Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!think!mintaka!bloom-beacon!primerd!barry From: barry@primerd.prime.com Newsgroups: comp.sys.mac Subject: Info on Mac Databases Needed Message-ID: <160700093@primerd> Date: 7 Mar 90 20:11:00 GMT Lines: 66 Nf-ID: #N:primerd:160700093:000:3210 Nf-From: primerd.prime.com!barry Mar 7 15:11:00 1990 For a number of years I've been maintaining a large database (about 5K records) on a Prime minicomputer using Prime Information (Prime's version of the Pick DBMS). Prime Information and Pick are pseudo-relational databases that use hashed access to variable length records. In a properly structured database, any record can be retrieved in at most one disk read. The number of fields can vary from record to record as can the datatype and length of any field. New fields can be added to a database at any time. When a record is modified, the new field values aren't constrained by the previous values in the field. I'm not sure of the maximum length for a field, but it's at least 32K characters. All of these characteristics make Prime Information (or Pick) very well suited to maintaining a database where the fields are mostly strings that vary widely in length, e.g. between 10 and 400 characters, approximately for my database. The retrievals done on this database are of the form list records with field1 like "...string..." and field2 eq "a string" ... with various sorting options ('field like "...value..."' means field contains "value" as a substring). The retrieval language is comparable to a subset of SQL. One other feature of PI and Pick that is used in this database is repeating fields. A repeating field is one that can have 1-N instances it it. In effect, a repeating field is an array. The key point is the user doesn't specify an upper bound. For all practical purposes, the limit is infinite - it's at least 32K. I'd like to move this database to my Mac where access to it will be much more convenient. I've inherited a copy of Fox/Base Mac from a colleague who left the company and I've skimmed thru the Fox/Base manuals. Since Fox/Base requires fixed length fields, it doesn't appear appropriate to my needs. Fox/Base has a memo data type that allows very long strings, but it seems to be a second class citizen as far as searching is concerned. I couldn't find anything that appears comparable to repeating fields. I'm looking for suggestions on what Mac databases would be appropriate. The articles I've seen in MacWorld and MacUser don't indicate which Mac DBMSs require fixed format records and which, if any, allow variable length records or repeating fields of the sort described above. Thus far, Filemaker II seems to be the most appropriate choice. I'd appreciate hearing from anyone who's used Filemaker II (or anyother relevant DBMS) for a mostly textual database. Is Fourth Dimension overkill? I'll post a summary if the replies appear to be of general interest. Thanks in advance for your help, Barry ------------------------------------------------------------------------ Barry Wolman | barry@s66.prime.com Principal Technical Consultant | 500 Old Connecticut Path Prime Computer | Framingham, MA 01701 | 508/879-2960, ext. 1100 ------------------------------------------------------------------------ Nothing in this posting reflects a position of Prime Computer. /* End of text from primerd:comp.sys.mac */ /* End of text from primerd:comp.sys.mac */