Nicholas Bohm to Satoshi: six not accepted in a row (April 6, 2009)

What rate of not-accepted did you see? I didn’t see anything unusual on my end. If you had more than, say, 4 in a row, that would be abnormal and probably a loss of network communication. If it’s scattered and less than 25%, just random bad luck. It’s normal and harmless to randomly get some per cent of not-accepted, and of course randomness can sometimes bunch up and look like a pattern.

This was the odd run (shown in context):

132 blocks      05/04/2009 09:19    Generated
168 blocks      05/04/2009 03:08    Generated
191 blocks      04/04/2009 22:08    Generated
193 blocks      04/04/2009 21:52    Generated
219 blocks      04/04/2009 16:52    Generated
230 blocks      04/04/2009 14:52    Generated
270 blocks      04/04/2009 05:48    Generated
311 blocks      03/04/2009 20:57    Generated
324 blocks      03/04/2009 18:01    Generated
330 blocks      03/04/2009 16:50    Generated
336 blocks      03/04/2009 15:54    Generated
0/unconfirmed    03/04/2009 10:10    Generated (not accepted)
0/unconfirmed    03/04/2009 08:00    Generated (not accepted)
0/unconfirmed    03/04/2009 04:43    Generated (not accepted)
0/unconfirmed    02/04/2009 21:43    Generated (not accepted)
0/unconfirmed    02/04/2009 19:09    Generated (not accepted)
0/unconfirmed    02/04/2009 16:05    Generated (not accepted)
452 blocks      02/04/2009 14:46    Generated

It could be loss of connectivity, though I wasn’t aware of it at the time.

The idea of an option to View/Hide unaccepted blocks is a good one, as well as View/Hide all generated blocks so you can more easily see incoming transactions. Seeing the unaccepted blocks is just annoying and frustrating. Everyone faces the same rate of unaccepted, it’s just a part of the process. It would probably be best to default to hide unaccepted blocks, so as not to show giving and taking away something that never was, and not show new generated blocks at all until they have at least one confirmation. It would only mean finding out you have a generated block 15 minutes later than normal, and then you still have 119 blocks to go before it matures anyway. This is on the to-do list for v0.1.6.

Sounds good to me. I assume you still have plans for multi-processor support.

NBQuote from: Nicholas Bohm on April 03, 2009, 11:39:00 AM UTC

I’m having an unusual run of failures, and thought I’d let you know in case this was of any significance.

Regards,

Nicholas