I'm down to just a single machine, now, as work cut back a bit on the flexibility to "burn-test" new servers for a month or two before putting them into production, plus we've had to be a bit more restrictive on power-usage in our test-lab!

That single machine is chugging away quite nicely, though, as a dual 6-core Xeon.

I think that would make an impressive resume addition and a positive talking point.

One thing I could suggest would be a kind of Moores law penalty. I feel sad for all the people who were crunching a decade ago on single, dual or even quad core machines, religiously tuning them for tens of thousands of points per day. These were the real trailblazers and it seems far too easy now to overtake these legends. An average android phone has 4 cores and consumes 5w, a few of these and you can overtake years of run time effort. At the other end of the scale I've got a PC (admittedly a large one, cue Viz joke ) with 96 cores that'll be the same as 40 PC's back in the day the team was formed, for considerably less personal effort.
Perhaps a league table with this bias built in could level the playing field a bit . League of Extraordinary Crunchers... lol
So we'll need to agree on a scale, perhaps there's something out there already!? and also try and extract or recreate the historic data for the team members - maybe that's not possible. But it feels the fair thing to do - thoughts?

Dave,

I've been thinking about your request for suggestions regarding functionality changes on your web site. But before I mention any, I want to say how much I appreciate what we already have. I think it's a great way of attracting people to the team, and of letting the team know how both the team and the individuals are doing. I take a peep at all the tabs on most days.

The one reason I do sometimes visit BOINCStats is that it doesn't just rely on today's data, or all-time data. It keeps track of averages and movements over time-scales of a week, a month and 60 days too (IIRC). When a cruncher gets new kit, all-time averages don't give them an indication of how the new kit's doing. Seven-day averages do, and that's what you have on your site. Great! But averages over longer time-scales even out glitches, holidays and the like. At the other end, an all-time maximum isn't very helpful either because system hiccoughs, or the addition/loss of temporarily available machines, can seriously skew these one-time numbers. This is something that you do have on your site, and I would like to see them replaced by a max/min over a shorter time-scale -- perhaps a month?

I'm rambling a bit, because different things are better/worse in different circumstances. I'm just requesting a little thought about what time-scales to use for averages and max/min data. Maybe even make it available over a couple of different time-scales.

This data could perhaps also be used to add individual overtakes in a similar manner to team overtakes, and I'd like to see it for country and world-wide ranking changes too (among team members -- it would be way too much data to do it for everyone!). BOINCStats only does overtakes for points, not for run-time or results. But it also gives the number of places gained/lost in the last week and month, not just that day.

Thanks for listening, and input from others on this would be great!

For the first time in my 12yr marriage I'm not making Christmas dinner, going to my father-in-laws for it. As usual I'll be doing a fair amount of driving over the next few days visiting relatives etc, so not much drinking for me.
Now that my son is 6yr old, we have some learning toys covering basic physics, chemistry and electronics, start them young eh!
Also got a Scalextric set which will be fun for all of us.

Have a great Christmas everyone and stay safe.

You are quite right Apis, I believe that Mac's OS 10 and up is based on FreeBSD.
I cannot be certain that they gave anything back to FreeBSD.