Missing one of my computers

This is what I do.  It's what I'm doing in this reply.

1. Right click the 'quote' link (top line of message you're responding to) and select "open in a new tab" so the reply is in a separate browser tab from the original thread.
2. In the new tab, you will see the opening tag (including the poster's name) at the start and the closing tag at the very end, all inside a text composition box.
3. In my opinion it's quite lazy to leave a longish message if you're only responding to a particular sentence so select and remove those portions that are irrelevant to your reply.  If somebody wants to read the full original message they can go to the original without having to suffer a full copy of the whole thing being shoved in their face once again :-).
4. If you are responding to multiple sentences from the original message, just leave those sentences and copy the original opening quote tag and paste that in front of each additional sentence.  Also insert a closing quote tag (type it in or paste a copy) at the end of each additional sentence.  That way it's very clear as to what you're responding to and what you're response exactly is.
5. If you get stuck on how to do something, make sure you expand the drop-down BBCODE HELP menu item that's always found below the text composition box you are using to compose your message.  There's lots of useful stuff in there.  There are 2 lines about quote tags quite close to the bottom of the multiple pages of BBCODE instructions.

Exactly what I've already suggested.

1. Leave the setting for % of cores BOINC is allowed to use at 50%.
2. Make sure you tick the setting for "Leave tasks in memory when suspended".
3. Remove the restriction that would cause BOINC to suspend if CPU use exceeds 25%.
4. Make sure you scroll to the bottom and click "Save" when finished.
5. You can also click "Update" in BOINC Manager to force the client to contact the server immediately to become aware of the changes you have made.

I would be very surprised if you could notice any effect of these changes on the performance of your main job. If you do and if these effects are unacceptable, then perhaps you should consider not using BOINC on that machine.

What's not 'concrete' about that set of commands? :-).

As I understand it (google his name for yourself) Gianfranco Costamagna is a Debian developer of some repute.  Ubuntu is a Debian derivative. I also think he may be involved in BOINC development at some level.  Without doing any research, I would tend to think that what he suggests would be the correct way to upgrade your BOINC installation.  Having said that and as I mentioned earlier, there should be nothing wrong with your current version 7.2.42.  For your very straight forward use case, you could easily leave it where it is.  I'm still running 7.2.42 on a number of machines and there doesn't seem to be much problem with that.

Do you believe everything you read on the internet? ;-) :-).  A number of things are immediately obvious.

1. The post was made in 2010.  A lot of things change in 9 years.
2. He has a total of 3 posts in all that time.
3. His very next post was an "Oops" in response to the first one.
4. He is a user with no apparent affiliation with the BOINC Devs.
5. He is obviously unhappy with the way that Ubuntu creates a separate user:group of boinc:boinc under which BOINC runs.  That way is OK for the 'set and forget' people who don't want to fiddle but quite annoying and frustrating for the tinkerers who don't have a good grasp of Unix/Linux file ownership and user and group permissions.  Even though I do have sufficient understanding to handle all that, I would still find it incredibly frustrating if I had to do things the "Ubuntu way". :-).
6. What he has produced is a mishmash of the original Berkeley shell archive installer as modified to (sort of) cope with how Ubuntu needs to have things done.  If you had a good understanding of Linux, you could get something to work along those lines.  From your comment about your Linux knowledge, you should be very wary about trying to follow this.

Those figures point to a finish of the primary calculations (~90%) in a total of just over 12.5hrs as compared to 11.5hrs that was indicated earlier.  In my experience, a total predicted at the ~30% complete stage is usually reasonably accurate so I think that after your app finished recording, maybe it did some heavier CPU use work that took some CPU time away from the science app.  If (when the task is finished and returned) you look on the website and see a big difference between the CPU time and the elapsed time (say more than an hour) you could suspect that this was going on at some point after the 30% stage probably.

You are right!!! Sorry about that!!!