home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ukma!darwin.sura.net!zaphod.mps.ohio-state.edu!magnus.acs.ohio-state.edu!usenet.ins.cwru.edu!cleveland.Freenet.Edu!an662
- From: an662@cleveland.Freenet.Edu (Michael Leach)
- Newsgroups: comp.databases
- Subject: Clipper and record locks
- Date: 13 Nov 1992 23:57:26 GMT
- Organization: Case Western Reserve University, Cleveland, Ohio (USA)
- Lines: 38
- Message-ID: <1e1fd7INN1j3@usenet.INS.CWRU.Edu>
- NNTP-Posting-Host: hela.ins.cwru.edu
-
-
- After reading the discussion on how Clipper implements it's
- record locks, I can't agree or disagree completely, nor do I
- know (beyond all doubt) how Clipper handles them, but I can
- state several things:
-
- 1. The EXE does NOT store the record locks internally. If it
- did, how would clip1.exe know what records clip2.exe has locked
- if both are accessing the same file at the same time.
-
- 2. The DOS SHARE command (and similar type things on networks)
- do not all use the same method for locking ranges of bytes.
- Most do hook into the same DOS interrupt (I don't remember whick
- one).
-
- 3. Clipper does not tell the server (or pc or whatever) to lock
- a range of bytes, the way some database managers do. If they
- did, 3Com's 3Share network would be able to detect it, and it
- doesn't.
-
- I remember a discussion about a year and a half ago at a user's
- group meeting on this subject, and the only answer that anyone
- could get from Nantucket on the subject was that Clipper locks
- a record by telling the system that a single byte (or bytes)
- is (are) locked. This byte is outside the range of the offset
- of the actual record. Nantucket wouldn't say where or how
- this info was stored, or what was stored in it beyond the fact
- that it indicated what record was locked. The general consensus
- at the group was that it was stored (somehow) within the database
- (we didn't agree on where). Since DBFSIX allows multiple record
- locks within a single workarea that are Clipper compatible,
- I would assume that Successware has figured out (or coaxed it
- out of CA/Nantucket) what it does.
-
- Mike Leach
- an662@cleveland.freenet.edu
- leacmj@morekypr.bitnet
- leachmj@uc.edu
-