I belong to 6 projects running across 6 commputers (they 2 fastest run all 6, 4 slower machines share all except Climate and Seti). The CPID on each website keeps changing/toggling between two codes. Therefore, BOINCsynergy will not properly combine my totals. I have been using BOINC for a long time. This toggling back and forth has been happening for at least 2 weeks.
I periodlically do manual project "updates" from manager to remove the "Ready to Report" results.

Question 1 is "how can I get the CPID to sync up and stay sync'd.

Question 2 is "do credits get assigned on upload or when the "ready to report" are updated?"

---

Question 1 is "how can I get the CPID to sync up and stay sync'd.

The only way that I can think of involves editing xml files... are you comfortable doing that? If not, someone may have a better idea.

Question 2 is "do credits get assigned on upload or when the "ready to report" are updated?"

That one is easy - the validator assigns the credit after a quorum is met for the work unit, and a quorum is only met after all required results have been _reported_. The upload is (possibly) to a separate server.

---

[quote]Question 1 is "how can I get the CPID to sync up and stay sync'd.

The only way that I can think of involves editing xml files... are you comfortable doing that? If not, someone may have a better idea.


I am comfortable editing XML. just give me the instructions. THanks.

---

I am comfortable editing XML. just give me the instructions. THanks.

Be warned that I have not tried this, and it may not work at all - the CPID may be _downloaded_ from the projects during an update, without reference to the values in the xml. However, that is not my understanding of how it works, where (within a single xml file) mismatched CPIDs will be 'reset' to that of the oldest-connected project. Regardless, it would be a good idea to work with only one PC at a time, and backup the BOINC folder on each before starting...

cross_project_id is stored in client_state.xml, and sent/received in sched_request_projectname.xml and sched_reply_projectname.xml. You should only need to touch client_state.xml.

I would first disable network access, or stop BOINC completely, on all systems. You don't need a PC you aren't working with connecting to the projects and changing things while you are fixing a different PC. Then I would look at the client_state.xml file on one machine, and verify that the CPID is the same for all projects on that machine. If it's not, then you would change it so that it is consistent. Run BOINC and hit "update" on all projects. (Do this even if you haven't changed anything.) Stop BOINC and verify that CPIDs are all still the same. At this point, if you check the projects websites, all CPIDs _should_ match. Note that the CPID as seen in the xml files is not the same value as that seen on the websites - I don't know if this is some security issue, or what, but there is obviously some way that one is converted to the other. All that matters is that they are the same across all projects in both places, website and xml, not that the website value matches the xml value.

Go on to the next machine, and make sure that all CPIDs there are the SAME as the one on the first machine. At some point you will find a machine that has the second CPID, that is causing the "argument" over which is correct. Whenever that machine connects, it is causing that project to be set to the 'extra' CPID - then when a 'good' PC connects, it's switching it back. If you're lucky, only one machine will be 'different'. If your luck is like mine, you'll find they're half-and-half!

Please let us know if you try this approach, and if this works for you. From my understanding of the system, it should, but there are always bits that I don't have completely "down" yet...

---

Do you use the same email addy and name for each project?

---

BOINC Wiki

Do you use the same email addy and name for each project?

Yes. I double checked. they call have the same email and full name (matching case too).

---