Subject RE: [IB-Architect] Generator bugfix RFP
Author Ann Harrison
At 02:09 PM 10/23/2000 -0400, Ann Harrison wrote:

>Aside from calling the new value dbb_gv_per_gp, that was my thought as
>well. Users should examine their generators

Sorry. Sloppy multi-processing. What I meant to say a new
element in the pgc structure of pgc_gpp, not a new element
in the dbb, which is pudgy enough as it is.

typedef struct pgc {
struct blk pgc_header;
SLONG pgc_high_water; /* Lowest PIP with space */
SLONG pgc_ppp; /* Pages per pip */
SLONG pgc_pip; /* First pointer page */
int pgc_bytes; /* Number of bytes of bit in PIP */
int pgc_tpt; /* Transactions per TIP */
SLONG pgc_gpg; / * generators per generator page */
} *PGC;

I recognize that changing the generator page to make it look more
like a pointer page by using bits at the bottom would allow us to
mark generators as "in use" or not, and thus drop generators. That
doesn't seem like a reason for an ODS change at the moment.