|[AMD Users] IMDb Addict|
|I guess I will be getting flamed for this because everyone asks this, but what the hell.|
Does anyone (Stephen?) have any clue when v5 will be comming out? How long do you think the test phase will be necessary? I can't wait to contribute te the 'real' design (I know this phase isn't bullshit, but it just doesn't feel like the real thing, y'know ).
If this question was allready answered, just say which tread I should read.
|Yeah I kinda feel the same way, is there a muon site with upto-date news about the project that I don't know about.. I'm running afew pc's 24/7 and I don't really know what I am accomplishing.|
What will be accomplished in v5 what we haven't done in 4.31c?
|Also the DPC boys are waiting for the new client. A lot of them are changing to other projects.|
|Well I'm hoping to be working full-time on v5 from September, because that is when my official permanent placement starts. There are some peculiar (but occasional) bugs in v4.31c that might warrant a bugfix version earlier on during this summer vacation.|
For now I'd recommend going onto other projects for the time being (at least on the majority of your machines), and I'll of course update here when the v5 stuff is sorted out.
I was pretty-much out of action for a couple of weeks after finishing university, but in the last two weeks I've started programming more again. I've been uploading a few things I had around to other parts of the site, and also overhauled the front page a bit. After a few more odds and ends, I'll get the new Muon stats page (for v4.3+, designed to be v5-compatible) up and then maybe look at the client again.
But from September it's a question of - if v5 doesn't come out and work - I don't have a job
Oh; and it was 32°C here yesterday, which is not great to work in - late night would be best, but I'm with my parents so have to keep vaguely sociable hours. A dust-devil formed in the road outside my house today, which caused several doors to slam and a car alarm to go off! Tomorrow promises large thunderstorms in places...
Today's weather in %region is Sunny/(null), max. temperature #NAN°C
|Maybe any news ??|
|Right now I'm writing a report - or rather updating the report that was at http://stephenbrooks.org/ral/report (you'll notice that page now supports a list), which I wrote after about version 4.11, to reflect the results of the rest of the project up until about now. The finished report ought to cover everything before what I call version 5, which will be part of a different design. So that will take probably until next week...|
Today I started working on the Muon code itself again, gradually doing things on the to-do list for v4.31d, like allowing user-specified designs to be inputted via queue.txt (without weird checksumming 'cheating' that was possible before).
Also some time there's going to be a new optimisation called "SolenoidsTo15cm", which will be before v5, and unlike the current "SolenoidsOnly", it will have a more sensible physical meaning. I've locked the last solenoid to 15cm radius - it'd be good if I could get results from that for this report-update I'm doing too. But I've got to test the thing first (actually right now my workstation is plugging away at it, so I can see if it's doing the right thing).
It doesn't make any sense: that's why they call it "virtual"
|Look what I dug up out of the past |
I was just curious what is thought about information gleaned from runs up through V4.33a. It seems like we're not making a lot of new prgoress toward yield, meaning we seem to be at or very near optimal for this design. How does the yield simulated through current results compare to expected data? Or maybe there wasn't a good equivalent estimate that compares against the solenoidsto15cm that's running now? Will the chicane be re-introduced to what we are running now before particle accelerators are introduced? Is a chicane still planned for the project? And, finally, is there a timeframe in mind for the next modeling software release?
|Well as you know, SolenoidsTo15cm eventually came out (October 2003?) and the results from that allowed me to finish that 2nd version of my report by December. What I find interesting about the results from it right now are that they're still increasing, albeit slowly. For a month while I was finishing the report they were constant! I will at some point have a look at how today's approaching-9.9% scores compare to the under-9.7% ones I wrote about in my report, because despite the apparently-small change in yield, something interesting might be going on there, as I was fully expecting a complete standstill!|
If the optimisation really has finished, the only "record breakers" will be statistical fluctuations around the best mean, and succesive records will get rarer with exponentially-increasing periods between them. What I'm seeing on the graph is a comparatively rather more regular increase, though slowing down, which suggests we're rolling uphill to a standstill rather than having already stopped.
There wasn't really any precise expected data for how well this simulation would turn out, as this particular thing hasn't been simulated before. The only direct comparison is the, er, 3.1% that the baseline design I was already given got.
Here is where the plan has changed a bit. A while ago I was told the chicane had been dumped in favour of a cooling ring design, but then I realised that (a) the first components I would have to put in for the cooling ring would be RF cavities, which could nicely complete the 3-section chicane design and (b) in order to sensibly compare the cooling ring option with the chicane one, I'd better simulate/optimise both on Muon1 anyway. Now I've said that, the guy who came up with both designs is interested in the chicane option again.
So I suspect the next _major_ release will be v4.4, which will include a new optimisation script including the chicane plus linac. At a guess I'd say late March for that. Then v5 will have this cooling ring too, perhaps by the end of May? If all goes to plan.
HB Pencils, also sold as "Moron's Choice" Graphite Cigars.
|Thanks Stephen! You have a real knack for explaining things in a way I can understand|
|I agree also! Thanks Stephen.|
Life the Universe and Everything!
|1 year and 13 days later...|
Still no V5 in sight?
Or did you meant May 2005 or even 2006?
|Well at least I got the v4.4x series out last year, if not v5. It turned out that when I started investigating the "muon cooling" last autumn, it wasn't really quite as simple as I thought, and I have some data now that I need to process in various ways to see what the most accurate way of modelling it is.|
That's going to happen after I've sorted some stuff out with the rod target. I'm going to be giving a talk on that at CERN later this month, so need to summarise and complete my investigations. What it means for you is that the 2.2GeV pion distribution I've been using thus far wasn't the optimal one, and I will introduce a new lattice that will use a 15-30 GeV one. This should give better yields.
Right now I'm just about to use the optimised ChicaneLinacB and PhaseRotB results to plot some charts of the dynamics to go in the talk...
|kk thank you for your time to answer my posting.|
Is there something like a mailing list now were we can subscribe, so we get informed if version 5 got finished?
|Whatever happened to v5? |
(If I remember correctly, you somewhere posted a graphic of the proposed cooling ring).
|Other projects in Muon1 are possible but that particular cooling ring I don't think would have worked. I did some tests of it and eventually found the beam would not go around through the fieldmaps stably.|
Updates to the client are at least six months away. The examiners have accepted my thesis (good news) but want more, fairly substantial, corrections done to it first (bad news). They will take a month or more. I've got a conference (HB2010) at the end of September for which I'm designing an accelerator - actually part of a proton accelerator with possible applications to the neutrino factory, so before the section Muon1 simulates.
Those two things put me in October before I can think straight about Muon1. I need to compile a new build, test the results and hopefully get some of you to do some beta testing and benchmarking. For some reason people have complained that newer builds are slower than the older ones and I want to figure that out. Also possible is converting the code to compile with Dev-C++ (a MinGW port of gcc) rather than LCC. So it'll take a while until I can just get back to releasing a stable working version (by end of 2010 hopefully).
From there, it starts to get more interesting. It would be good to update the solenoid field function to allow more different coil geometries (and also simulate/optimise the "baseline" official design, which has overlapping solenoids in places). The optimiser has been getting "fairly good" results, but it seems to often get stuck or not progress as much as I would have expected from similar lattices - I'm wondering if I need to do something to discourage it from getting stuck too quickly in one place, which might entail a rather different algorithm.
|Update— The examiners have accepted my thesis and I've just finished the HB 2010 paper and poster. All this happened more-or-less on the same day! I'll be in Switzerland from the 26th until October 1st.|
The new reports can be found here. My time estimate for client updates remains where it is (so ~4 months now).