I think we're getting close to considering merging Ola's Rej-based code
into trunk, since although there's more perf work to do on it, it
already eliminates the byte[]/char[] conversion overhead.
Right now it's not compiling for me, so look into that if you can Ola.
But what about some discussion on this? Ola's numbers for most apps show
it helps quite a bit, but there's some areas where it's much slower than
what we have now. Ola is going to try to quantify what's good and bad so
we have a better picture.
Thoughts?
- Charlie
---------------------------------------------------------------------
To unsubscribe from this list please visit:
http://xircles.codehaus.org/manage_email
