Ann Harrison wrote:

> Jim Starkey wrote:
>
>> I'm a little sorry to see the original journalling code was long
>> gone,
>>
> I also remember something about a performance issue?

Maybe the problem was also the very low performance of old hard disk?

> By the way - do get rid of all the WAL and journalling stuff.

I need to finish my current cleaning. I'm getting rid of some uppercase
typedef's.
Then, I will address Nickolay's cleaning needs. Being able to debug
correctly is a priority.
Finally, I will use Sean Leyne's award winning, patent pending, worldwide
famous, robust "monkey at the keyboard TM" approach, where I delete JOURNAL
and WAL and try to compile.
:-)

Let me abuse this list for details that belong into fb-devel: I assume
anything that reads
if (dbb->dbb_wal) {
...
}
can be yanked and wherever I face
if (!dbb->dbb_wal)
return;
the rest of the function below that point can be yanked, too, and the same
for journal. Is this correct?

Now, what we do with Dave's suggestion to review the page header? Do we
reserve those bytes for something else?

Thanks.

C.