Message: Re: Question about Openscientist. Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

None Re: Question about Openscientist. 

Keywords: OpenScientist installation structure
Forum: Analysis
Re: None Question about Openscientist.
Date: 08 Apr, 2002
From: Guy Barrand <Guy Barrand>

There are multiple questions here.

RedHat-7.x binary installation :
 The binary tarballs (zip and rpm) for Linux had been built on a 
RedHat-6.1 and clearly there are forward compatibility problems. 
I have put a note under the OpenScientist download page for a 
diagnostic and a possible solution. 

 See :

 Right now (08/04/2002) I have no RedHat-7.x at hand at LAL 
to rebuilt at least the .zip file. I will try to look around
CERN, I think that there are test machines to help migration
from 6.x to 7.x...

OpenScientist have so many packages....
 Right. And this is a global architecture wish.
We (it is a LAL computing team point of view, but 
also for example an LHCb one....) do not want to put everything
in one big package, mainly for question of flexibility.
 We think that this is a crucial aspect, especially 
for analysis environment where part may be discarded
(for exa Objectivity storage...) due to various
good reasons (change of technology) or bad ones (no more money !).

 You must see the packages as the "categories" of Geant4 with
the difference that here the package/category are versionned...

 OpenScientiat, having the pretention to be an interactive
environment for doing analysis, must then have at least
an histogram package, a plotting package, something to store
things, some GUI, some scripting, etc, etc, etc....
Then there are at least one package for each "computing domain".

 BUT, you must note, than compare to other project around, we
do not attempt to rewrite everything and then we rely, as much 
as possible to code written elsewhere ; for exa HEPVis, Midnight.
Oftenly, packages are only CMT repacking of other packages around

 Having subpackages permits for example to reuse the fitting
(Midnight) elsewhere without most of the rest, reuse CINT
elsewhere without the rest, etc...

and there is such few instruction for a beginner...
 here I agree, documenting a whole analysis system
is not a simple task, whatever it is. There are some
tutorial for the OnX and Lab packages. They come
with some demos and more are going to come in next
release, but despite the Doxygen for the classes
it is clear that it is not sufficient.

 Under "OpenScientist and Geant4" there is some
explanation if starting to do some AIDA analysis
for Geant4. We are working on the AIDA repository to be 
able to handle a bunch of C++ examples, etc...

 Here, people must be aware that the user support
for all that is at the beginning and that the task
is a huge one. The best is to give
feedback on obscure/unclear points, we will try to do our best
to improve (here I speak for OpenScientist but also for
AIDA itself, AIDA in Geant4, etc...).

installing from source....
 here, I need more feedback to help.

 The experience shows that 99% of problems
with the source installation came from a
bad access to "external packages" like Python, OpenGL, 
freetype, etc...

 With OpenScientist/v8 you have to take care
of the Interfaces/<external package>/cmt/requirements
file to parametrize the access to external packages.

 If that is correctly done and if the external packages
are here (oftenly people have not them on their machine !!!)
things should be ok (not straightforward but feasable).

Inline Depth:
 1 1
 All All
Outline Depth:
 1 1
 2 2
 All All
Add message: (add)

 Add Message Add Message
to: "Re: Question about Openscientist."

 Subscribe Subscribe

This site runs SLAC HyperNews version 1.11-slac-98, derived from the original HyperNews

[ Geant 4 Home | Geant 4 HyperNews | Search | Request New Forum | Feedback ]