Consider default pageSize value in toPullStream

[arj03]

I was looking at some numbers today and noticed that there was a significant overhead of using toPullStream in some cases. The following query takes 2.4 seconds:

pull(
      ssb.db.query(
        where(
          and(
            author('@QlCTpvY7p9ty2yOFrv1WU1AE88aoQc4Y7wYal7PFc+w=.ed25519')
          )
        ),
        toPullStream(),
        pull.collect((err, results) => {
            ....

while:

    ssb.db.query(
      where(
        and(
          author('@QlCTpvY7p9ty2yOFrv1WU1AE88aoQc4Y7wYal7PFc+w=.ed25519')
        )
      ),
      toCallback((err, results) => {
            ....

takes 1.2s.

It turns out that toPullStream actually does pagination internally in all cases. Tweaking this number made quite a difference. So I'm wondering how this number came about? And if it should be something like maybe 1000 instead?

[staltz]

Hmm, yeah 20 might be too low. I think I came up with the number arbitrarily.

[arj03]

Remember this one? :)

[staltz]

Hmmm