Below is the current state of the code, corresponding to the release v5r26 of CalRecon.OBSOLETE PAGE TO BE UPDATED SOON
Panel |
---|
title | Structure of CalRecon v5r26 |
---|
borderStyle | solid |
---|
|
|
The proposal below is trying to go one step closer to TkrRecon, and one step toward future TDS classes. The original main algorithm is splitted in two actors, one for clustering, one for corrections. The latter, CalEnergyAlg, has Above is the current state of the code. The next planned step is a step toward something more similar to TkrRecon. First of all, we will extract from CalClustersAlg a new top level algorithm called CalEventEnergyAlg, letting in the former only the clustering related stuff. CalEventEnergyAlg will have default tools which depends on the instance name (same mechanism as in TkrReconAlg). In the second daigramAlso, we show that the former kernel tool is made a package service, which is closer to its actual rolewant to transform the current kernel into a service, which any client will access as a classical Gaudi Service. This makes the class CalReconActor useless. The expected result is show below. It should be close to what will be used for DC2.
Panel |
---|
title | Proposal 1: two top level algorithmsThe DC2 structure of CalRecon |
---|
borderStyle | solid |
---|
|
Image Removed |
Panel |
---|
title | Proposal 2: the former kernel tool is made a package service |
---|
borderStyle | solid |
---|
|
Image RemovedImage Added |
If you wish more details about the arguments which have lead to the current decisions, you can have a look at those pages, which are more
or less under development :