The server status page hadn’t updated since June 16, 14:50:03 UTC, and BOINC‘s scheduler requests had only resulted in messages that the project was temporarily shut down for maintenance, so in case anyone still had work units and wasn’t just holding on to them, they couldn’t report them either. But at least the message boards still worked, the whole site usually only reporting that it was down for maintenance at the usual time, on Tuesdays. However, this week that changed, and now it seems that SETI@home truly did shut down for good, without even an announcement stating that it was going to happen, since it’s Friday and it’s yet to return from this week’s maintenance outage, and I just noticed yesterday that the response received by BOINC had changed as of Monday evening. Had last attempted a connection and received the maintenance response at 1:34:14 PM UTC on Monday, July 6, and then connected again at 11:43:39 PM UTC and the response was “Server error: feeder not running”.
The last work unit I had listed had one last user who hadn’t reported a result when the lists stopped being updated, the deadline being June 25, but there were others saying they had some left that had users with deadlines even in the second half of July, and I believe someone digging through lists found something only expiring in August as well, though I’m not sure of that anymore. And there were also some saying that a few of the most active hosts were still returning some results, after likely using tricks to hoard huge numbers. But, while the questions remained unanswered and this was therefore never confirmed, it was almost certain that all work units had already been validated, these results only serving to allow those users to gather more credit before the end, and also keep a higher recent average credit, and some were suggesting that this permanent “maintenance” reported to BOINC even though the site worked was just a way to “punish” them for trying this, albeit while also hurting those who still actually had computers working on the remaining work units. But with that work being unnecessary, nobody being able to find any work unit that wasn’t already validated among what those hosts still had in progress, and some demanding all along to have those remaining tasks canceled, there were few complaints, people just hoping that the message boards will remain operational. And the original announcement clearly stated that the site and message boards will continue to operate, so there seemed to be little reason for concern… Yet, on top of it being a rotten move to shut it down completely in this manner, without saying so or even answering those asking what’s happening, at the moment it doesn’t seem like the project managers are keeping their word even when it comes to the message boards.
On a personal level, this also means that I’ll only get to 21 years in the project in the sense that I’ll keep BOINC attached to it, but since that day will come on July 15, it’ll be just about a month after the project stopped allowing any sort of computing activity and, unless the situation changes again by then, over a week after it was taken down for good, even the message boards and much of the site going away. But, of course, I will consider myself as having stayed with the project for the full 21 years even in that case, and will keep BOINC attached to it and leave it with the highest resource share indefinitely, just in case it’ll ever come back to life… And for old times’ sake even if it won’t.