Nobody knows since it only ran on UD agent which was quite rich in sending out copy after copy after copy, it knew no better. Under BOINC the grid runs 3 times more efficient, batch for batch, which is a tech' provided value from before zero redundancy.

Because of the nature RICE and HPF2 were already running as were they ZR. DDDT & FAAH followed and HCC is not possible because of the nature of the beast (see other posts discussing this).

So, HCMD-2: Toss up, expect it to run in Quorum 2 until it's data has acquired enough scientist validation. I'll dig for anything said in that field for I cannot remember this moment.

AC@H you find in the FAQs. 1 job cut up in 10 parts due result transmission size.

And for more history, WCG started out with Init Distro 4 on BOINC, changed to 3 in about 2006 and pushed on to 2 a while later.

HCMD was available only for the UD agent, therefore there was no redundancy visible for the cruncher: you got your points immediately after returning the result files (since in UD there was no next job waiting to start). That does not mean that there was no redundancy behind the curtain, and actually I think there was some but it was not a member's concern.

AC@H WUs were distributed to 10 members. Because the result file was huge each member of the quorum was returning one tenth of it and all pieces were put together after reception of all 10 parts by the servers.

Cheers. Jean.

I always use Preview to correct my mistakes, and usually I also scroll down to see if an answer has come while I was typing. That's not always enough to avoid duplicate answers if we are very well synchronized!

This time I had seen your answers but I also thought that I could bring a little more to them. So I decided to let mine go and to add my piece to this demonstration of CA responsiveness.

Cheers. Jean.