It appears that LHC@Home is soon to go down for a month or so for maintenance, upgrade, funding, or whatever.

Predictor@Home has been down for a month or more and is soon to return (hopefully).

It seems that the server end could send a message that says something to the effect that "I'm down for a while" that would automatically cause the clients to check back in some relatively long time (a week?) instead of trying every hour or two.

It seems cleaner to do that. I detached from Predictor (and will detach from LHC) when I figured out that they were going to be down a while. The error messages and red "i" were really not useful and were just extra junk to wade through. Now, I have to manually check the web site to see when I should re-attach.

Just a thought.

---

When the LHC server is down, it recommends a 4 hours timeout (the client doesn't understand that yet though), maybe they will send a longer timeout when they take the project down for a while.

> When the LHC server is down, it recommends a 4 hours timeout (the client
> doesn't understand that yet though), maybe they will send a longer timeout
> when they take the project down for a while.
>
Predictor sends a 24 hour wait, however the BOINC client generates a 'countdown' message every hour.

---

BOINC WIKI

BOINCing since 2002/12/8