home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.vms
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!saimiri.primate.wisc.edu!ames!pacbell.com!unet!porsche!nowacki
- From: nowacki@porsche.net.com ( Michael Nowacki I.T. Database Contractor)
- Subject: Re: cdd, c, and rdb 4.1
- Message-ID: <1992Dec23.193113.12390@unet.net.com>
- Sender: news@unet.net.com
- Nntp-Posting-Host: porsche
- Organization: Network Equipment Technologies, Redwood City
- References: <9212231344.AA20322@top.magnus.acs.ohio-state.edu>
- Date: Wed, 23 Dec 1992 19:31:13 GMT
- Lines: 40
-
- In article <9212231344.AA20322@top.magnus.acs.ohio-state.edu> Ung-Ho Yi <ungyi@magnus.acs.ohio-state.edu> writes:
- >thank you to people who have responded to my c and cdd question.
- >
- >some one asked what the DEC said about my problem.
- >i only spoke with one technical support person, and he said i
- >should create another record for C language, which is the same advice
- >i got from most of you.
-
- one alternative would be to declare an integer in cdd that is the length
- of the data element, so that routines that need to declare buffers for
- the field in question can do the equivalent of
-
- char[sizeof(CDD_INT) + 1] buffer_fld;
-
- then you would have to have the 'c' source code ( a header file) declare
- all the 'c' versions of data. you have to have 2 different mechanisms to
- provide the same field in different languages. this option allows one field
- and 1 size descriptor, rather than 2 distinct field definitions suporting
- the same domain. just an alternative, they both are a nuisance.
-
- >I would also like to respond to the rdb v4.1 because we just went
- >through with much difficulty.
- >
- >first was segmented string. i am not sure about the exact problem,
- >but here it goes. our application program first stores 3
- > bytes as segmented string,
- >then later on the program adds more data on to the segmented string.
- >this has always worked with rdb v4.0, but on rdb v4.1 it corrupted
- >our database. the DEC engineers had manually remove the bad pages.
- >
- >second problem was with selecting rows. i heard this has to do with the
- >optimization of the rdb.... some how, when we select rows and
- >sort by some order, the query becomes very slow, much much slower then
- >when we were using rdb v4.0. so, we had to change some of our application
- >programs again.
-
- this news group is a great place to share the gory details of the querys
- involved and the indexes on the tables.
- Did DEC concede that theses are bugs and not side effects of the architectural
- changes in 4.1?
-