[Halld-offline] Fwd: Re: inclusion of more Root libraries

[Mark M. Ito]

fyi, a recent email exchange

-------- Original Message --------
Subject: 	Re: inclusion of more Root libraries
Date: 	Thu, 29 Mar 2012 21:11:08 -0400
From: 	Mark M. Ito <marki at jlab.org>
To: 	David Lawrence <davidl at jlab.org>, senderovich at phys.uconn.edu



David and Igor,

I would prefer to avoid getting tied more into ROOT than we already are.
My view is that ROOT is its own universe; and very nice if you have a
task that fits in that universe. Picking and choosing parts of it for
GlueX software is the slippery slope to swallowing the entire whale that
is ROOT. If it can be avoided at reasonable cost, it is a cost worth paying.

On the other hand, if there is a immediate need to be addressed, fine.
No need to be dogmatic about everything. I would always like to keep an
eye toward being ROOT-free, as long as that is reasonable.

   -- Mark

PS: Can I forward this exchange to the whole offline group?

On 03/28/2012 02:51 PM, David Lawrence wrote:
>
>  Hi Igor,
>
>   The choice of ROOT libraries was just made by what was needed at the
>  time to get things to link. At that time though, changes could be made
>  that would only affect maybe 3 people total since no one else was
>  using the code! Now, I guess the minimum one would need to do for such
>  a change is make the the Software Coordinator (Mark) aware of it in
>  case he wants to object.
>
>   Aside from that, I'm not familiar with the EG library. It looks like
>  it is there in the versions of ROOT I'm using (including the ifarm
>  machines here at JLab). So, I'd guess your first suggestion would
>  probably be fine. I'll admit there is some appeal too though of just
>  putting a utility for it in particleType.h. I think personally, I'd be
>  fine with either solution.
>
>  Regards,
>  -David
>
>  On 3/28/12 2:33 PM, Igor Senderovich wrote:
>>  Hi David,
>>
>>  One more question on BMS policy:
>>  I know we have a "particleType.h" in our sim-recon tree for particle
>>  geant code lookup, but it's useful to look up PDG codes in
>>  bggen/Pythia output. The classes in ROOT's MonteCarlo/EG are nice,
>>  but require dependence on libEG. Could I add this dependence (e.g.
>>  with "-lEG" added to MISC_LIBS in Makefile.ROOT) or was there some
>>  policy or design choice to limit dependence on ROOT libraries in
>>  sim-recon. In the latter case, should we create a "particlePDGType.h"
>>  to have our own lookup table?
>>
>>  Thanks,
>>  -Igor


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20120404/e255322e/attachment-0002.html>