Re: Any risk in increasing BLCKSZ to get larger tuples?

Joseph Shraibman <jks(at)selectacast(dot)net> writes:
> since WAL isn't ready yet and TOAST is,
> why not release 7.1 now and release 7.2 in January with WAL?

Unfortunately that's not the scenario we're in. What we've got is
beta-quality TOAST, beta-quality outer joins, and a long list of
unrelated bugs still to be fixed before 7.1 can go out the door.
Even if we decided today that we wanted to make a release without WAL,
I don't think we'd be ready to go beta much before the end of this
month, and certainly no final release less than a month after that.
The reason is that other people have been designing their schedules
around WAL --- for example, if we'd been trying to make an Oct 1 beta
date, I'd have been doing bug fixes not outer joins last month.

Pulling WAL would allow us to move up 7.1 release somewhat, but I don't
think two months worth.

Furthermore, if we did do it that way then 7.2 wouldn't happen in
January. A beta test/release cycle consumes a lot of developer time,
which would translate directly to pushing back the 7.2 dates.

If we'd foreseen the schedule more accurately back around July, we
might've chosen to push out a 7.1 with only TOAST, no WAL or outer
joins, but it's too late for that now.

regards, tom lane