode T?
[Fixed in OSI]
8. 810 /t should index as title as well as series.
Add dcode T?
[Fixed in OSI]
9. 811 /t should index as title as well as series.
Add dcode T?
[Fixed in OSI]
10. 400 /t should index separately as title.
400 should be consistent with 800 AFTER 800 is
fixed (see above).
Add second layer for /t that is consistent with 800.
Change dcode to 1AMPX (add X to indicate 2nd
layer).
[Fixed in OSI, sorta]
11. 410 /t should index separately as title.
410 should be consistent with 810 AFTER 810 is
fixed (see above).
Add second layer for /t that is consistent with 810.
Change dcode to 2AMPX (add X to indicate 2nd
layer).
[Fixed in OSI, sorta]
12. 411 /t should index separately as title.
411 should be consistent with 811 AFTER 811 is
fixed (see above).
Add second layer for /t that is consistent with 811.
Change dcode to 3AMPX (add X to indicate 2nd
layer).
[Fixed in OSI, sorta]
13. 840 should index as title as well as series.
Add dcode T?
[Fixed in OSI]
14. 780 should be indexed.
Not in B.Index at all. In Marcive index as Author.
Should /t be indexed as title? Or what?
15. 785 should be indexed.
Not in B.Index at all. In Marcive index as Author.
Should /t be indexed as title? Or what?
16. 711 dcode value should be 3AMX, not just 3AM, to indicate
2nd layer of index for /t (as done for 700 and 710).
17. 711 /t T index should include other title-specific subfields in
2nd layer index (like /n, /p, etc.) as done for 700 /t
and 710 /t. [Should I list all possible subfields
here?] BUT do not add /h to 2nd layer (see #22
below.)
18. 111 Shouldn't there be a "Join 240" program (Prep column) for
111 like there is for 100 and 110?
19. 870
871
872
873 In B.Index, these are indexed as
cross-reference (dcode = C).
--If they stay that way, will they interfere with authority
control processing we have done by a vendor? How would they
interact with cross-references generated from auth. records?
--Should they be indexed?
--Should we instead consider globally deleting 870-873
from entire database? Or adding them to strip program?
--Or should we just leave them alone, indexed as is, and
worry about problems later?
20. 700 /h should not index.
Delete /h from Valid column (in there twice;
once for 1AMX and once for T)
21. 710 /h should not index.
Delete /h from Valid column (in there twice;
once for 2AMX and once for T)
22. 711 /h should not index.
Delete /h from Valid column (in there once,
for 3AMX --do not add to 2nd layer T index when
fleshing out to be consistent with 710 and 711--
see #17 above).
23. 730 /h should not index.
Delete /h from Valid column.
