Meeting Notes for 1 October 2008 CMS Week Report from Dan K. CMS e-Lab Prototype Testbeam vs. Monte-Carlo e-Lab (What makes an e-Lab). Wiki/Blogs Upgrade of Machines at Argonne Hardware Needs

Meeting Notes for 1 October 2008

Present: Mihael, Eric, Liz, Tom J., Bob, Tom L. , Dan K.
Absent: Mike W., Randy, Kris, Joao

CMS Week Report from Dan K.

Dan talked to Dave Barney, the CMS Outreach person, about using the CMS e-Lab for a European Master Class project. Previous Master Classes have used LEP data. He is planning a Master Class in the 1st week of April where he will show Monte Carlo data from CMS and ATLAS. Dave is also hoping to show the Amelia software. It is a 2-3 day workshop where teachers work with a web browser, but would not download Amelia to the individual machines. The presenter could put it on his machine and demo it. Dave would like to test out the CMS e-Lab on the 10% level in the April Master Class. ELabs will have to have an e-Lab up and running using Monte Carlo data. We could do this on a non-grid system.

Action Item: Mihael will start working on the current Testbeam e-Lab. (He estimates he just needs to move the data and do some small changes. )

CMS e-Lab Prototype

Jill Weigler has made a wiki version of a CMS Monte-Carlo e-Lab, hosted at wikispaces. It could be refined there and then imported to our site. Doug Berry is also working on this. Jill wants to work on this. Holding tank before they move it to an e-Lab. Use the wiki for composing/editing. Kernel of the page can be extracted and put into pages with our navbars and decorations.

Eric is thinking there might be a way to tie them together. Tom L. will send the URL to Eric. Eric did a proof of concept test where he used mediawiki where he strips away all the extra stuff and accesses the content from an e-Lab page. Eric would have to export data from wikispaces to put into mediawiki. Eric is suggesting that we can add a lot of content from the wiki. Hardwire the pages from the wiki pages.

Miheal is in favor of making a small change over time. Mihael’s highest priority is to make the current e-Labs them good for the analysis, it may not be the highest priority. The references, glossaries might be a good starting point. Putting wikis all over the place might require too much effort. CMS Monte-Carlo could be a good target for experimenting with this. Jsp pages have to get the content. Use include files that get the content from the wiki. People can work on the content through the wiki.

Testbeam vs. Monte-Carlo e-Lab (What makes an e-Lab).

We had a long discussion about whether the analysis of Monte-Carlo data should be a completely new e-Lab or if it could be thought of a new study with a CMS e-Lab that had all possible CMS-related analyses. Tom L. argued that it should be a completely new e-Lab because it is conceptually very different from the Testbeam e-Lab. It would require different learner outcomes, pretest-posttest,milestones and references. The European Master Class will need these ready.

Eric asked whether John Kerr’s magnetometer work would be a study or new e-Lab for LIGO?

Cosmic – same body of data; multiple analysis tools (studies) CMS – different data; studies are fundamentally different for the different data.

Put the information together to make a coherent experience so we can make a Monte Carlo e-Lab. Do we need an infrastructure that allows a student to move from testbeam studies, to Amelia and then to statistical studies with CMS LHC?

CMS Testbeam – get to know your detector; CMS - LHC – look at data.

Jean Young plans for year 1: CMS and LIGO are in the same cycle. LIGO will be first. Year 2 – we will be testing the CMS-LHC. Jean probably does not know about the Monte-Carlo e-Lab. Fundamentally different between Testbeam and CMS-LHC. Eric – which product are we putting out?

LIGO – multiple e-Labs for each of the topics (weather, seismic)? Or is it one e-Lab with multiple investigations.

We may not get clarity on what makes an e-Lab? until we use it some more. Are we tying content learner outcomes to one e-Lab?

Can we think of a body of data as a separate e-Lab?

Tom J. – how does moving from Testbeam to Collision data change the type of study you do?

Testbeam might be comparable to Performance study for CMS.

We need this as well for our audience. (Eric)

Wiki/Blogs

Marge has sent out a summary of all the design document for wikis, blogs, etc. There may be more than one wiki: one for the public, one for teachers.

Upgrade of Machines at Argonne

Rick – is going to start the upgrades; he will let us know.

Hardware Needs

Grant money came in so we could by new hardware. Mihael says it is good for now.

-- Main.LizQuigg - 06 Oct 2008