[m-rev.] for review: --trace-table-io-all

[Zoltan Somogyi]

On 02-Sep-2002, Fergus Henderson <fjh at cs.mu.OZ.AU> wrote:
> That comment should also go in the log message, as rationale.

Done.

> Does this actually causes the I/O to be tabled,
> or just it just generate executables which are capable of
> doing I/O tabling (based on the setting of some mdb option)?

The latter. In the absence of a "table_io start" command, no I/O tabling
actually takes place.

> But I'm not yet completely convinced.
> Doesn't I/O tabling cost a lot?

Without "table_io start", there is no space cost, and the time cost is only
one or two percent.

> Also, is it always desirable?

It makes "table_io start" meaningful, and that is desirable.

> For programs that just do I/O via the terminal,
> it may be nicer to redo the I/O on retries.

If you don't issue "table_io start", that is exactly what will happen.

In other words, this change adds flexibility at minimal cost.

Zoltan.
--------------------------------------------------------------------------
mercury-reviews mailing list
post:  mercury-reviews at cs.mu.oz.au
administrative address: owner-mercury-reviews at cs.mu.oz.au
unsubscribe: Address: mercury-reviews-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-reviews-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------