3 files all become hung up, and won't finish downloading.
setiathome_6.03_windows_intelx86.exe 98%
seitgraphics_6.03_windows_intelx86.exe
libfftw3f-3-1-1a_upx.dll

Tried uninstalling, deleting directories, rebooting.

Any ideas.

3 files all become hung up, and won't finish downloading.
setiathome_6.03_windows_intelx86.exe 98%
seitgraphics_6.03_windows_intelx86.exe
libfftw3f-3-1-1a_upx.dll

Tried uninstalling, deleting directories, rebooting.

Any ideas.

Have tried clicking 'Retry Now' in the Transfers Tab?

Claggy

I've found a bug in 6.6. Go to the activity menu and suspend networking. Wait a few seconds, then set network to always, not to schedule if you have that set.

And then what? What is it supposed to do, or does it not do in your case? Also, which BOINC 6.6?

I suspect the original problem of not being able to download all of the Seti files was due to the connection into and out of Seti being overloaded. It should've been solved by now.

---

I've found a bug in 6.6. Go to the activity menu and suspend networking. Wait a few seconds, then set network to always, not to schedule if you have that set.

Thank you for the suggestion. It has not worked for me, though. I have a whole stack of SETI WUs waiting to upload (for well over a day). When I tried your suggestion, BOINC tells me initially that I do not have a network connection and need to configure it, which is totally misleading as my network connection is fine - I can update other projects. What it (probably) actually means is that it cannot access the SETI servers nor the project reference server, probably because they are both overloaded or because my client is seen as a DoS threat because BOINC keeps trying to upload the individual WUs over and over again.

I, too, am getting frustrated. The discussion around server communications and the way BOINC handles them has been going on since 2005. With SETI and BOINC housed under the same roof, I would have hoped for a more sensible upshot than is currently apparent.

---

Sekerob's suggestion was posted to address a download problem - nothing to do with uploads.

You cannot upload to SETI at the moment, because their upload server has been disabled by staff - as is clearly visible on the project's Server Status page (though no explanation of the need for the extended outage has been posted yet, AFAIK).

No amount of fiddling with the client at your end can restart that server at Berkeley.

With SETI and BOINC housed under the same roof, I would have hoped for a more sensible upshot than is currently apparent.

I'm not sure how you come to that conclusion, but BOINC and Seti aren't housed under the same roof. They may both be with the University of Berkeley, but do you know how big that one is? The BOINC domain server may be living in the Seti server cabinet, but it's not even using the 100Mbit pipeline for communications.

David Anderson lives and works in California, partially in the SSL building but mostly from home.
Charlie Fenton lives and works in California, does all his work from home.
Rom Walton lives in Florida, does all his work from home.

And that's your main BOINC development team. Others attributing either send their code from home or their work (other universities and companies, the world over).

It's best to step away from the thought that Seti == BOINC and BOINC == Seti. They aren't. BOINC is a whole lot more than just Seti and is being developed for all those other projects as well as Seti.

---

However, David Andseron is also listed on the SETI 'project personnel' page as a Director of SETI@home. He certainly has the power (= access codes) to work directly on SETI's servers, updating code and so on - though whether he has the authority to do so is a moot point. I would have thought that the originator of BOINC as a robust infrastructure would at least take an interest in its behaviour at the extremes of the envelope.

I would have thought that the originator of BOINC as a robust infrastructure would at least take an interest in its behaviour at the extremes of the envelope.

I'm sure he does, but as long as no-one knows what is ailing the upload server at Seti, it's all circumstantial to think he's on top of things.

Instead of seeing these people as being 24/7 busy with BOINC (and in your sense Seti), know that they are people as well. People with families and friends, other interests than 'the code', who need to eat, sleep, relax and sometimes completely be away from it all. Just the same way as everyone else does.

The libCurl part doing the communications in BOINC can only do so much anyway. When a server is down, or the connection into a project is so clogged up with computers trying to get through, it won't magically find other ways. It will back-off and try again at a later time.

So there is nothing to be frustrated about. Your computer isn't frustrated, your BOINC isn't frustrated. That's a human quality and a bad one at that. If you can't take it anymore, don't take it out on the people of the project who know there are problems and are trying to do their best to fix them. In that case just go do something else than stupidly stare at BOINC, as you can't change the situation anyway.

You only end up posting heated posts on a forum, getting moderated over something you let slip through, getting more heated and over what exactly? It's not as if you get paid millions of real-world bucks or that the world will end because you couldn't upload.

And if all those tasks trying to upload are interfering with your normal work on your computer, turn BOINC off. As and when the server is fixed, you won't get through anyway as all those other computers out there will also be trying to get through at much the same time.

---