On 12-11-2007 13:42:12 -0800, jsolderitsch wrote:
So -- the good news is my problem is not completely operator error.
But the bad news is that something is broken at a level for which there is no immediate work-around so I can't do what I set out to do -- at least with the sources I used to build: MonetDB-Mars-SuperBall-SR4
Let me know if you foresee an improvement to this situation -- say in a nightly build.
Would I be more likely to have success if my jdbc client was running on another platform -- linux or Windows instead of a Mac?
Well, not really. The problem now is to determine who is at fault. JDBC thinks it gets a too large block, but the question remains whether the server really sends a too large block or not. I've never been able to reproduce it, meaning I can't just put the debugger on it to pinpoint who is at fault here. The debug needs to be done on the server side in stream.mx, however, for that code I need Niels to be able to quickly analyse it to see if it is theoretically possible at all that too big blocks are being sent, or that maybe somewhere some bitshifting doesn't take something into account which it should have...