Re: [Amelia-development] Some AMELIA Updates

[Joao P. <jpp...@lb...>]

Hi Karol,

As you said, I've been very busy with Episode III (actually for really 
bad reasons. Jana's computer fried yesterday and I spent too much time 
trying to recover it with no success. At least was able to use Ubuntu 
from the live CD in console mode, and backup her work). So you'll have 
very few updates on AMELIA from me over the next few days.
I see you've been committing some of the changes proposed in the  last 
email. That's great! But let me get it point by point:
>
> 1) I've reworked (just slightly ;) ) the QIrrWidget again and we no
> longer have the problem with OpenGL and the main menu on Linux.
> Another advantage is that there is a "working" MacOSX port of AMELIA.
> By working I mean that my cube example works, but MacOSX locks up most
> of the time when I start AMELIA (although I had few successful runs.).
> I should be able to commit the basic code soon (just need to clean it
> up).
>   
Hooray!
> 2) We've had an undergraduate physics party, where we showed some
> demos, including AMELIA. I wrote a short documentation outlining some
> basic features of AMELIA. You might be interested in temporary using
> it as the "Getting Started Guide". It is here:
> http://hep.phys.sfu.ca/~kka34/amelia-doc/amelia.html
>   
It's a nice starting point. I think I'll use what you did, modify it 
slightly (and add some details about the Interesting Tracks table, 
detector tab and "Guided Tours") and post on the "Getting Started" 
section. I'll try to do it tomorrow.
> 3) For the above party, I had to make a few modifications and you
> might be interested in having some of them in the official version.
> Let me know if I should commit any of them. Here they are:
> - I've added a fourth tab to the tabbar that says "Hints". It
> basically listed signatures of common events as a quick reference. I
> think something like this (but more general) would be quite useful.
>   
I agree that we need something like a permanently accessible reference 
(like your "hints") but thought already about a nicer (and easy) way to 
implement it. Basically we could add a AMainView page to AGeometryFrame 
(which is basically a StackedWidget) like we're now doing for the 
tables, and activate it trough a button ("Need Help?" or "?" or 
"Hints"...) on the toolbar. It would also allow for more space available 
for the help, and could run an html page on screen. Another thing I'd 
like to implement whenever the current project allows me so, is a HUD 
class (AHudInfo) to display contextual information over AGeometry.
> - AXMLEvent now reads the "run number" and "event number" info from
> the JiveXML files and displays them in the "Event Info" graphics view
> thing.
> - Since we had a touch screen, I've added zoom in and zoom out buttons
> to control the sphere camera. Same thing for enabling selection of
> multiple tracks. You can see these buttons in the screenshots used in
> the documentation mentioned above.
>   
That's something I had already thought about before. Simple yet really 
useful. This should definitely be standard.
> 4) I'm going to start to play around with displaying the track hits
> and LAr calorimeters in the way we talked about while I was at
> Berkeley. So if you have any hints (ei: place to find good
> documentation on the JiveXML format?)  or you think I should wait
> until you are less busy, just let me know.
>
>
>   
I have a good idea about how to implement this, but no time for it for 
the time being. One thing that could be hard for you to do is to 
implement the 3D structure of the cells and pixels of the detector. This 
involves a lot more than simply understanding all the intricate details 
of the JiveXML. The .xml files only tell you the ID of the cells or 
pixels hit, and the energy deposited in the calorimeters, not *where* in 
the 3D space these cells and pixels are.
So a first step should be to implement a 3D map of these structures and 
translate it to the Irrlicht space.
I suggest you wait a little bit (until I move to CERN, which will be in 
one month) and then we can work on this together. Also the guys who 
created JiveXML and who hold the information about the 3D structure of 
the cells will be close to me then...


Cheers,

Joao


-- 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Joao Pequenao
    Mailstop 50R-6222
    Lawrence Berkeley National Laboratory
    1 Cyclotron Road
    Berkeley, CA 94720  USA
 Tel:  (510) 486-4290
 Fax:  (510) 486-4799
 Email: jpp...@LB...
 Webpage: http://pdg.lbl.gov/~pequenao
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=